Removal of References to Credit Ratings From Regulation M, 39962-39994 [2023-12591]

Agencies

• SECURITIES AND EXCHANGE COMMISSION

[Federal Register Volume 88, Number 117 (Tuesday, June 20, 2023)]
[Rules and Regulations]
[Pages 39962-39994]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2023-12591]



[[Page 39961]]

Vol. 88

Tuesday,

No. 117

June 20, 2023

Part III





 Securities and Exchange Commission





-----------------------------------------------------------------------





17 Parts 240 and 242





Removal of References to Credit Ratings From Regulation M; Final Rule

Federal Register / Vol. 88, No. 117 / Tuesday, June 20, 2023 / Rules 
and Regulations

[[Page 39962]]


-----------------------------------------------------------------------

SECURITIES AND EXCHANGE COMMISSION

17 CFR Parts 240 and 242

[Release No. 34-97657; File No. S7-11-22]
RIN 3235-AL14


Removal of References to Credit Ratings From Regulation M

AGENCY: Securities and Exchange Commission.

ACTION: Final rule.

-----------------------------------------------------------------------

SUMMARY: The Securities and Exchange Commission (``Commission'') is 
adopting rule amendments to implement section 939A(b) of the Dodd-Frank 
Wall Street Reform and Consumer Protection Act of 2010 (``Dodd-Frank 
Act''), which requires, among other things, that the Commission remove 
from its regulations any references to credit ratings and substitute in 
their place alternative standards of creditworthiness. The amendments 
remove certain existing rule exceptions that reference credit ratings 
for nonconvertible debt securities, nonconvertible preferred 
securities, and asset-backed securities and substitute in their place 
new exceptions that are based on alternative standards of 
creditworthiness. These substitutes include exceptions for 
nonconvertible debt securities and nonconvertible preferred securities 
(together, ``Nonconvertible Securities'') of issuers who meet a 
specified probability of default threshold, as well as exceptions for 
asset-backed securities that are offered pursuant to an effective shelf 
registration statement filed on a certain form that is tailored to 
asset-backed securities offerings. The Commission is also adopting an 
amendment to a recordkeeping rule applicable to broker-dealers in 
connection with their reliance on an exception involving probability of 
default determinations.

DATES: Effective date: The final rules are effective on August 21, 
2023.

FOR FURTHER INFORMATION CONTACT: Jessica Kloss, Attorney-Adviser, Laura 
Weber, Branch Chief, Josephine Tao, Assistant Director, Office of 
Trading Practices, or Carol McGee, Associate Director, Office of 
Derivatives Policy and Trading Practices, at (202) 551-5777, Division 
of Trading and Markets, Securities and Exchange Commission, 100 F 
Street NE, Washington, DC 20549-7010.

SUPPLEMENTARY INFORMATION: The Commission is amending the following 
rules adopted under the Securities Exchange Act of 1934 (``Exchange 
Act''):
---------------------------------------------------------------------------

    \1\ 17 CFR 242.100 through 242.105. Regulation M is also adopted 
under the Securities Act of 1933 (``Securities Act'') and under the 
Investment Company Act of 1940.

------------------------------------------------------------------------
          Commission reference                     CFR citation
------------------------------------------------------------------------
Rule 17a-4.............................  17 CFR 240.17a-4
------------------------------------------------------------------------
Regulation M: \1\
  Rule 100.............................  17 CFR 242.100
  Rule 101.............................  17 CFR 242.101
  Rule 102.............................  17 CFR 242.102
------------------------------------------------------------------------

Table of Contents

I. Introduction
II. Discussion of the Final Rule Amendments
    A. Rule 101(c)(2) of Regulation M: Implementing Section 939A(b) 
in Certain Exceptions for Distribution Participants
    1. Rule 101(c)(2)(i): Nonconvertible Securities of Issuers Who 
Meet a Specified Probability of Default Threshold
    2. Rule 101(c)(2)(ii): Asset-Backed Securities Offered Pursuant 
to an Effective Shelf Registration Statement Filed on Form SF-3
    B. Rule 102(d)(2) of Regulation M: Implementing Section 939A(b) 
in Certain Exceptions for Issuers and Selling Security Holders
    1. Rule 102(d)(2)(i): Nonconvertible Securities of Issuers Who 
Meet a Specified Probability of Default Threshold
    2. Rule 102(d)(2)(ii): Asset-Backed Securities Offered Pursuant 
to an Effective Shelf Registration Statement Filed on Form SF-3
    C. Exchange Act Rule 17a-4(b)(17): Adding a Record Preservation 
Requirement for Broker-Dealers in Connection With Probability of 
Default Determinations
III. Other Issues
IV. Other Matters
V. Economic Analysis
    A. Baseline
    1. The Investment Grade Fixed Income Market
    2. The Investment Grade Exception
    B. Benefits of the Amendments
    C. Costs of the Amendments
    1. Costs Associated With Obtaining the Estimate of the 
Probability of Default
    2. Costs Associated With Maintaining Records Related to the 
Probability of Default Estimation
    3. Costs Associated With Structural Credit Risk Model Based 
Probability of Default Being an Imperfect Proxy for Creditworthiness
    4. Costs Associated With Asset-Backed Securities' Amendments
    5. Indirect and Other Costs of the Amendments
    D. Efficiency, Competition, and Capital Formation
    E. Reasonable Alternatives
    1. Alternative Threshold for Probability of Default
    2. Exception Based on Security Characteristics
    3. Exception Based on Issuer Characteristics
    4. Exception Based on Issuer and Issue Characteristics
    5. Elimination of the Investment Grade Exception From Rule 101
    6. Alternative for Asset-Backed Securities
    7. Alternatives for Rule 102 Exception
    8. Alternative for the Record Preservation Requirement
VI. Paperwork Reduction Act
    A. Respondents
    B. Use of Information
    C. Collection of Information
    1. Burden and Cost Estimates Related to the Rule 101 Amendments
    2. Burden and Cost Estimates Related to the Rule 17a-4 
Amendments
    D. Collection of Information Is Mandatory
    E. Confidentiality of Responses to Collection of Information
    F. Retention Period for Record Preservation Requirement
VII. Regulatory Flexibility Act Certification
Statutory Authority

I. Introduction

    To reduce reliance on credit ratings,\2\ section 939A(b) of the 
Dodd-Frank Act requires the Commission, among other things, to ``remove 
any reference to or requirement of reliance on credit ratings'' and 
``substitute in such regulations such standard of credit-worthiness'' 
as the Commission determines to be appropriate for those 
regulations.\3\ In making such a determination, the Commission must 
seek to establish, to the extent feasible, uniform standards of 
creditworthiness for use by the Commission, taking into account the 
entities it regulates and the purposes for which those entities would 
rely on such standards of creditworthiness.\4\
---------------------------------------------------------------------------

    \2\ See Joint Explanatory Statement of the Committee of 
Conference, Conference Committee Report No. 111-517, to accompany 
H.R. 4173, 864-79, 870 (June 29, 2010).
    \3\ See Public Law 111-203, sec. 939A(b), 124 Stat. 1376, 1872-
90 (2010). Section 939A of the Dodd-Frank Act also requires the 
Commission to ``review any regulation issued by [the Commission] 
that requires the use of an assessment of the credit-worthiness of a 
security or money market instrument and any references to or 
requirements in such regulations regarding credit ratings.'' Public 
Law 111-203, sec. 939A(a). The Commission must transmit a report to 
Congress upon the conclusion of the review required in section 
939A(a). Public Law 111-203, sec. 939A(c); see U.S. Securities and 
Exchange Commission Staff, Report on Review of Reliance on Credit 
Ratings: As Required by Section 939A(C) of the Dodd-Frank Wall 
Street Reform and Consumer Protection Act (2011), available at 
https://www.sec.gov/news/studies/2011/939astudy.pdf. Staff reports, 
Investor Bulletins, and other staff documents (including those cited 
herein) represent the views of Commission staff and are not a rule, 
regulation, or statement of the Commission. The Commission has 
neither approved nor disapproved the content of these documents and, 
like all staff statements, they have no legal force or effect, do 
not alter or amend applicable law, and create no new or additional 
obligations for any person.
    \4\ Public Law 111-203, sec. 939A(b).
---------------------------------------------------------------------------

    Regulation M, which is a set of prophylactic anti-manipulation 
rules

[[Page 39963]]

that is designed to preserve the integrity of the securities trading 
markets as independent pricing mechanisms by prohibiting activities 
that could artificially influence the market for an offered security, 
contains references to credit ratings in identical exceptions under 17 
CFR 242.101 (``Rule 101'') and 242.102 (``Rule 102'') for investment 
grade Nonconvertible Securities and asset-backed securities.\5\ The 
Investment Grade Exceptions are two of several exceptions to Rule 101's 
and Rule 102's general prohibitions: in connection with a distribution 
\6\ of covered securities,\7\ distribution participants,\8\ issuers, 
selling security holders, and their affiliated purchasers are 
prohibited from, directly or indirectly, bidding for, purchasing, or 
attempting to induce any person to bid for or purchase, a covered 
security \9\ during the applicable ``restricted period.'' \10\ These 
prohibitions exist to protect the integrity of the offering process by 
precluding activities that could artificially influence the market for 
the offered security.\11\
---------------------------------------------------------------------------

    \5\ See 17 CFR 242.101(c)(2) (``Rule 101(c)(2)''), 17 CFR 
242.102(d)(2) (``Rule 102(d)(2)''). Both of these rules except 
Nonconvertible Securities and asset-backed securities that are rated 
by at least one nationally recognized statistical rating 
organization, as that term is used in 17 CFR 240.15c3-1 (``Rule 
15c3-1''), in one of its generic rating categories that signifies 
investment grade. Throughout this release, each exception in Rule 
101(c)(2) or Rule 102(d)(2) that references credit ratings is 
referred to as an ``Investment Grade Exception,'' and, together, 
those exceptions are referred to as the ``Investment Grade 
Exceptions,'' as applicable.
    \6\ See 17 CFR 242.100(b) (``Rule 100(b)'') (defining 
``distribution'' as ``an offering of securities, whether or not 
subject to registration under the Securities Act, that is 
distinguished from ordinary trading transactions by the magnitude of 
the offering and the presence of special selling efforts and selling 
methods'').
    \7\ See 17 CFR 242.100(b) (defining ``covered security'' as any 
security that is the subject of a distribution or any reference 
security, and ``reference security'' as a security into which a 
security that is the subject of a distribution may be converted, 
exchanged, or exercised or which, under the terms of the subject 
security, may in whole or in significant part determine the value of 
the subject security'').
    \8\ See 17 CFR 242.100(b) (defining ``distribution participant'' 
as any ``underwriter, prospective underwriter, broker, dealer, or 
other person who has agreed to participate or is participating in a 
distribution'').
    \9\ See 17 CFR 242.100(b) (defining ``covered security'' as any 
security that is the subject of a distribution or any reference 
security, and ``reference security'' as a security into which a 
security that is the subject of a distribution may be converted, 
exchanged, or exercised or which, under the terms of the subject 
security, may in whole or in significant part determine the value of 
the subject security).
    \10\ See 17 CFR 242.100(b).
    \11\ Anti-Manipulation Rules Concerning Securities Offerings, 
Release No. 34-38067 (Dec. 20, 1996) [62 FR 520 (Jan. 3, 1997)] 
(``Regulation M Adopting Release''), 62 FR 521. Rule 101's 
prohibitions apply to distribution participants and their affiliated 
purchasers, while Rule 102's prohibitions apply to issuers, selling 
security holders, and their affiliated purchasers.
---------------------------------------------------------------------------

    In adopting the Investment Grade Exceptions, the Commission stated 
that certain securities and activities should be excepted to allow for 
activities that are necessary for the distribution to occur; to limit 
adverse effects to the trading market that could result from these 
prohibitions absent such exceptions; and to permit conduct that is not 
likely to have a manipulative impact.\12\ The Investment Grade 
Exceptions were premised on the principle that investment grade 
Nonconvertible Securities and asset-backed securities are less likely 
to be subject to the type of manipulation that Regulation M seeks to 
address because they are largely fungible and trade primarily on the 
basis of yield and creditworthiness (traditionally measured by credit 
ratings),\13\ rather than the identity of the particular issuer.\14\
---------------------------------------------------------------------------

    \12\ See Trading Practices Rules Concerning Securities 
Offerings, Release No. 34-37094 (Apr. 11, 1996) [61 FR 17108 (Apr. 
18, 1996)] (``Regulation M Proposing Release''), 61 FR17111, 17120.
    \13\ See Regulation M Adopting Release, 62 FR 527; see also 
infra note 38 (discussing how the ability to substitute similar 
securities in the market for the security in distribution limits the 
potential impact a covered person might attempt to exert on the 
market and distribution of such security). The Investment Grade 
Exceptions trace back to a 1975 Commission staff no-action position 
regarding Exchange Act Rule 10b-6, the predecessor to Rules 101 and 
102 of Regulation M. See Letter from Robert C. Lewis, Assoc. Dir., 
Div. Mkt. Reg., SEC, to Donald M. Feuerstein, Gen. Partner & 
Counsel, Salomon Bros. (Mar. 4, 1975) (emphasizing the following 
representations from the lead underwriter-requestor in taking its 
position: (1) ``because the non-convertible bonds of particular 
issuers are not considered unique and because of the concept of 
relative value, it is simply not possible to manipulate the price of 
a corporate bond that has broad investor interest,'' and (2) 
purchasing activities in such securities generally are ``unlikely to 
materially affect the price of [a nonconvertible debt security being 
offered] because of the availability of large amounts of securities 
of other issuers which have comparable quality yield [spreads]''). 
For a further discussion of the history of the Investment Grade 
Exceptions, see Removal of References to Credit Ratings From 
Regulation M, Release No. 34-94499 (Mar. 23, 2022) [87 FR 18312 
(Mar. 30, 2022)] (``Proposal''), 87 FR 18315.
    \14\ Regulation M Proposing Release, 61 FR 17112.
---------------------------------------------------------------------------

    In accordance with section 939A(b)'s requirements, in 2022, the 
Commission proposed rule amendments to remove the Investment Grade 
Exceptions and substitute them with new exceptions that are based on 
alternative standards of creditworthiness.\15\ The Commission proposed 
to except from Rule 101 (1) Nonconvertible Securities of issuers who 
meet a specified probability of default threshold,\16\ and (2) asset-
backed securities that are offered pursuant to an effective shelf 
registration statement filed on Form SF-3.\17\ The Commission proposed 
to eliminate, without replacing, the Investment Grade Exception from 
Rule 102.\18\ The Commission also proposed to amend 17 CFR 240.17a 
(``Rule 17a-4''), specifically paragraph (b) of Rule 17a-4 (``Rule 17a-
4(b)''), to require broker-dealers to preserve written probability of 
default determinations pursuant to Rule 101.\19\
---------------------------------------------------------------------------

    \15\ See Proposal, 87 FR 18316-24. The Commission previously 
proposed two alternatives to the Investment Grade Exceptions. See 
Removal of Certain References to Credit Ratings Under the Securities 
Exchange Act of 1934, Release No. 34-64352 (Apr. 27, 2011) [76 FR 
26550 (May 6, 2011)]; References to Ratings of Nationally Recognized 
Statistical Rating Organizations, Release No. 34-58070 (July 1, 
2008) [73 FR 40088 (July 11, 2008)] (``2008 Proposing Release''), 73 
FR 40095-97. The Commission did not adopt any rule amendments with 
regard to the Investment Grade Exceptions based on either of these 
proposals.
    \16\ See Proposal, 87 FR 18317-19.
    \17\ See Proposal, 87 FR 18321-22.
    \18\ See Proposal, 87 FR 18323-24.
    \19\ See Proposal, 87 FR 18324-25.
---------------------------------------------------------------------------

    The Commission received comments from an industry group, a data 
provider, nonprofit organizations, and individuals.\20\ Commenters 
broadly recognized and acknowledged the objectives of the Proposal. 
Some commenters, including individual commenters, provided general 
support for the Proposal \21\ and stated that reliance on credit 
ratings is outdated \22\ and can be harmful to investors or the 
markets.\23\ Another commenter

[[Page 39964]]

supported the Proposal and stated that its adoption will lead to 
increased market competition.\24\ Some commenters opposed or expressed 
concerns about the Proposal, and offered certain recommendations with 
regard to particular aspects of the proposed rule amendments,\25\ which 
are addressed below, in Parts II.A through C. After reviewing and 
carefully considering the public comments and recommendations, and in 
accordance with the requirements of section 939A(b), the Commission is 
adopting final rule amendments, with targeted modifications to address 
comments received and to streamline and clarify the rule text from the 
Proposal. As discussed below in Parts II.A and II.B, the Commission 
believes that its original basis for excepting investment grade 
Nonconvertible Securities and asset-backed securities from Rules 101 
and 102 continues to apply to the securities that are captured by the 
amendments' substitute standards of creditworthiness.
---------------------------------------------------------------------------

    \20\ Comments received in response to the Proposal are contained 
in File No. S7-11-2022, available at https://www.sec.gov/comments/s7-11-22/s71122.htm.
    \21\ See, e.g., Letter from Chris Carr (May 19, 2022); Letter 
from Daniel Kuo (May 19, 2022); Letter from Fred Carter (May 19, 
2022); Letter from Biren Patel (May 19, 2022); Letter from Robert 
Tso (May 23, 2022); Letter from Stephen W. Hall, Legal Dir. & Secs. 
Specialist, Better Mkts, Inc., to Vanessa A. Countryman, Sec'y, SEC 
(May 23, 2022) (``Better Markets Letter''), at 3.
    \22\ See, e.g., Letter from Alexandra Merz (May 18, 2022) 
(``Merz Letter''); Letter from Gerhard Krohmer (May 19, 2022); 
Andriy Granovsky (May 19, 2022); Letter from Jason Smith (May 20, 
2022); Letter from Craig Faison (May 20, 2022); Letter from Jaymin 
Patel (May 20, 2022); Letter from Paul K. Sacco (May 21, 2022); 
Letter from David Navari (May 22, 2022) (``Navari Letter''); Letter 
from Jim Protsenko (May 24, 2022); Letter from John Hall (May 26, 
2022); Letter from Andrew Macafee (May 30, 2022). The Commission 
also received two anonymous comments on May 19, 2022, both of which 
stated that credit rating agencies ``have become obsolete.''
    \23\ See, e.g., Merz Letter; Letter from Robert Long (May 19, 
2022); Letter from James R. Brown (May 19, 2022); see also Letter 
from William Desavigny (May 19, 2022); Letter from Kevin Price (May 
19, 2022); Letter from Jason MacKenzie (May 19, 2022); Letter from 
James Zarbock (May 19, 2022); Letter from Carsten Hensch (May 19, 
2022); Letter from Thomas Sutton (May 19, 2022); Letter from Harold 
VanPatten (May 19, 2022); Letter from Aaron Grimshaw (May 19, 2022); 
Letter from Andre M (May 19, 2022); Letter from Andrew Oshea (May 
19, 2022); Letter from Steven Calvino (May 19, 2022); Letter from 
Dennis Smith (May 19, 2022); Letter from Devin Dasbach (May 19, 
2022); Letter from Mark A. Fritzke (May 19, 2022); Letter from 
Cameron Beebe (May 19, 2022); Letter from Nick Parasiris (May 19, 
2022).
    \24\ See Letter from Jacob Rajan (May 19, 2022).
    \25\ See, e.g., Letter from Joseph Corcoran, Managing Dir. & 
Assoc. Gen. Counsel, Secs. Indus. & Fin. Mkts. Ass'n, to Vanessa 
Countryman, Sec'y, SEC (May 23, 2022) (``SIFMA Letter 1''); Better 
Markets Letter.
---------------------------------------------------------------------------

II. Discussion of the Final Rule Amendments

    The amendments remove the existing Investment Grade Exceptions from 
both Rule 101 and Rule 102 of Regulation M. For distributions of 
Nonconvertible Securities, the Commission is adopting two new 
exceptions--one in 17 CFR 242.101(c)(2)(i) (``Rule 101(c)(2)(i)''), for 
reliance by distribution participants and their affiliated purchasers, 
and one in 17 CFR 242.102(d)(2)(i) (``Rule 102(d)(2)(i)''), for 
reliance by issuers, selling security holders, and their affiliated 
purchasers. Both exceptions are based on the requirements more fully 
described in Parts II.A.1 and B.1 that relate to the determination of 
an issuer's probability of default as derived from a structural credit 
risk model.\26\ As discussed below, in Part II.A.1, final Rule 
101(c)(c)(i) differs from the Proposal with regard to the exception's 
conditions involving who is eligible to make probability of default 
determinations pursuant to Rule 101(c)(2)(i) and when such probability 
of default determination must be made to rely on the exception. While 
the Proposal would have allowed any distribution participant to make 
the probability of default determination in meeting the conditions of 
Rule 101(c)(2)(i), the final amendments require the probability of 
default determination to be made by the distribution participant acting 
as the lead manager (or in a similar capacity) of a distribution.\27\ 
In addition, final Rule 101(c)(2)(i) will require five additional 
business days before the price determination date from what was 
proposed to make the probability of default determination in satisfying 
the exception's conditions.\28\ Finally, the Commission is making some 
technical, non-substantive changes from the Proposal with regard to the 
wording of the standard,\29\ as well as some clarifying changes to the 
proposed definition of ``structural credit risk model.'' \30\
---------------------------------------------------------------------------

    \26\ See 17 CFR 242.101(c)(2)(i), as amended, 242.102(d)(2)(i), 
as amended (requiring, for reliance by issuers, selling security 
holders, and their affiliated purchasers, that the distribution 
participant acting as the lead manager (or in a similar capacity) of 
a distribution have made the probability of default determination, 
as applicable to the subject security, pursuant to Rule 
101(c)(2)(i), as amended).
    \27\ See infra Part II.A.1.
    \28\ See infra Part II.A.1.
    \29\ See infra note 126.
    \30\ See infra Part II.A.1.
---------------------------------------------------------------------------

    Because the term ``structural credit risk model'' is used 
identically in both Rule 101(c)(2)(i) and Rule 102(d)(2)(i), as 
amended, the Commission is adding a definition for the term 
``structural credit risk model'' in Rule 100(b) of Regulation M.\31\ In 
addition, the Commission is adopting new paragraph (b)(17) of Rule 17a-
4 (``Rule 17a-4(b)(17)'') requiring the preservation of the written 
probability of default determination, relied upon by a broker-dealer, 
pursuant to new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i), as 
applicable, to facilitate Commission staff examinations.\32\
---------------------------------------------------------------------------

    \31\ See 17 CFR 242.100(b).
    \32\ See 17 CFR 240.17a-4(b)(17), as amended.
---------------------------------------------------------------------------

    For distributions of asset-backed securities, the Commission is 
adopting identical, new exceptions--one in 17 CFR 242.101(c)(2)(ii) 
(``Rule 101(c)(2)(ii)''), for reliance by distribution participants and 
their affiliated purchasers, and one in 17 CFR 242.102(d)(2)(ii) 
(``Rule 102(d)(2)(ii)''), for reliance by issuers, selling security 
holders, and their affiliated purchasers--requiring that such 
securities be offered pursuant to an effective shelf registration 
statement filed on Form SF-3.\33\
---------------------------------------------------------------------------

    \33\ See 17 CFR 242.101(c)(2)(ii), as amended, 
242.102(d)(2)(ii), as amended.
---------------------------------------------------------------------------

A. Rule 101(c)(2) of Regulation M: Implementing Section 939A(b) in 
Certain Exceptions for Distribution Participants

    The application of Rule 101's prohibitions to distributions of 
Nonconvertible Securities and asset-backed securities generally is 
limited because, under Regulation M, bids for and purchases of 
outstanding Nonconvertible Securities are not restricted unless the 
security being purchased is identical in all of its terms to the 
security being distributed.\34\ For example, Rule 101's restrictions do 
not apply for a security if there is a single basis point difference in 
coupon rates or a single day's difference in maturity dates from the 
security in distribution.\35\ In addition, as stated in the Proposal, 
commenters on the Commission's previously proposed alternatives to the 
Investment Grade Exception \36\ stated that reliance on the Investment 
Grade Exceptions largely is limited to two situations: re-openings 
(e.g., when an issuer may want to make a series of offerings of its 
fixed-income securities via a re-opening to match its funding needs or 
the desires of its target investor class, or when a foreign sovereign 
issuer may conduct a re-opening for public financing purposes) and 
sticky offerings.\37\ The securities that meet the

[[Page 39965]]

requirements of Rule 101's Investment Grade Exception are less likely 
to be subject to the type of manipulation that Rule 101 seeks to 
prevent because these securities trade on the basis of their yield and 
creditworthiness (traditionally measured by credit ratings), rather 
than the identity of the particular issuer, and are largely 
fungible.\38\
---------------------------------------------------------------------------

    \34\ Regulation M Adopting Release, 62 FR 524.
    \35\ See Regulation M Adopting Release, 62 FR 524. To illustrate 
with a simple example, absent an exception, a broker-dealer who is 
participating in a distribution of XYZ Corp.'s 3% bonds maturing 12/
31/2029 would be prohibited from making a market in bonds with those 
terms prior to completing the distribution. The broker-dealer would 
not, however, be prohibited from making a market in XYZ Corp.'s 3% 
bonds maturing 12/31/2030 because the date of maturity, a term of 
the bond, is different from that of the security in distribution.
    \36\ See supra note 15.
    \37\ Proposal, 61 FR 18316. In addition, the Commission also 
stated in the Proposal that another example provided by a commenter 
is a ``best-efforts'' offering. Proposal, 61 FR 18316. One commenter 
on the Proposal stated that firms rely on the Investment Grade 
Exceptions in the context of ``sticky deals'' and ``re-openings'' of 
debt issuances. See SIFMA Letter 1, at 4. As discussed below, in 
Part V.E.5, any offering can become a sticky offering. In such case, 
it may become challenging for the issue to trade based solely on its 
yield and maturity, notwithstanding the issuer's creditworthiness. 
Therefore, a sticky offering does not necessarily indicate a lack of 
creditworthiness on the part of the issuer. In the Proposal, the 
Commission asked if sticky offerings of creditworthy issuers 
disprove the underlying premise for excepting certain Nonconvertible 
Securities. See Proposal, 87 FR 18320. The Commission also asked if 
the Investment Grade Exception should be removed from Rule 101, 
without a replacement, because whether an offering will become 
sticky is unknown at the beginning of the Regulation M restricted 
period. See Proposal, 87 FR 18320. One commenter stated that it is 
unaware of any manipulative issues associated with reliance on the 
Investment Grade Exceptions in connection with sticky offerings. See 
SIFMA Letter 1, at 2. However, no commenter suggested that the 
Commission should remove the Investment Grade Exception from Rule 
101, without a replacement.
    \38\ See Regulation M Adopting Release, 62 FR 527; see also 
Regulation M Proposing Release, 61 FR 17112. For purposes of 
Regulation M, securities of issuers of a certain credit quality 
trade on the basis of their yield and creditworthiness 
(traditionally measured by credit ratings) and are less susceptible 
to manipulation because other similar Nonconvertible Securities or 
asset-backed securities are available to investors as an 
alternative. If the pricing of an offering is inconsistent with 
pricing in the overall secondary market for similar Nonconvertible 
Securities or asset-backed securities, an investor may purchase 
alternative securities that have a better yield, yet are of 
comparable creditworthiness, in relation to the security being 
distributed. Accordingly, the ability to substitute similar 
Nonconvertible Securities or asset-backed securities for the 
security in distribution limits the ability of a distribution 
participant to impact the market and distribution of such security.
---------------------------------------------------------------------------

1. Rule 101(c)(2)(i): Nonconvertible Securities of Issuers Who Meet a 
Specified Probability of Default Threshold
    The Commission proposed to except the Nonconvertible Securities of 
issuers for which the probability of default, estimated as of the day 
of the determination of the offering pricing and over the horizon of 12 
calendar months from such day, is less than 0.055%, as determined and 
documented in writing by the distribution participant as derived from a 
structural credit risk model.\39\ The Commission included a definition 
for the term ``structural credit risk model'' as a proviso in proposed 
Rule 101(c)(2)(i) to mean ``any commercially or publicly available 
model that calculates the probability that the value of the issuer may 
fall below a threshold based on an issuer's balance sheet.'' \40\ 
Accordingly, as proposed, a distribution participant's (or its 
affiliated purchaser's) reliance on proposed Rule 101(c)(2)(i) would 
have been conditioned on a probability of default determination that 
was made by use of any commercially or publicly available model that 
calculates the probability that the value of the issuer may fall below 
a threshold based on an issuer's balance sheet.
---------------------------------------------------------------------------

    \39\ Proposal, 87 FR 18338. The Commission stated that, as 
discussed in that release, based on an analysis of the probability 
of default and investment grade ratings of a sample of 
Nonconvertible Securities available on the market as of Oct. 22, 
2021, this was an appropriate substitute standard of 
creditworthiness in place of the reference to credit ratings in the 
Investment Grade Exception for Nonconvertible Securities. See 
Proposal, 87 FR 18318.
    \40\ Proposal, 87 FR 18338.
---------------------------------------------------------------------------

    As discussed in the Proposal, since 1974, structural credit risk 
models, such as the model first proposed by Robert C. Merton and its 
successor models, have become widely relied upon to determine the 
probability that an issuer will default on its loan obligations.\41\ 
Many commercial data providers, as part of software suites that allow 
users to analyze securities, employ certain structural credit risk 
models that are based on the Black-Scholes option pricing model as a 
way to measure the creditworthiness of companies. These types of 
structural credit risk models typically use measures from company 
accounting statements and company-specific and aggregate market prices 
and require input variables to calculate an estimated probability of 
default for a specified horizon, including the market value and 
volatility of the assets, as well as assumptions regarding the 
threshold for company asset values, below which the equity owner would 
default on its obligations (``Default Point'').\42\ In addition, these 
structural credit risk models provide the probability that a company's 
assets will fall below the Default Point at or by the expiration of a 
defined period.
---------------------------------------------------------------------------

    \41\ See generally Proposal, 87 FR 18316-17 (providing a history 
and overview of structural credit risk models). Generally, these 
models assume that owners of a company's equity will continue to pay 
the company's liabilities if the company's value exceeds its 
liabilities. Equivalently, if the equity owners were considered to 
own a call option on the value of the company with a strike price 
equivalent to the liabilities owed, the equity owners would exercise 
the call on the value of the company. If, however, the company's 
liabilities exceed the company's value, the models assume that the 
equity owners will choose to default on the company's liabilities, 
or equivalently, the equity owners would not exercise the call on 
the value of the company. Accordingly, these structural credit risk 
models provide a method, based on the Black-Scholes option pricing 
model, to estimate the probability that a company might default on 
its liabilities. See Proposal, 87 FR 18317.
    \42\ The Default Point frequently is calculated as all short-
term liabilities plus half of the long-term liabilities. See Mario 
Bondioli et al., The Bloomberg Corporate Default Risk Model (DRSK) 
for Public Firms (Mar. 2021), available at https://ssrn.com/abstract=3911300.
---------------------------------------------------------------------------

    Generally, the following variables are needed to derive an issuer's 
probability of default from a typical structural credit risk model: (1) 
the issuer's value, which can be based on observed market prices of an 
issuer's equity security or estimated based on an issuer's balance 
sheet; (2) the volatility of the issuer's equity or assets, which also 
can be based on market observations or estimated based on an issuer's 
balance sheet; (3) the risk-free rate; (4) a time horizon; and (5) the 
Default Point. A structural credit risk model's application may be 
limited in the absence of a market for an issuer's equity securities if 
the market price of the issuer's assets, which, as discussed above, is 
required to calculate the probability of default, is difficult to 
determine.\43\
---------------------------------------------------------------------------

    \43\ These models calculate the probability of default based on 
inputs from an issuer's balance sheet. Transactions in equity 
securities frequently are used as a proxy to determine the value of 
the firm and the overall volatility of the issuer's assets. Even the 
absence of a market for an issuer's equities alone does not preclude 
the ability of a distribution participant to use certain structural 
credit risk models because the issuer's balance sheet will include 
the liabilities, assets, and equity, which, with further analysis, 
can be used to determine the inputs for the models. Distribution 
participants, based on their activities as an underwriter, broker-
dealer, or other person who has agreed to participate in a 
distribution, can access an issuer's balance sheet to calculate the 
issuer's probability of default.
---------------------------------------------------------------------------

    Some commenters supported the proposed exception for Nonconvertible 
Securities that is based on an issuer's probability of default.\44\ One 
commenter stated that the proposed application of a structural credit 
risk model requirement will provide additional transparency for 
investors and other market participants.\45\ Another commenter agreed 
with the Commission that the estimated probability of default of a debt 
security ``is and should be a central component of the analysis of the 
credit risk'' \46\ and that the ``expected probability of default can 
be independently determined by structural credit risk models based on 
observable market events and information available on a firm's balance 
sheet,'' without having to rely on an investment grade rating.\47\ 
However, this commenter also stated that balance sheet measures 
frequently are inaccurate and that comparisons of market values to book 
values are subject to concerns about ``garbage in, garbage out.'' \48\ 
While the Commission requested comments regarding whether the exception 
proposed in Rule 101(c)(2)(i) should require the issuer's balance sheet 
to be audited,\49\ it did not receive any

[[Page 39966]]

comments in response to this question. In addition, the Commission 
considered including in the exception models that may not necessarily 
rely on an issuer's balance sheet to determine a firm's 
creditworthiness, such as reduced-form models. Reduced-form models, 
however, do not necessarily predict future defaults better than 
structural credit risk models do, and they suffer from a lack of 
theoretical foundation of the assumed relationships, or the intuitive 
interpretation of the model dependencies and why the defaults 
occur.\50\ For these reasons, and as discussed throughout this Part, 
the Commission is adopting a standard that is based on the use of a 
structural credit risk model as it is appropriately designed to measure 
creditworthiness of Nonconvertible Securities in new Rule 101(c)(2)(i), 
in accordance with the requirements of section 939A(b).
---------------------------------------------------------------------------

    \44\ Letter from Gregory Babyak, Global Head of Reg. Affairs, 
Bloomberg L.P., to Vanessa A. Countryman, Sec'y, SEC (May 23, 2022) 
(``Bloomberg L.P. Letter''), at 1; Letter from Robert E. Bishop, 
Fellow, Ctr. Law & Bus., UC Berkeley School of Law, & Frank Partnoy, 
Adrian A. Kragen Professor of Law, UC Berkeley School of Law, to 
Vanessa Countryman, Sec'y, SEC (May 23, 2022) (``IILF Letter''), at 
2; Better Markets Letter, at 2.
    \45\ Bloomberg L.P. Letter, at 1.
    \46\ IILF Letter, at 2.
    \47\ IILF Letter, at 5.
    \48\ IILF Letter, at 6.
    \49\ Proposal, 87 FR 18320. Such a requirement could present 
operational challenges in connection with deriving an issuer's 
probability of default from a structural credit risk model. As 
discussed below, in this Part, the determinations must be 
``estimated as of the sixth business day immediately preceding the 
determination of the offering price,'' which helps to ensure that 
timely information regarding the issuer is used as a model input. A 
lead manager may encounter difficulties in obtaining model input 
information from an issuer's audited balance sheet each time it 
needs to determine an issuer's probability of default for purposes 
of reliance on the exception for Nonconvertible Securities if such a 
determination were being made in between audits.
    \50\ See infra Part V.E.3.
---------------------------------------------------------------------------

    One commenter suggested an exception based on alternative standards 
of creditworthiness that do not utilize structural credit risk models. 
First, this commenter suggested that the Commission adopt an exception 
for Nonconvertible Securities that are offered pursuant to an effective 
registration statement filed on any of the following forms: (1) Form S-
3; \51\ (2) Form S-4; \52\ (3) Form F-3; \53\ (4) Form F-4; \54\ or (5) 
Form F-10,\55\ provided that, for an offering registered on Form F-10, 
the offering also meets the transactional requirements of General 
Instruction I.B.2 of Form F-3.\56\ The commenter stated several reasons 
for which the use of this standard would be desirable,\57\ and that 
allowing some amount of high yield issuers to be eligible for the 
exception would be an acceptable compromise in light of the benefits 
the commenter's proposed standard otherwise provides.\58\ In addition, 
this commenter stated that the ``consistently very high percentage of 
registered nonconvertible debt tranches that were investment grade 
demonstrates that Securities Act registration alone serves as a 
reliable proxy for identifying offerings of nonconvertible debt 
securities that trade primarily based upon their yield and 
creditworthiness.'' \59\
---------------------------------------------------------------------------

    \51\ 17 CFR 239.13.
    \52\ 17 CFR 239.25.
    \53\ 17 CFR 239.33.
    \54\ 17 CFR 239.34.
    \55\ 17 CFR 239.40.
    \56\ SIFMA Letter 1, at 5-8; see Letter from Joseph Corcoran, 
Managing Dir. & Assoc. Gen. Counsel, Secs. Indus. & Fin. Mkts. 
Ass'n, to Vanessa Countryman, Sec'y, SEC (Oct. 28, 2022) (``SIFMA 
Letter 2'').
    \57\ SIFMA Letter 1, at 3-4 (stating that this standard would 
provide a straightforward, uniform standard; align with how the 
Commission addressed the Dodd-Frank Act-related removal of 
references to credit ratings from the eligibility criteria for use 
of certain provisions under the Securities Act and related forms; 
promote the conduct of offerings on a registered basis by limiting 
the exception to qualifying registered offerings; afford 
predictability; avoid complex calculations that could lead to errors 
or differing results, depending on the particular structural credit 
risk model used; allow the availability of the exception to be 
readily and independently verified through a review of the issuer's 
Electronic Data Gathering, Analysis, and Retrieval (EDGAR) database 
filings, which would minimize additional regulatory burdens and 
obviate the need for any additional broker-dealer recordkeeping 
obligations; ease the burden on all involved, including for 
regulators; and provide greater legal certainty to the affected 
issuers and any selling shareholders).
    \58\ SIFMA Letter 2, at 1; SIFMA Letter 1, at 5-7.
    \59\ SIFMA Letter 2, at 3.
---------------------------------------------------------------------------

    The Commission acknowledges that the commenter's suggested standard 
may capture a consistently very high percentage of registered 
nonconvertible debt tranches that were rated investment grade and may 
seem operationally easier to determine whether the new exception in 
Rule 101(c)(2)(i) for Nonconvertible Securities is available, may help 
to promote the conduct of offerings on a registered basis, and allow 
for the use of an exception that can be verified with publicly 
available information, among other things. The commenter's suggested 
standard, however, would not be appropriate because it does not 
sufficiently focus on creditworthiness. When the Commission revised the 
eligibility criteria for use of Forms S-3 and F-3 to remove any 
references to credit ratings, it specifically stated that the 
eligibility criteria included in those forms did not distinguish among 
issuers by the quality of their credit but rather focused exclusively 
on whether the issuer has a wide following in the marketplace to 
identify issuers who should be eligible for short-form registration and 
faster access to capital markets through the shelf registration 
process.\60\ This is distinct from the basis for adopting the 
Investment Grade Exceptions, including under Rule 101(c)(2), which was 
that Nonconvertible Securities are appropriate to except from 
Regulation M's requirements because they are fungible and traded on the 
basis of their yield and creditworthiness, and therefore are less 
likely to be manipulated.\61\
---------------------------------------------------------------------------

    \60\ Security Ratings, Release No. 33-9245 (July 27, 2011) [76 
FR 46603 (Aug. 3, 2011)] (``Form S-3 and Form F-3 Release''), 76 FR 
46607. When the Commission revised the eligibility criteria for use 
of Form S-3 and Form F-3 to remove any references to credit ratings, 
it noted that none of the criteria are a standard of 
creditworthiness. Form S-3 and Form F-3 Release, 76 FR 46607 n.60. 
The Commission stated that ``any alternative standard for Forms S-3 
and F-3 eligibility that does not refer to credit ratings should 
preserve the forms and access to the shelf registration process for 
issuers who have a wide following in the marketplace.'' Form S-3 and 
Form F-3 Release, 76 FR 46607. Form SF-3, the shelf registration 
statement form for asset-backed securities that is discussed below, 
in Part II.A.2, differs from these other forms raised by the 
commenter that rely on the eligibility criteria for use of Form S-3 
or Form F-3. Whereas the eligibility criteria included in Forms S-3 
and F-3 focus exclusively on whether the issuer has a wide following 
in the marketplace to identify issuers who should be eligible for 
short-form registration and faster access to capital markets through 
the shelf registration process, the eligibility criteria and 
offering requirements included in Form SF-3 help to ensure that 
asset-backed securities issued in shelf offerings are designed to 
help ensure that that the securitization is designed to produce 
expected cash flows that are sufficient to service payments or 
distributions in accordance with their terms; that obligated parties 
more carefully consider the characteristics and quality of the 
assets that are included in the pool; and that asset-backed 
securities shelf offerings have transactional safeguards and 
features that make those certain securities appropriate to be issued 
without prior Commission staff review. See Asset-Backed Securities 
Disclosure and Registration, Release No. 34-72982 (Sept. 4, 2014) 
[79 FR 57184 (Sept. 24, 2014)] (``Regulation AB II Adopting 
Release''), 79 FR 57267, 57278, 57283.
    \61\ See Regulation M Adopting Release, 62 FR 524; supra note 
13. For these reasons discussed above, as related to the eligibility 
criteria of Forms S-3 and F-3, the commenter's suggestion of 
excepting Nonconvertible Securities that are offered pursuant to an 
effective registration statement filed on Form S-4 or F-4 would not 
be an appropriate substitute standard of creditworthiness in place 
of the reference to credit ratings in the Investment Grade Exception 
pursuant to section 939A(b) because Forms S-4 and F-4 include the 
Forms S-3 and F-3 eligibility criteria by allowing registrants that 
meet the registrant eligibility requirements of Form S-3 or F-3 and 
that are offering investment grade securities to incorporate by 
reference certain information. See Form S-3 and Form F-3 Release, 76 
FR 46611 (citing General Instruction B.1 of Forms S-4 and F-4). 
Similarly, the commenter's suggestion of excepting Nonconvertible 
Securities that are offered pursuant to an effective registration 
statement filed on Form F-10, provided that the offering also meets 
the transactional requirements of General Instruction I.B.2. of Form 
F-3, would not be an appropriate substitute standard of 
creditworthiness in place of the reference to credit ratings in the 
Investment Grade Exception pursuant to section 939A(b) because that 
measure references transactional requirements that have a distinct 
purpose from the Commission's original basis for adopting the 
Investment Grade Exception. As discussed in this Part, the 
probability of default is an appropriate measure to identify low 
manipulation risk of Nonconvertible Securities because it allows for 
the selection of issuers whose securities trade on the basis of 
yield and creditworthiness.
---------------------------------------------------------------------------

    Specifically, securities of issuers of a certain credit quality 
trade based on

[[Page 39967]]

yield and creditworthiness \62\ and are less susceptible to 
manipulation because other similar Nonconvertible Securities are 
available to investors as an alternative to the security in 
distribution. If pricing of a Nonconvertible Security offering is 
inconsistent with pricing in the overall secondary market for similar 
Nonconvertible Securities, an investor may purchase alternative 
Nonconvertible Securities that have a better yield, yet are of 
comparable creditworthiness, than the security being distributed. 
Accordingly, the ability to substitute similar Nonconvertible 
Securities in the market for the security in distribution limits the 
potential impact that a distribution participant might attempt to exert 
on the market and distribution of such security. In addition, when debt 
has a very low probability of default, the cash flows are close to 
risk-free. Thus, the price of the debt is mainly subject to 
fluctuations based on aggregate interest rates rather than issuer-
specific or security-specific news.
---------------------------------------------------------------------------

    \62\ Bonds trade among investors and dealers in secondary 
markets at prices that depend on economy-wide interest rates, as 
well as on market perceptions regarding the likelihood that the 
issuing company will make the promised payments. Hendrik 
Bessembinder & William Maxwell, Markets: Transparency and the 
Corporate Bond Market, 22 J. ECON. PERSP. 217, 220 (2008).
---------------------------------------------------------------------------

    The probability of default is an appropriate measure to identify 
low manipulation risk of such securities, as it allows for the 
selection of issuers whose securities trade on the basis of yield and 
creditworthiness (traditionally measured by credit ratings). For 
issuers with sound creditworthiness, the pricing of securities is 
unrelated to other risks associated with the identity of the issuer, 
greatly reducing their uncertainty and manipulation risk. A standard 
based on a criterion such as being widely followed in the market does 
not allow for such a clear distinction because such a standard does not 
differentiate securities that are traded solely on their yield and 
creditworthiness from securities that trade solely on the issuer's 
identity and thus could present a high manipulation risk.\63\ 
Accordingly, it is appropriate to implement the section 939A(b) mandate 
by adopting an exception that is based on a standard that is likewise 
premised specifically on creditworthiness rather than on whether a 
particular issuer has a wide following in the marketplace.
---------------------------------------------------------------------------

    \63\ See infra Part V.E.3.
---------------------------------------------------------------------------

    Second, the commenter suggested that its recommended standard 
described above, which is based on the Forms S-3 and F-3 eligibility 
criteria, could be modified by prohibiting reliance on the exception 
for Nonconvertible Securities that include both a ``limitation on 
restricted payments covenant'' and a ``limitation on sales of assets 
and subsidiary stock covenant,'' which the commenter stated are two 
covenants that typically are associated with non-investment grade debt 
securities and are almost never used in investment grade debt 
securities.\64\
---------------------------------------------------------------------------

    \64\ SIFMA Letter 1, at 8.
---------------------------------------------------------------------------

    However, despite the commenter's suggestions, the covenant 
restrictions are features of current market practices \65\ but are not 
necessarily inherent characteristics of the securities related to their 
creditworthiness. Conditioning the exception on the absence of certain 
covenants poses the risk that, should market practice change, the 
exception would quickly become outdated. Therefore, even with the 
commenter's two suggested modifications, a standard that is focused on 
the Form S-3 or Form F-3 eligibility criteria and is premised 
exclusively on whether an issuer is widely followed,\66\ rather than on 
an issuer's creditworthiness, is not an appropriate substitute standard 
of creditworthiness in place of the references to credit ratings in 
Rule 101's Investment Grade Exception for Nonconvertible Securities to 
sufficiently respond to the requirements of section 939A(b).
---------------------------------------------------------------------------

    \65\ See, e.g., SIFMA Letter 1, at 8.
    \66\ Form S-3 and Form F-3 Release, 76 FR 46607.
---------------------------------------------------------------------------

    Finally, this commenter suggested that the Commission adopt a 
modified version of the 2008 Proposing Release involving a well-known 
seasoned issuer (``WKSI'') standard \67\ to except: (1) a WKSI that, as 
of a date within 60 days of the applicable determination date, has a 
worldwide market value of its outstanding voting and non-voting common 
equity held by non-affiliates of $700 million or more; and (2) a non-
WKSI to the extent it is carved out of the WKSI definition solely by 
virtue of the application of paragraph (1)(v), (vi), or (ix) of the 
definition of ``ineligible issuer'' under Rule 405 under the Securities 
Act.\68\ The Commission recognizes the advantage a WKSI-based standard 
might have in terms of its simplicity and straightforward calculation. 
As noted in the Proposal, however, a WKSI-based standard as proposed in 
2008 was criticized for allowing many risky, high-yield issues to be 
excepted and preventing issues by smaller but otherwise credit-worthy 
issuers from being eligible for the exception.\69\ This WKSI-based 
standard, however, unlike the probability-of-default-based standard, 
would fail to capture the pricing point where the sound 
creditworthiness of the issuer eliminates other risks associated with 
the issuer identity. The Nonconvertible Securities of such issuers 
trade solely based on their yields and creditworthiness and not on 
issuer characteristics, where pricing uncertainty and manipulation risk 
are at their minimum.\70\ The WKSI-based standard, therefore, is a less 
effective measure of manipulation risk as compared to the probability 
of default measure.
---------------------------------------------------------------------------

    \67\ In 2008, prior to the enactment of the Dodd-Frank Act, the 
Commission proposed to substitute credit ratings references in Rules 
101 and 102 with a standard for Nonconvertible Securities that was 
based primarily on the WKSI concept from 17 CFR 230.405 (``Rule 
405''), as well as a standard for asset-backed securities that were 
registered on Form S-3. See 2008 Proposing Release, 73 FR 40095-97. 
The WKSI-based approach, consistent with the definition of WKSI 
under Rule 405, would have excepted the Nonconvertible Securities of 
companies that have issued at least $1 billion aggregate principal 
amount of nonconvertible securities, other than common equity, in 
primary offerings for cash, not exchange, registered under the 
Securities Act. See 17 CFR 230.405, paragraph (1)(i)(B)(1) of the 
definition of WKSI; see also 2008 Proposing Release, 73 FR 40096.
    \68\ See SIFMA Letter 1, at 10. This commenter stated that its 
suggested exception based on Forms S-3 and F-3 is more appropriate 
than this WKSI-based approach because the Form S-3/F-3-based 
approach ``recognizes several different means of qualifying under 
the transactional requirement, only one of which is based upon the 
aggregate principal amount of non-convertible securities issued over 
the preceding three years.'' SIFMA Letter 1, at 9-10. For the 
reasons discussed in this Part, neither of these approaches is an 
appropriate standard of creditworthiness in place of the reference 
to credit ratings in the Investment Grade Exception.
    \69\ Proposal, 87 FR 18334-35.
    \70\ See infra Part V.E.3.
---------------------------------------------------------------------------

    For these reasons, and pursuant to the requirements of section 
939A(b), the probability of default measure is a more appropriate 
substitute of creditworthiness for the reference to credit ratings in 
the existing Investment Grade Exception than is the commenter's 
suggested WKSI-based standard.\71\ As discussed above, in this Part, 
when debt has a very low probability of default, its price fluctuations 
are mainly based on aggregate interest rates rather than on company-
specific or security-specific news. The probability of default measure, 
in contrast to the commenter's suggested WKSI-based standard, continues 
to rely on the premise underlying the Investment Grade Exception: 
Nonconvertible Securities

[[Page 39968]]

that trade primarily based on their yield and creditworthiness are less 
susceptible to the type of manipulation that Rule 101 seeks to prevent.
---------------------------------------------------------------------------

    \71\ See Proposal, 87 FR 18317. Similar to how securities 
covered by the existing Investment Grade Exception are excepted from 
Rule 101's prohibitions, Nonconvertible Securities that trade based 
on their yield and creditworthiness would be excepted under Rule 101 
as amended to include the probability of default-based standard.
---------------------------------------------------------------------------

    Some commenters stated that the Commission should specify a 
particular structural credit risk model to be used by all parties in 
making probability of default calculations.\72\ These commenters stated 
their concerns regarding the potential for inconsistent outcomes 
resulting from the discretion to choose what structural credit risk 
models to apply.\73\ One commenter stated that the adoption of the 
proposed model-based standard would create the risk that the new 
standard would be manipulated because firms would have a wide variety 
of models from which to select.\74\ This commenter stated that, while 
the proposed probability-of-default-based standard is a reasonable 
alternative standard of creditworthiness, the use of structural credit 
risk models would create challenges for the Commission, with regard to 
implementing the probability of default standard, and for investors, 
with regard to confidence in the consistency and reliability of 
determinations made under the new standard.\75\ In addition, this 
commenter stated that setting no minimum standards for the models and 
allowing market participants the discretion to choose among a wide 
range of models threatens to create a ``race to the bottom'' as market 
participants seek to avoid competitive disadvantages that will arise 
from having an appropriately rigorous risk of default evaluation.\76\ 
Another commenter, however, stated that it would be ``risky'' for the 
Commission to engage in ``model preferencing.'' \77\
---------------------------------------------------------------------------

    \72\ See SIFMA Letter 1, at 10; Better Markets Letter, at 5.
    \73\ See SIFMA Letter 1, at 5; Better Markets Letter, at 4.
    \74\ Better Markets Letter, at 4 (stating, in part, that the use 
of structural credit risk models will create a lack of uniformity 
that conflicts with the mandate in section 939A(b) for the 
Commission to establish, to the extent feasible, uniform standards 
of creditworthiness). But see Bloomberg L.P. Letter, at 2 (stating 
that, although the application of a particular threshold across 
multiple models may have some unintended consequences (e.g., 
different point-in-time probability of default models may produce 
different results for the same issuance), the proposed exception 
provides an alternative measure of creditworthiness that is 
practical, appropriately based on objective factors, and can be 
consistently applied by market participants). The mandate in section 
939A(b) to seek to establish uniform standards of creditworthiness 
is limited ``to the extent [that it is] feasible.'' Public Law 111-
203, sec. 939A(b). As discussed in this Part, the use of structural 
credit risk models to derive an issuer's probability of default is 
an appropriate standard of creditworthiness in accordance with 
section 939A(b)'s requirements.
    \75\ Better Markets Letter, at 4.
    \76\ Better Markets Letter, at 4.
    \77\ See IILF Letter, at 6 (stating that requiring a particular 
type of model could potentially distort the behavior of market 
participants in their estimations of probability of default and 
discouraging further and alternative inquiries into the probability 
of default). The use of structural credit risk models is required 
only for purposes of deriving an issuer's probability of default 
pursuant to new Rule 101(c)(2)(i). Distribution participants, as 
well as other market participants, may use other types of models in 
evaluating the creditworthiness of an issuer outside of making a 
Rule 101(c)(2)(i) probability of default determination.
---------------------------------------------------------------------------

    The Commission agrees with the comment against requiring the use of 
a specific structural credit risk model.\78\ On balance, the use of a 
structural credit risk model to derive an issuer's probability of 
default pursuant to Rule 101(c)(2)(i), as amended, provides an 
appropriate degree of flexibility in terms of model selection while 
also providing certainty to distribution participants as to the 
standards for the structural credit risk model required for purposes of 
compliance in making probability of default determinations. The ability 
to use an unrestricted universe of models for purposes of meeting the 
conditions of new Rule 101(c)(2)(i) could provide distribution 
participants with the opportunity to choose model specifications that 
enable abuse of the exception for Nonconvertible Securities. In this 
regard, the definition of ``structural credit risk model,'' as 
discussed below in this Part, sets minimum standards for the structural 
credit risk models that may be used to derive an issuer's probability 
of default to meet the conditions of new Rule 101(c)(2)(i). These 
minimum standards include that the model be a commercially or publicly 
available model and that it calculate, based on an issuer's balance 
sheet, the probability that the value of the issuer will fall below the 
Default Point, at or by the expiration of a defined period. As 
discussed below, in Part V.C.3, the standard's use of structural credit 
risk models could incentivize lead managers to select models and 
estimation specifics in such a way to ensure the resulted estimates are 
below the threshold, thus allowing securities of issuers with low 
creditworthiness and high manipulation risk to be eligible for the 
exception. The public availability of alternative estimates for 
investors, however, should mitigate this concern.\79\ Specifically, the 
limitation that the structural credit risk model must be a commercially 
or publicly available model would limit a distribution participant's 
ability to develop models for the purpose of abusing the exception.\80\ 
In this regard, use of a structural credit risk model that is not 
commercially or publicly available, or one that does not calculate, 
based on an issuer's balance sheet, the probability that the value of 
the issuer will fall below the Default Point, at or by the expiration 
of a defined period, would not be permissible in meeting the conditions 
of the exception.
---------------------------------------------------------------------------

    \78\ See IILF Letter, at 6.
    \79\ Also, as discussed below, in Part II.C, because probability 
of default estimates may be subjective to some extent and not 
comparable across different issuers or for the same issuer across 
different issues if estimates are based on different models, or done 
by different researchers or vendors, the requirement associated with 
reliance on new Rule 101(c)(2)(i) to preserve written probability of 
default determinations is designed to facilitate the Commission's 
examinations of broker-dealers who rely on the exception in new Rule 
101(c)(2)(i) or new Rule 102(d)(2)(i).
    \80\ See infra Part V.C.3.
---------------------------------------------------------------------------

    While a standard that relies on the use of a structural credit risk 
model retains a certain level of subjectivity,\81\ this standard also 
leaves room for improvement if the market adopts more accurate 
structural credit risk models in the future. As the Commission stated 
in the Proposal, the use of any model to estimate creditworthiness 
necessarily provides an imperfect measure.\82\ This flexibility in 
selection may result in an outcome-oriented selection of structural 
credit risk models, as one commenter suggested.\83\ However, a 
selection that meets the definition of ``structural credit risk 
model,'' as provided in Rule 100(b), as well as the requirements of new 
Rule 101(c)(2)(i), would be consistent with the aims of section 939A as 
well as those of Regulation M.\84\ In addition, the new record 
preservation requirement set forth in Rule 17a-4(b)(17), as discussed 
below in Part II.C, is designed to aid Commission examinations of 
broker-dealers who rely on the exception in Rule 101(c)(2)(i) or Rule 
102(d)(2)(i), as amended, and can help deter improper adjusting of the 
estimation to meet the conditions of either of the exceptions.\85\
---------------------------------------------------------------------------

    \81\ See infra Part V.C.3; see also Proposal, 87 FR 18334.
    \82\ Proposal, 87 FR 18332.
    \83\ SIFMA Letter 1, at 5.
    \84\ See infra Part V.B (discussing how structural credit risk 
models, as defined in Rule 100(b), are designed to measure 
creditworthiness, and creditworthiness itself is considered to be a 
good measure of manipulation risk).
    \85\ See 17 CFR 240.17a-4(b)(17), as amended; infra Part II.C.
---------------------------------------------------------------------------

    The Commission requested comment on whether there are ``any reasons 
why the Rule should not permit a distribution participant to perform 
its own calculation (subject to recordkeeping requirements, as 
proposed).'' \86\ To address concerns that the proposed flexibility in 
structural

[[Page 39969]]

credit risk model selection could lead to different underwriters coming 
to different conclusions on the availability of an exception from 
Regulation M based on which structural credit risk model they use, as 
well that this flexibility could contribute to inefficiencies, 
confusion, and dissension among distribution participants,\87\ the 
final amendments limit the universe of those who are eligible to 
determine an issuer's probability of default under new Rule 
101(c)(2)(i) to include only the distribution participant who is acting 
as the lead manager (or in a similar capacity) \88\ of a 
distribution.\89\ This limitation will help to ensure consistent 
reliance on the exception across all distribution participants for the 
same distribution through use of the same written probability of 
default determinations and through the new record preservation 
requirements under Rule 17a-4(b)(17).\90\ The lead manager's role and 
responsibilities in overseeing the distribution process \91\ should, 
for these same reasons, help alleviate concerns regarding the 
consistency and reliability of the determinations within any particular 
distribution.\92\
---------------------------------------------------------------------------

    \86\ Proposal, 87 FR 18320.
    \87\ SIFMA Letter 1, at 5.
    \88\ Distribution participants who act as the ``lead manager'' 
of a distribution for purposes of the exception in Rule 101 may, as 
a practical matter, also use or be known by different titles, such 
as ``lead underwriter,'' ``managing lead underwriter,'' ``syndicate 
manager,'' ``stabilizing manager,'' ``lead bookrunner,'' or ``co-
managing underwriter.'' The parenthetical ``(or in a similar 
capacity)'' is included in FINRA's underwriting-related rules, such 
as FINRA Rule 5110, to recognize this common industry practice, as 
well as to prevent evasion by persons attempting to avoid regulatory 
responsibility under a particular provision by using a different 
title or term to refer to themselves, even though they perform the 
same or similar function.
    \89\ See 17 CFR 242.101(c)(2)(i), as amended.
    \90\ As discussed below, in Part II.C, broker-dealers who rely 
on the new exception for Nonconvertible Securities in new Rule 
101(c)(2)(i) or new Rule 102(d)(2)(i), as applicable, must preserve 
certain records pursuant to Rule 17a-4 under the Exchange Act. New 
paragraph (b)(17) of Rule 17a-4 requires broker-dealers to preserve 
the written probability of default determination, relied upon 
pursuant to the exception for Nonconvertible Securities. 
Accordingly, broker-dealers relying on the exception for 
Nonconvertible Securities are required to preserve for a period of 
not less than three years, the first two years in an easily 
accessible place, the written probability of default determination.
    \91\ The term ``lead manager'' under new Rule 101(c)(2)(i) is 
consistent with how the term ``manager'' is applied, for 
recordkeeping purposes, in 17 CFR 240.17a-2(b)(1) with respect to 
any person who acts as a manager of a distribution for its sole 
account or for the account of a syndicate or group in which it is a 
participant with respect to keeping records of any syndicate 
covering transactions, penalty bids, and all related stabilizing 
activity, all three of which are governed under 17 CFR 242.104, 
which cross-references the recordkeeping requirement in 17 CFR 
240.17a-2, as well as the ``managing underwriter'' in connection 
with TRACE-reporting of eligible fixed-income securities, or FINRA 
Rule 5131's requirement that the lead managing underwriter of a 
distribution disclose indications of interest and final allocation 
information to the issuer's pricing committee, or notify the issuer 
of any impending release or waiver of lock-ups. Thus, similar to the 
traditional role played by the lead or managing underwriter in firm 
commitment offerings--which generally include overseeing the 
offering process to ensure that the marketing, pricing, and 
allocation processes all go smoothly; providing critical advice on 
the structure, size, timing, and price of the offering; and advising 
on how to best present the issuer's business in the prospectus or 
other offering documents--the distribution participant acting as the 
lead manager (or in a similar capacity) of a distribution is the 
only market participant who is eligible to derive the issuer's 
probability of default for purposes of meeting the conditions of new 
Rule 101(c)(2)(i), in recognition that it is in the best position to 
do so. There may be distributions with more than one distribution 
participant acting as the lead manager (or in a similar capacity). 
In such a distribution, because the rule text refers to ``the 
distribution participant acting as the lead manager,'' only one of 
the distribution participants acting as the lead manager would be 
permitted to make the probability of default determination for the 
particular distribution.
    \92\ See, e.g., infra Part V.C.3 (discussing that the 
requirement related to the lead manager's probability of default 
determination should mitigate the subjectivity (of the analysis 
involved in probability of default estimation, as well as of the 
selection of the model and data sample specifics) and the concerns 
regarding non-uniform probability of default estimates for the same 
issue--and to some degree across issues for the same issuer to the 
extent the same parties are engaged by the issuer for different 
issues).
---------------------------------------------------------------------------

    The lead-manager requirement is intended to broadly reflect current 
market practices.\93\ Lead managers will be incentivized to share their 
probability of default determinations with other distribution 
participants and their affiliated purchasers (as well as with the 
issuer, selling security holders, and their affiliated purchasers) in 
order to rely on the exception for Nonconvertible Securities given 
their primary role and responsibilities in overseeing the distribution 
process, which can include providing liquidity and facilitating an 
orderly distribution and aftermarket in connection with the 
offering.\94\ While Regulation M does not require the lead manager to 
coordinate the activities of the other syndicate members, lead managers 
are, as a practical matter, concerned that the other underwriters in 
the syndicate are, among other things, complying with Regulation M's 
trading prohibitions \95\ so as not to extend the Regulation M 
restricted period.\96\ However, Rule 101(c)(2)(i), as amended, does not 
require that the lead manager making the probability of default 
determination share the determination with other distribution 
participants or their affiliated purchasers in order for those parties 
to rely on the exception.\97\ Therefore, non-lead manager distribution 
participants and their affiliated purchasers (as well as issuers, 
selling security holders, and their affiliated purchasers, as discussed 
below, in Part II.B.1) may not be able to rely on the exception for 
Nonconvertible Securities if the lead manager does not share the 
probability of default determination or there is no distribution 
participant to act as the lead manager for the distribution, such as 
with self-underwritten offerings, at-the-market offerings, or other 
shelf offerings, to the extent such an offering meets the definition of 
a ``distribution'' under Rule 100(b) of Regulation M. \98\
---------------------------------------------------------------------------

    \93\ See, e.g., supra notes 88, 91.
    \94\ See, e.g., Regulation M Adopting Release, 62 FR 534-35.
    \95\ See, e.g., supra notes 88, 91; SIFMA, Model Form of Master 
Agreement Among Underwriters (Dec. 10, 2018), https://www.sifma.org/wp-content/uploads/2017/08/SIFMA-Model-MAAU.pdf https://www.sifma.org/wp-content/uploads/2017/08/SIFMA-Model-MAAU.pdf.
    \96\ See Regulation M Adopting Release, 62 FR 522-23 (discussing 
Rule 100's definition of ``completion of participation in a 
distribution'' when underwriters in a syndicate are involved).
    \97\ The Commission asked in the Proposal whether distribution 
participants should be required to post or make the probability of 
default public on their website to rely on the exception. See 
Proposal, 87 FR 18320. As discussed below, in Part II.C, one 
commenter stated that the Commission also could publish, or require 
publication of, default probability estimates that market 
participants derive from various models, along with default 
probabilities implied by both market prices and credit default swap 
spreads. See IILF Letter, at 8. However, the Commission did not 
receive any comment suggesting that distribution participants making 
probability of default determinations should be required to post or 
make the probability of default determinations public on their 
website in order to rely on the exception. Nor did the Commission 
receive any comment suggesting that the sharing of probability of 
default determinations among other covered persons should be 
included as a condition to the exception.
    \98\ Rule 100 of Regulation M defines the term ``distribution'' 
as ``an offering of securities, whether or not subject to 
registration under the Securities Act that is distinguished from 
ordinary trading transactions by the magnitude of the offering and 
the presence of special selling efforts and selling methods.'' With 
regard to shelf offerings, each takedown of a shelf is to be 
individually examined to determine whether such offering constitutes 
a ``distribution'' (i.e., whether it satisfies the ``magnitude'' of 
the offering and ``special selling efforts and selling methods'' 
criteria of a distribution). Regulation M Adopting Release, 62 FR 
526. In those situations where a broker-dealer sells shares on 
behalf of an issuer or selling security holder in ordinary trading 
transactions into an independent market (i.e., without any special 
selling efforts), the offering will not be considered a 
distribution, and the broker-dealer will not be subject to Rule 101. 
Regulation M Adopting Release, 62 FR 526.
---------------------------------------------------------------------------

    This potential impact is mitigated, however, if the syndicate 
adjusts the way it interacts with lead managers. For example, the 
syndicate could decide, by contract (e.g., in an agreement among 
underwriters) and for the same reason of consistency as discussed 
above, to specifically authorize the lead manager

[[Page 39970]]

to share its probability of default determination with other 
distribution participants. The syndicate could also make the decision 
not to allow the lead manager to share its probability of default 
determination with unrelated distribution participants in order to keep 
a tighter control of the distribution. Because the facts and 
circumstances vary across issues, it is reasonable to let the syndicate 
decide how widely the probability of distribution determination is 
shared by the lead manager. If such information is shared, it must be 
used consistently across the syndicate to rely on the new exception in 
Rule 101(c)(2)(i) because such reliance is conditioned on the lead 
manager's probability of default determination.
    The lead-manager requirement, however, could affect current market 
practices by resulting in fewer Nonconvertible Securities being 
excepted under the new standard in comparison to those currently 
excepted under the Investment Grade Exception if, as a practical 
matter, only Nonconvertible Securities that are subject to underwritten 
offerings become eligible for the new exception in Rule 101(c)(2)(i). 
With regard to the types of distributions covered, as a result of the 
condition requiring that the lead manager perform the probability of 
default determinations in order for reliance on the exception, this 
exception will be available to a subset of distributions of 
Nonconvertible Securities covered under the existing Investment Grade 
Exception and to a subset of the distributions that would have been 
captured under the Proposal.\99\ Accordingly, any potential challenges, 
as a commenter suggested,\100\ are likely to be faced in carrying out 
obligations in order to rely on the new exception in Rule 101(c)(2)(i). 
The estimated costs associated with the requirement related to the lead 
manager making the probability of default determination are included 
below, in Part V.C.1.\101\
---------------------------------------------------------------------------

    \99\ See Proposal, 87 FR 18330.
    \100\ See Better Markets Letter, at 4.
    \101\ However, as discussed below, in Parts V.C.1 and VI.C.1, 
some lead managers may rely on third party vendors rather than 
internally calculate the probability of default.
---------------------------------------------------------------------------

    The bright-line requirements of Rule 101(c)(2)(i), as amended, and 
the corresponding record preservation requirements of new Rule 17a-
4(b)(17) also help to promote confidence in the consistency and 
reliability of the lead manager's determinations by limiting the degree 
to which there are variances between probability of default 
calculations within any one distribution as well as by deterring any 
improper tweaking of model inputs.\102\ The bright-line threshold of 
0.055%, as well as the pre-determined time horizon, are model inputs 
that are uniform and predictable and, thus, should provide the 
necessary clarity as to what is expected in evaluation and 
documentation. In addition, the bright-line threshold of 0.055% will 
help to ensure that only those Nonconvertible Securities that trade on 
the basis of yield and creditworthiness, and are fungible,\103\ will 
meet the exception, regardless of the model picked. Accordingly, the 
probability-of-default-based standard articulates an appropriate 
alternative measure of creditworthiness that is practical and is 
appropriately based on objective factors.
---------------------------------------------------------------------------

    \102\ As discussed below, in Part II.C, the record preservation 
requirements provided in new Rule 17a-4(b)(17) are sufficient to 
help the Commission's examinations of broker-dealers relying on the 
new probability-of-default-based standard.
    \103\ See, e.g., supra notes 13, 38.
---------------------------------------------------------------------------

    One commenter stated that commercially or publicly available 
structural credit risk models are not used by all firms in the context 
of evaluating whether to underwrite a security and that such a 
decision, instead, is focused on the adequacy and accuracy of 
disclosures.\104\ The Commission acknowledges that, when a firm 
evaluates whether to underwrite a distribution of securities, it 
typically focuses on the adequacy and accuracy of an issuer's 
disclosures. However, the adequacy and accuracy of disclosures are a 
separate question from the creditworthiness of securities for purposes 
of the exception for Nonconvertible Securities. It is possible that an 
accurate disclosure statement did not reveal the issuer's low 
creditworthiness, making the offered securities inappropriate to be 
eligible for the exception for Nonconvertible Securities that trade on 
the basis of their yield and creditworthiness.
---------------------------------------------------------------------------

    \104\ SIFMA Letter 1, at 5.
---------------------------------------------------------------------------

    One commenter stated that the proposed probability of default 
calculations would create a heightened risk for errors.\105\ Another 
commenter stated that market participants can reliably estimate the 
probability of default, not only by using the proposed ``structural 
credit risk models'' but also by using other statistical models,\106\ 
market measures of credit risk, and other credit risk measures.\107\ 
This commenter encouraged the Commission to adopt a final rule that 
references market measures of credit risk as part of the estimation of, 
or as an alternative to, the probability of default.\108\
---------------------------------------------------------------------------

    \105\ SIFMA Letter 1, at 5.
    \106\ IILF Letter, at 2 (stating that the inability to reliably 
estimate the probability of default on a debt security using any of 
a variety of statistical models and market measures is strong 
evidence that the security should not fall within an exception to 
Regulation M and could also be evidence that the market participant 
is not in a position to trade or hold that specific security). As 
discussed in this Part, new Rule 101(c)(2)(i) requires that the 
probability of default determination be derived from a structural 
credit risk model. Distribution participants and their affiliated 
purchasers may not avail themselves of the exception in new Rule 
101(c)(2)(i) with regard to any security that does not meet the 
requirements of that exception.
    \107\ IILF Letter, at 2.
    \108\ IILF Letter, at 2, 6-8 (stating its concerns about the use 
of balance sheets and suggested that the Commission reference more 
flexible alternatives, such as the probability of default threshold 
could vary annually on an ongoing basis depending on a similar 
analysis of more recent data going forward or peg the annual 
probability of default threshold based on an analysis of a sample of 
securities from the previous year). The Commission has considered 
this comment and, on balance, concerns about the use of an issuer's 
balance sheet should be addressed by the rigorous theoretical 
justification as well as by the economic interpretation of the 
resulting relationships between the inputs that are embedded in such 
structural credit risk models. See, e.g., infra note 243. However, 
allowing a more flexible threshold, as would be done under the 
commenter's suggestion, would result in increased subjectivity and 
non-uniformity of the application of the exception.
---------------------------------------------------------------------------

    The Commission agrees with the comments that market participants 
can reliably estimate the probability of default derived from a 
structural credit risk model.\109\ The Commission also acknowledges 
that probability of default estimates are not free of subjectivity and 
can vary across structural credit risk models, researchers, or 
vendors.\110\ The use of any model or market measure to estimate issuer 
creditworthiness is imperfect.\111\ The new exception's bright-line 
probability of default threshold and time horizon, however, provides 
predictability and allows for the exception to be applied consistently 
by distribution participants who are eligible to make probability of 
default determinations.\112\ As discussed below, in Part V.E.3, the 
Commission has considered other types of models, such as reduced-form 
models, which would generally provide less stable predictions than 
structural credit risk models do because they can be so flexible that 
they suffer from a lack of theoretical foundation and a lack of 
intuitive interpretation of why the defaults occur. Also, unrestricted 
use of these models might also provide more opportunity to choose a 
reduced-form model specification to enable use of the exception for 
Nonconvertible Securities. The Commission therefore is adopting

[[Page 39971]]

an exception that is based on the use of a structural credit risk model 
as this model is appropriately designed to measure creditworthiness of 
Nonconvertible Securities in Rule 101(c)(2)(i), as amended, in 
accordance with the requirements of section 939A(b).
---------------------------------------------------------------------------

    \109\ IILF Letter, at 2.
    \110\ See Proposal, 87 FR 18332.
    \111\ See Proposal, 87 FR 18332.
    \112\ Bloomberg L.P. Letter, at 2.
---------------------------------------------------------------------------

    One commenter recommended that the proposed probability of default 
threshold should be increased to 0.5% in order to capture the maximum 
amount of issuers who, currently, are eligible under the existing 
exception (i.e., the proposed threshold of 0.055% is too restrictive 
with regard to scoping in securities that currently are excepted).\113\ 
Other commenters supported the proposed probability of default 
threshold of 0.055% as reasonable.\114\ One commenter stated that the 
bright-line threshold of 0.055% will provide clarity as to what is 
expected in evaluation and documentation.\115\
---------------------------------------------------------------------------

    \113\ SIFMA Letter 1, at 10. The Commission estimated in the 
Proposal that, while the proposed threshold of 0.055% would capture 
approximately 90% of the investment grade securities in its sample 
of nonconvertible fixed income securities, a threshold of 0.5% would 
capture about 98.6% of investment grade securities. See Proposal, 87 
FR 18330, 18334.
    \114\ See Bloomberg L.P. Letter, at 2; IILF Letter, at 6.
    \115\ Bloomberg L.P. Letter, at 2. Another commenter stated that 
it is not necessary to state a precise bright-line measure. IILF 
Letter, at 6. The exception's use of a bright-line threshold, by 
imposing specific and clear requirements, helps to ensure that the 
exception captures only those Nonconvertible Securities that trade 
on the basis of yield and creditworthiness. It also helps to ensure 
that the exception is based on objective factors and can be 
consistently applied by market participants. See Bloomberg L.P. 
Letter, at 2.
---------------------------------------------------------------------------

    The Commission has considered these comments and concluded that the 
0.055% threshold appropriately calibrates the probability of default to 
determine the creditworthiness of an issuer whose Nonconvertible 
Securities trade based on their yield and creditworthiness.\116\ While 
the higher threshold of 0.5% captures a larger set of securities of 
creditworthy issuers whose securities are eligible for the existing 
Investment Grade Exception, it also allows for an exception that 
captures a larger set of securities that could be prone to manipulation 
risk in comparison to the 0.055% threshold (i.e., non-investment grade 
securities).\117\ Because the commenter's suggested 0.5% threshold, in 
comparison to the 0.055% threshold, risks capturing a majority of the 
securities that are not traded on the basis of their yield and 
creditworthiness in the same way that Nonconvertible Securities 
excepted under the existing Investment Grade Exception are traded, too 
many distributions of these types of securities would be included, 
which reflects that a 0.5% threshold may not be an appropriate 
replacement standard of creditworthiness, in accordance with of section 
939A(b).
---------------------------------------------------------------------------

    \116\ See infra Part V.B; Proposal, 87 FR 18319, 18330.
    \117\ See also Proposal, 87 FR 18334. Based on an analysis of 
the available data as of Mar. 2023, a 0.5% threshold would make the 
new exception less restrictive and would result in 124 additional 
non-investment grade securities being captured by the standard, from 
64 non-investment grade issues under the 0.055% threshold to 188 
issues under the 0.5% threshold. See infra Part V.B. These figures 
differ from those included in the Proposal because they are based on 
an analysis of the available data as of Mar. 2023, whereas the 
Proposal's figures were based on an analysis of the data available 
as of Oct. 2021. See infra Part V.B.
---------------------------------------------------------------------------

    Further, even if the 0.055% threshold does not capture the exact 
same set of Nonconvertible Securities captured by the Investment Grade 
Exception, the 0.055% threshold nevertheless identifies Nonconvertible 
Securities that are less susceptible to the manipulation that Rule 101 
is designed to prevent because they trade based on their yield and 
creditworthiness, As discussed above, Regulation M seeks to protect the 
offering price of a security during a distribution, when there are 
heightened incentives on the part of those who are involved in the 
offering process to influence the subject security's price. Because 
these Nonconvertible Securities are traded on the basis of their yield 
and creditworthiness, and are largely fungible, they are less 
susceptible to manipulation.\118\
---------------------------------------------------------------------------

    \118\ See Regulation M Adopting Release, 62 FR 527; supra note 
38.
---------------------------------------------------------------------------

    In other words, the ability of distribution participants and their 
affiliated purchasers to bid up the price of a Nonconvertible Security 
of an issuer that meets the 0.055% probability of default threshold is 
limited by investors' ability to substitute the security with other 
securities that are similar and of comparable creditworthiness. In 
contrast, a non-investment grade security that has a much higher 
probability of default tends to have idiosyncratic risks that make them 
less substitutable and hence more susceptible to manipulation. The 
threshold of 0.5% would capture more than the majority of non-
investment grade securities (approximately 69.9% of non-investment 
grade securities),\119\ which indicates that it may not be an 
appropriate measure of creditworthiness to replace the reference to 
credit ratings in Rule 101's Investment Grade Exception. Accordingly, 
the 0.055% threshold appropriately calibrates the probability of 
default to determine the creditworthiness of an issuer whose 
Nonconvertible Securities should trade based on yield and 
creditworthiness and is an appropriate substitute standard of 
creditworthiness to replace the credit ratings reference in the 
Investment Grade Exception pursuant to section 939A(b).
---------------------------------------------------------------------------

    \119\ See infra Part V.E.1.
---------------------------------------------------------------------------

    While the probability of default measure uses a threshold of 
0.055%, as was proposed, the final rule text is changed from the 
proposed rule text of ``less than 0.055%'' to state ``0.055% or less.'' 
The Commission is clarifying that a determination of a 0.055% 
probability of default is eligible for the exception, so long as all 
other conditions of the exception are met. This change is consistent 
with the estimates included in the Proposal, including with how the 
Commission calibrated the probability of default threshold in the 
Proposal.\120\
---------------------------------------------------------------------------

    \120\ See Proposal, 87 FR 18330; see also Proposal, 87 FR 18319, 
18332.
---------------------------------------------------------------------------

    The same commenter also suggested that the probability of default 
calculations should be permitted to be made within a specified duration 
of time in advance of pricing (rather than as of the day of determining 
the offering price), for example, within 10 calendar days prior to 
pricing of the offering, similar to the approach taken with respect to 
average daily trading volume (``ADTV'') calculations, to afford 
distribution participants adequate time to adjust their market 
activities as necessary.\121\ The Commission acknowledges that lead 
managers who make probability of default determinations pursuant to 
Rule 101(c)(2)(i) may need additional time prior to the pricing of an 
offering to make the required calculations. However, in light of the 
comments received, the 10-calendar-day period suggested by the 
commenter would be unnecessarily long for the lead manager to determine 
and document in writing the issuer's probability of default because the 
determination is likely to be highly automated.\122\ In addition, the 
suggested 10-calendar-day period may not encourage as timely of 
information about the issuer as possible if the model inputs are taken 
farther away from the day of the determination of the offering price, 
as proposed.
---------------------------------------------------------------------------

    \121\ SIFMA Letter 1, at 10.
    \122\ See infra Part VI.C.1. For example, because commonly 
available spreadsheet software can be used to calculate the 
probability of default, lead managers would not need 10 calendar 
days to derive an issuer's probability of default. See, e.g., 
Proposal, 87 FR 18319.
---------------------------------------------------------------------------

    The Commission is, therefore, modifying the proposed time horizon 
of ``the day of the determination of the

[[Page 39972]]

offering pricing'' to allow the lead manager to make its probability of 
default determination as of ``the sixth business day immediately 
preceding the determination of the offering price'' for purposes of the 
new exception. This change from the proposed time horizon of ``the day 
of the determination of the offering price'' is being made in response 
to comment that additional time is needed because ``[i]t would be very 
damaging to the issuer to launch a re-opening, subsequently determine 
that there is no exception under the probability of default 
calculation, and then have to extend the pricing of the offering by at 
least one (or five) business days.'' \123\ The Commission agrees with 
the commenter that more time would be useful to address the potential 
of an offering by at least one to five business days but is concerned 
that the model inputs supporting the probability of default 
determination may become stale with additional time beyond that.\124\ 
Accordingly, the Commission is extending the time horizon to allow for 
the potential of an offering being extended up to five business days. 
Further, this will allow the determination to be made before Regulation 
M's otherwise applicable five-business-day restricted period (i.e., 
preceding the determination of the offering price) and should provide a 
sufficient amount of additional time for the lead manager to account 
for any relevant market activities and timely information regarding the 
issuer as a model input in determining the probability of default.\125\ 
This essentially allows these distribution participants, in relation to 
the proposed ``day of the determination of the offering price'' 
requirement, five additional business days, as defined in Rule 100(b), 
before the actual pricing and launch of the offering to make the 
probability of default determination.
---------------------------------------------------------------------------

    \123\ SIFMA Letter 1, at 11.
    \124\ See, e.g., infra Part V.B (discussing how probabilities of 
default implied by structural credit risk models generally use 
current estimates of equity valuation and volatility based on the 
recent trading activity, and hence incorporate more recent news 
affecting the valuation and perceived volatility of the firm).
    \125\ See, e.g., Proposal, 87 FR 18330.
---------------------------------------------------------------------------

    In view of the nature and trading characteristics of Nonconvertible 
Securities, the impact, if any, of a corporate or market event in the 
intervening five business days would be unlikely to result in the 
manipulation that Regulation M seeks to prevent. Nonconvertible 
Securities are priced and traded differently than equity securities in 
that the focus (with Nonconvertible Securities) is placed on receiving 
periodic interest payments during the life of the instrument rather 
than on any potential equity upside or increase in the current trading 
or offering price. Therefore, trading activity in Nonconvertible 
Securities at or around the time of the distribution is unlikely to 
influence the pricing or trading of such securities, particularly 
during Regulation M's (otherwise applicable) restricted period.
    Accordingly, the Commission is adopting, with targeted 
modifications in consideration of the comments received, as discussed 
in this Part, as well as certain technical changes,\126\ a new 
exception in Rule 101(c)(2)(i) for Nonconvertible Securities of issuers 
for which the probability of default, estimated as of the sixth 
business day immediately preceding the determination of the offering 
price and over the horizon of 12 full calendar months from such day, is 
0.055% or less, as determined and documented, in writing, by the 
distribution participant acting as the lead manager (or in a similar 
capacity) of a distribution, as derived from a structural credit risk 
model. For the reasons discussed above, in this Part, and as supported 
by an analysis of the probability of default and investment grade 
credit ratings of a sample of Nonconvertible Securities available on 
the market,\127\ the standard used in the new exception is an 
appropriate substitute standard of creditworthiness to replace the 
credit ratings reference in the Investment Grade Exception for 
Nonconvertible Securities.
---------------------------------------------------------------------------

    \126\ The final amendments make the technical change of deleting 
the word ``the'' from the beginning of the exception in order to 
mirror the beginning of new Rule 101(c)(2)(ii). The final rule 
amendments also make an edit to use the words ``as derived from'' 
instead of ``by using,'' as proposed, to clarify that a structural 
credit risk model must be the only method of determining an issuer's 
probability of default, as opposed to one method among others. This 
change also conforms the standard to how it was discussed in the 
Proposal. See, e.g., Proposal, 87 FR 18318-19. The final amendments 
also make a conforming edit from the proposed rule text to add the 
word ``full'' preceding the time horizon of 12 calendar months to 
make the phrasing of the exception's time horizon consistent with 
Regulation M's other time horizons. See, e.g., 17 CFR 242.100(b) 
(defining the terms ``ADTV,'' which uses a time horizon of ``two 
full calendar months,'' and ``principal market,'' which uses the 
time horizon of ``12 full calendar months'').
    \127\ See infra Parts V.A through E.
---------------------------------------------------------------------------

    Finally, the Commission is adopting, substantially as proposed, 
with certain clarifying \128\ and technical \129\ changes, a definition 
under Rule 100(b) of Regulation M for the term ``structural credit risk 
model'' that means ``any commercially or publicly available model that 
calculates, based on an issuer's balance sheet, the probability that 
the value of the issuer will fall below the threshold at which the 
issuer would fail to make scheduled debt payments, at or by the 
expiration of a defined period.'' \130\ Accordingly, a covered person's 
reliance on Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as amended, is 
conditioned on a probability of default determination that was derived 
from any commercially or publicly available structural credit risk 
model that calculates, based on an issuer's balance sheet, the 
probability that the value of the issuer will fall below the threshold 
at which the issuer would fail to make scheduled debt payments, at or 
by the expiration of a defined period.
---------------------------------------------------------------------------

    \128\ The Commission is making clarifying changes to the 
proposed definition of the term ``structural credit risk model'' 
that conform it to its description in the Proposal. As discussed 
above, in Part II.A.1, the Proposal stated that a structural credit 
risk model ``provide[s] a probability that a firm's assets will fall 
below the Default Point at or by the expiration of a defined period 
of time.'' Proposal, 87 FR 18317. The final rule amendments conform 
the proposed definition to that description and clarify that the 
``threshold'' referenced in the proposed definition is the Default 
Point.
    \129\ The Commission is also making a non-substantive, technical 
change from the proposed definition of the term ``structural credit 
risk model.'' As discussed below, because the final rule amendments 
include the term ``structural credit risk model'' in both new Rule 
101(c)(2)(i) and new Rule 102(d)(2)(i), the Commission is adding to 
Rule 100(b) a definition for the term ``structural credit risk 
model.'' The addition of this definition in Rule 100(b) does not 
change the definition of the term ``structural credit risk model'' 
but rather simplifies the final text of new Rules 101(c)(2)(i) and 
102(d)(2)(i) by obviating the need for a proviso containing a 
definition in each of those rules. Accordingly, use of the term 
``structural credit risk model'' is identical across new Rules 
101(c)(2)(i) and 102(d)(2)(i).
    \130\ 17 CFR 242.100(b).
---------------------------------------------------------------------------

2. Rule 101(c)(2)(ii): Asset-Backed Securities Offered Pursuant to an 
Effective Shelf Registration Statement Filed on Form SF-3
    The Commission proposed an amendment to add a new exception in Rule 
101(c)(2)(ii) for asset-backed securities that are offered pursuant to 
an effective shelf registration statement filed on Form SF-3.
    In 2014, the Commission adopted shelf eligibility criteria for 
asset-backed securities offerings registered on new Form SF-3 in part 
to implement section 939A(b) of the Dodd-Frank Act.\131\ The Commission 
designed the shelf eligibility requirements to help ensure a

[[Page 39973]]

certain ``quality and character'' in light of the requirement to reduce 
regulatory reliance on credit ratings.\132\ The shelf eligibility 
requirements included in Form SF-3 are designed to help ensure that the 
securitization is designed to produce expected cash flows that are 
sufficient to service payments or distributions in accordance with 
their terms; \133\ that obligated parties more carefully consider the 
characteristics and quality of the assets that are included in the 
pool; \134\ that asset-backed securities shelf offerings have 
transactional safeguards and features that make those certain 
securities appropriate to be issued without prior Commission staff 
review; \135\ and that issuers design and prepare asset-backed 
securities offerings with greater oversight and care.\136\ The asset-
backed securities offered pursuant to an effective shelf registration 
statement filed on Form SF-3 should trade primarily on the basis of 
yield and creditworthiness, rather than on the identity of a particular 
issuer and its idiosyncratic risk.
---------------------------------------------------------------------------

    \131\ See Regulation AB II Adopting Release.
    \132\ See Regulation AB II Adopting Release, 79 FR 57189.
    \133\ See Regulation AB II Adopting Release, 79 FR 57267.
    \134\ See Regulation AB II Adopting Release, 79 FR 57278.
    \135\ See Regulation AB II Adopting Release, 79 FR 57283.
    \136\ Regulation AB II Adopting Release, 79 FR 57265, 57285.
---------------------------------------------------------------------------

    One commenter supported the adoption of the proposed exception for 
asset-backed securities that are offered pursuant to an effective shelf 
registration statement filed on Form SF-3 and stated that it 
appreciated the straightforward nature of the standard, which allows 
all interested parties to easily determine whether the exception is 
available.\137\
---------------------------------------------------------------------------

    \137\ See SIFMA Letter 1, at 11.
---------------------------------------------------------------------------

    Another commenter disagreed with the Commission's statement that an 
exception for asset-backed securities that is based on a probability of 
default threshold may be unfeasible.\138\ This commenter stated that 
market participants are able to estimate the probability of default for 
these securities, not only using statistical models but also based on 
market measures such as credit spreads.\139\ In response to Request for 
Comment (``RFC'') 29,\140\ this commenter stated that a probability of 
default standard based on market measures would have indicated that the 
exception to Regulation M no longer applied for certain debt securities 
during the months leading up to the collapse of Lehman Brothers, and 
available data shows that, if such a market measure-based standard had 
been implemented, the exceptions for Regulation M would not have been 
available for debt securities as early as fall 2007.\141\ Despite this 
example, the Commission has considered this comment and determined that 
an exception for asset-backed securities that is based on a structural 
credit risk model to derive an issuer's probability of default would be 
unfeasible because distribution participants (including those acting as 
the lead managers) may not be able to collect all of the information 
required to calculate the probability of default, such as the value and 
volatility of the equity.\142\ In other words, practical challenges of 
obtaining reliable fundamental information about the equity would make 
a probability of default measure unfeasible for an exception for asset-
backed securities that trade on the basis of their yield and 
creditworthiness. The Commission has also determined that a measure 
based on credit spreads or the use of other models, such as reduced-
form models, would not be appropriate to use due to their flexible or 
unstructured nature, which could result in a standard that can be used 
to abuse the exception.\143\
---------------------------------------------------------------------------

    \138\ IILF Letter, at 6.
    \139\ IILF Letter, at 6-7 (stating that asset-backed securities 
are widely traded and have frequently quoted prices and credit 
spreads and that it is straightforward to calculate the probability 
of default based on these market measures). For the reasons 
discussed above, in Part II.A.1, the Commission is requiring that an 
issuer's probability of default be derived from a structural credit 
risk model, and not from other market measures.
    \140\ Proposal, 87 FR 18323 (requesting comment on whether a 
probability-of-default-based standard would be appropriate for the 
exception for asset-backed securities; whether there are models that 
are used to calculate a probability of default threshold for asset-
backed securities that would be relevant to consider based on the 
type of security involved and, if so, what the threshold should be; 
what benefits this approach would provide; what other concerns this 
approach could raise; and how this approach would address potential 
conflicts of interest involving the distribution participant or 
affiliated purchaser making the determination).
    \141\ IILF Letter, at 7 (citing Flannery, Houston & Partnoy, 
Credit Default Swap Spreads, 158 U. PA. L. REV. 2085, 2087 (2010)). 
The available data referenced in the commenter's statement, as well 
as the article the commenter cited, pertains to debt and credit 
default swaps. It does not appear from the cited article that the 
analysis performed related to credit default swaps was performed 
with regard to asset-backed securities. Further, this commenter did 
not provide similar information about asset-backed securities.
    \142\ See infra Part VI.B (discussing model inputs).
    \143\ See infra Part V.E.6.
---------------------------------------------------------------------------

    The same commenter stated that ``any final rules governing asset-
backed securities also could reference expected recovery in the event 
of default and default correlation.'' \144\ The commenter suggested to 
require under the exception that the applicable market participant 
determine and document a conclusion that the credit risk associated 
with a security was ``minimal'' (or some other similar standard) based 
on these variables, without any requirement that they use a particular 
model.\145\ The commenter's suggested exception, by conditioning 
reliance on a list of variables and a judgment of ``minimal'' credit 
risk, without any bright-line requirements to help deter abuse of the 
exception through self-serving conclusions, would not be sufficiently 
objective.
---------------------------------------------------------------------------

    \144\ IILF Letter, at 8.
    \145\ IILF Letter, at 8.
---------------------------------------------------------------------------

    The Commission continues to believe that its original basis for 
excepting securities of a certain quality and character is appropriate 
and that such securities are less at risk of the manipulation that 
Regulation M addresses.\146\ As discussed above, the Commission 
excepted investment grade asset-backed securities from Rule 101 because 
such securities trade primarily on the basis of yield and 
creditworthiness (traditionally measured by credit ratings).\147\ In 
providing this rationale, the Commission stated that the principal 
focus of investors in the asset-backed securities market is on the 
structure of a class of securities and the nature of the assets pooled 
to serve as collateral for those securities rather than on the identity 
of a particular issuer.\148\ The Commission also stated that Rule 
101(c)(2) excepts investment grade securities that are ``primarily 
serviced by the cashflows of a discrete pool of receivables or other 
financial assets, either fixed or revolving, that by their terms 
convert into cash within a finite time period plus any rights or other 
assets designed to assure the servicing or timely distribution of 
proceeds to the security holders.'' \149\
---------------------------------------------------------------------------

    \146\ See Regulation M Adopting Release, 62 FR 527; see also 
Prohibitions Against Trading by Persons Interested in a 
Distribution, Release No. 34-19565 (Mar. 4, 1983) [48 FR 10628, 
10631 (Mar. 14, 1983)] (stating that the ``fungibility'' of certain 
types of securities makes manipulation of their price very 
difficult).
    \147\ See Regulation M Adopting Release, 62 FR 527.
    \148\ See Regulation M Adopting Release, 62 FR 527.
    \149\ See Regulation M Adopting Release, 62 FR 527 (citations 
omitted). The Commission stated that such rationale also applies to 
the existing identical exception in Rule 102(d)(2) of Regulation M. 
Regulation M Adopting Release, 62 FR 531.
---------------------------------------------------------------------------

    As discussed above, in this Part, the practical challenge of 
obtaining reliable fundamental information about the equity makes a 
probability of default determination difficult or infeasible. The 
Commission believes that an appropriate and pragmatic approach is

[[Page 39974]]

to add an exception based on Form SF-3, as proposed, because it 
sufficiently focuses on creditworthiness.\150\ In addition, as stated 
in the Proposal, a standard that relies on Form SF-3 with respect to 
Nonconvertible Securities would not be appropriate because the 
transaction requirements included in Form SF-3 are relevant only to 
asset-backed securities and thus would not be a sufficient measure of 
creditworthiness for securities that are not subject to the Form SF-3 
transaction requirements.\151\
---------------------------------------------------------------------------

    \150\ See, e.g., supra notes 60-61 (contrasting the focus of 
creditworthiness in the eligibility criteria and offering 
requirements included in Form SF-3 with that of other Commission 
Forms, such as Form S-3 and Form F-3).
    \151\ See Proposal, 87 FR 18321.
---------------------------------------------------------------------------

    The transaction requirements included in Form SF-3 allow for shelf 
offerings of only those asset-backed securities that share the 
qualities and characteristics of the investment grade asset-backed 
securities currently excepted in Rule 101(c)(2): with respect to either 
set of securities, the principal focus of investors is the structure of 
a class of securities and the nature of the assets pooled to serve as 
collateral for those securities, rather than on the identity of a 
particular issuer.\152\ First, eligibility for offering securities 
pursuant to a Form SF-3 is limited, in part, by the percentage of 
delinquent assets and, for certain lease-backed securitizations, by the 
portion of the pool attributable to the residual value of the physical 
property underlying the leases.\153\ For an asset-backed securities 
offering with an effective Form SF-3, delinquent assets cannot 
constitute 20% or more of the asset pool. Delinquent assets may not 
convert into cash within a finite period of time, as required by the 
definition of ``asset-backed security,'' because they are not 
performing in accordance with their terms and management or that other 
action may be needed to convert the assets into cash. However, as the 
Commission stated at the time it adopted the 20% delinquency limitation 
for shelf eligibility, in principle, asset-backed securities should be 
primarily dependent on the pool of assets self-liquidating instead of 
on the ability of the entity performing collection services.\154\ The 
application of the limitation on delinquent assets was designed to 
ensure that attention is focused on the ability of collateral of the 
underlying asset pool to generate cash flow rather than on the identity 
of the issuer and its ability to convert those assets into cash,\155\ 
consistent with the Commission's original basis for excepting 
investment grade asset-backed securities from Rule 101.\156\
---------------------------------------------------------------------------

    \152\ See supra note 149.
    \153\ See 17 CFR 239.45(b)(v), (vi); Form SF-3, General 
Instruction I.B.1(e) and (f).
    \154\ Asset-Backed Securities, Release No. 33-8518 (Dec. 22, 
2004) [70 FR 1506, 1517 (Jan. 7, 2005)] (``Regulation AB Release''). 
In adopting the 20% delinquency concentration level, the Commission 
codified a staff position that an asset-backed security will not 
fail to meet the definition of ``asset-backed security'' solely 
because such a security is supported by assets having total 
delinquencies of up to 20% at the time of the proposed offering. See 
Regulation AB Release, 70 FR 1517 (citing Bond Mkt. Ass'n, SEC Staff 
No-Action Letter, 1997 WL 634124 (Oct. 8, 1997) (``BMA NAL'')). This 
threshold was the same threshold that was applied to certain other 
matters affecting registration and disclosure requirements for 
asset-backed securities (e.g., non-recourse commercial mortgage 
securitizations, pooling of corporate debt securities, and 
securitizations involving third-party credit enhancement). See BMA 
NAL, 1997 WL 634124, at * 3. The staff position was based on the 
premise that such a threshold for total delinquency concentration 
would, by itself, not present a materially greater risk of asset 
non-performance or default at the security level. See BMA NAL, 1997 
WL 634124, at * 4.
    \155\ See Regulation AB Release, 70 FR 1517.
    \156\ See Regulation M Adopting Release, 62 FR 527.
---------------------------------------------------------------------------

    Second, Form SF-3 includes certain transaction requirements with 
respect to the structure of the asset-backed security being offered. 
Such structural requirements include: (1) a certification by the 
depositor's chief executive officer that, among other things, the 
securitization structure provides a reasonable basis to conclude that 
the expected cash flows are sufficient to service payments or 
distributions in accordance with their terms; (2) a review of the 
asset-backed security's pool of assets upon the occurrence of certain 
triggering events, including delinquencies, by a person that is 
unaffiliated with certain transaction parties, such as the sponsor, 
depositor, servicer, trustee, or any of their affiliates; and (3) a 
dispute resolution provision, contained in the underlying transaction 
documents, for any repurchase request.\157\ When adopting the 
transaction requirements included in Form SF-3, the Commission stated 
that sponsors may have an increased incentive to carefully consider the 
characteristics of the assets underlying the securitization and 
accurately disclose these characteristics at the time of offering.\158\ 
The Commission also stated that investors should benefit from the 
reduced losses associated with nonperforming assets because, as a 
result of this new shelf requirement, sponsors will have less of an 
incentive to include nonperforming assets in the pool.\159\ Because the 
transactional safeguards included in Form SF-3 provide incentives for 
obligated parties to, among other things,\160\ more carefully consider 
the characteristics and quality of the assets that are included in the 
pool,\161\ asset-backed securities that are offered pursuant to an 
effective Form SF-3 should trade based on their yield and 
creditworthiness rather than on the identity of a particular 
issuer.\162\
---------------------------------------------------------------------------

    \157\ Form SF-3, General Instruction I.B.1(a)-(c).
    \158\ See Regulation AB II Adopting Release, 79 FR 57283.
    \159\ See Regulation AB II Adopting Release, 79 FR 57283.
    \160\ See supra notes 132-136.
    \161\ See Regulation AB II Adopting Release, 79 FR 57278.
    \162\ See, e.g., Regulation AB II Adopting Release, 79 FR 57277-
78.
---------------------------------------------------------------------------

    The requirement regarding an effective shelf registration statement 
filed on Form SF-3 is an appropriate substitute for the reference to 
credit ratings in the Investment Grade Exception because the standard 
is designed to limit eligibility for that exception to only those 
asset-backed securities that should trade based on their yield and 
creditworthiness due to their particular qualities and characteristics. 
Because the ability of distribution participants and their affiliated 
purchasers to bid up the price of an asset-backed security offered 
pursuant to an effective Form SF-3, during a distribution, is limited 
by a market participant's ability to substitute the security with other 
securities that are similar and of comparable creditworthiness,\163\ 
such a security is less susceptible to the types of manipulation that 
Regulation M seeks to prevent. The application of the transaction 
requirements included in the Commission's Form SF-3, therefore, should 
result in the offering of asset-backed securities that have similar 
qualities and characteristics to the investment grade asset-backed 
securities currently excepted under the existing provision in Rule 
101(c)(2). In addition, the exception for asset-backed securities that 
are offered pursuant to an effective shelf registration statement filed 
on Form SF-3 carries over the standard of creditworthiness included in 
the Commission's Form SF-3 and helps to implement the mandate that, to 
the extent feasible, uniform standards of creditworthiness be 
used.\164\
---------------------------------------------------------------------------

    \163\ See Regulation M Adopting Release, 62 FR 527.
    \164\ Public Law 111-203, sec. 939A(b) (requiring agencies to 
``seek to establish, to the extent feasible, uniform standards of 
credit-worthiness for use by each such agency, taking into account 
the entities regulated by each such agency and the purposes for 
which such entities would rely on such standards of credit-
worthiness'').

---------------------------------------------------------------------------

[[Page 39975]]

    Accordingly, the Commission is adopting, substantially as proposed, 
with a technical change,\165\ an amendment to add a new exception in 
Rule 101(c)(2)(ii) for asset-backed securities that are offered 
pursuant to an effective shelf registration statement filed on Form SF-
3.
---------------------------------------------------------------------------

    \165\ The final rule amendments make a non-substantive, 
technical change that replaces the proposed reference to ``17 CFR 
239.45'' with a reference to ``Sec.  239.45 of this chapter'' when 
referencing Form SF-3.
---------------------------------------------------------------------------

B. Rule 102(d)(2) of Regulation M: Implementing Section 939A(b) in 
Certain Exceptions for Issuers and Selling Security Holders

    The Commission proposed to remove, without replacing, the 
Investment Grade Exception in Rule 102(d)(2) of Regulation M. The 
Commission stated in the Proposal that this removal without replacement 
was appropriate given that the retention of an exception for 
creditworthy Nonconvertible Securities and asset-backed securities 
would not likely be necessary to facilitate orderly distributions or 
limit disruptions in the trading market in light of issuers' limited 
market access needs and the apparent limited reliance on Rule 102's 
Investment Grade Exception, coupled with the incentive for issuers, 
selling security holders, and their affiliated purchasers to manipulate 
the market for the distributed security, regardless of the security's 
credit quality.\166\
---------------------------------------------------------------------------

    \166\ See Proposal, 87 FR 18323-24.
---------------------------------------------------------------------------

    Two commenters supported the proposed elimination, without 
replacement, of Rule 102(d)(2)'s Investment Grade Exception.\167\ One 
commenter stated that it is ``appropriate in contexts where deference 
arguably should be made to independent decisions and judgment by market 
participants, without the crutch of reliance on credit ratings.'' \168\
---------------------------------------------------------------------------

    \167\ See Better Markets Letter, at 3-4 (stating that the 
proposal to eliminate, without replacing, the exception in Rule 102 
for certain investment grade securities is appropriate because 
issuers and selling security holders have comparatively strong 
incentives to manipulate the price of the distributed security); 
IILF Letter, at 7. As discussed below, in Part II.B.1, the new 
exception for Nonconvertible Securities takes account of this 
consideration.
    \168\ IILF Letter, at 7.
---------------------------------------------------------------------------

    One commenter objected to the proposed elimination, without 
replacement, of the Investment Grade Exception in Rule 102(d)(2) and 
stated that the ``continued availability of such an exception would be 
important to broker-dealers who are affiliated with an issuer but are 
not, for whatever reasons, serving as an underwriter or other 
participant in connection with the distribution.'' \169\ This commenter 
stated that the Commission, instead, should substitute the existing 
standard in Rule 102(d)(2) with ``the same standards as used for 
purposes of the exceptions under Rule 101(c)(2).'' \170\ For reasons 
explained below, the Commission agrees with this commenter and is 
replacing the Investment Grade Exception in Rule 102, rather than 
eliminating the exception, without replacement, as proposed. The 
Commission continues to believe that its original basis for excepting 
Nonconvertible Securities and asset-backed securities of a certain 
quality and character from Rule 102's prophylactic prohibitions is 
appropriate and that the substitute standards discussed below are 
appropriate to ensure that those securities are less at risk of the 
manipulation that Regulation M addresses.\171\
---------------------------------------------------------------------------

    \169\ SIFMA Letter 1, at 12-13. This commenter stated that the 
Investment Grade Exception in Rule 102 is relied upon in the context 
of sticky offerings and re-openings of debt issuances. See SIFMA 
Letter 1, at 4.
    \170\ See SIFMA Letter 1, at 12. With regard to replacing the 
Investment Grade Exception in Rule 102 pursuant to section 939A(b)'s 
requirements, the Commission requested comment on whether it should 
adopt an exception based on either the probability-of-default-based 
standard for Nonconvertible Securities or the standard for asset-
backed securities that are offered pursuant to an effective shelf 
registration statement filed on Form SF-3 instead of removing the 
Investment Grade Exception, without substituting an alternative, and 
whether it should adopt an exception in Rule 102 if a distribution 
participant determines that a security is an excepted security 
pursuant to Rule 101(c)(2). Proposal, 87 FR 18324 (request for 
comment (RFC) 35). One commenter replied, in response to RFC 35, 
that, to the extent the Commission receives comments that market 
participants on their own cannot make decisions and judgments about 
credit risk related to Rule 102, an exception based on probability 
of default would be a viable alternative. See IILF Letter, at 7. The 
Commission did not receive any such comment in response to the 
Proposal.
    \171\ See Regulation M Adopting Release, 62 FR 527; see also 
Prohibitions Against Trading by Persons Interested in a 
Distribution, Release, No. 34-19565 (Mar. 4, 1983) [48 FR 10628, 
10631 (Mar. 14, 1983)].
---------------------------------------------------------------------------

    The Commission is adopting rule amendments that remove the 
Investment Grade Exception from Rule 102 of Regulation M and substitute 
in its place exceptions based on alternative standards of 
creditworthiness to except Nonconvertible Securities. The reasons for 
adding these new exceptions are discussed below with regard to 
Nonconvertible Securities, in Part II.B.1, and asset-backed securities, 
in Part II.B.2.
1. Rule 102(d)(2)(i): Nonconvertible Securities of Issuers Who Meet a 
Specified Probability of Default Threshold
    The Commission acknowledges that eliminating the exception, without 
replacement, may impact entities, such as broker-dealers who are not 
distribution participants (and are not eligible to quality for an 
exception under Rule 101) but may qualify for a comparable exception 
under Rule 102 as a result of being an affiliate of an issuer or 
selling security holder and meeting the exception's conditions. The 
continued availability of an exception for the Nonconvertible 
Securities will also provide issuers and selling security holders with 
more flexibility during distributions as compared to the Proposal. The 
elimination, without replacement, of the Investment Grade Exception 
from Rule 102 for issuers and selling security holders could increase 
issuance costs or deter market participants from issuing Nonconvertible 
Securities with low manipulation risk.\172\
---------------------------------------------------------------------------

    \172\ See Proposal, 87 FR 18333.
---------------------------------------------------------------------------

    In adopting the Investment Grade Exception, the Commission stated 
that it determined to include the Investment Grade Exception in Rule 
102 based, in part, on the rationales indicated for an identical 
exception to Rule 101.\173\ The Commission excepted investment grade 
Nonconvertible Securities from Rule 101 ``based on the premise that 
these securities traded on the basis of their yield and credit ratings, 
are largely fungible and, therefore, are less likely to be subject to 
manipulation.'' \174\ As discussed above, in Part II.A.1, the new 
exception in Rule 101(c)(2)(i) for Nonconvertible Securities of issuers 
for which the probability of default, estimated as of the sixth 
business day immediately preceding the determination of the offering 
price and over the horizon of 12 full calendar months from such day, is 
0.055% or less, as determined and documented, in writing, by the 
distribution participant acting as the lead manager (or in a similar 
capacity) of a distribution, as derived from a structural credit risk 
model, is an appropriate substitute standard of creditworthiness in 
place of the reference to credit ratings in the Investment Grade 
Exception for Nonconvertible Securities.
---------------------------------------------------------------------------

    \173\ Regulation M Adopting Release, 62 FR 531.
    \174\ Regulation M Adopting Release, 62 FR 527.
---------------------------------------------------------------------------

    The standard of creditworthiness, which was the basis of the 
Investment Grade Exception for Nonconvertible Securities in Rule 102, 
is still appropriate to use as the basis of an

[[Page 39976]]

exception to Rule 102 for Nonconvertible Securities.\175\ In addition, 
the standard of creditworthiness used in the exception in Rule 
101(c)(2)(i), as amended, is an appropriate standard of 
creditworthiness to use in place of the reference to credit ratings in 
the Investment Grade Exception in Rule 102 pursuant to the requirements 
of section 939A(b). That standard, which is based on an issuer's 
probability of default, is designed to identify Nonconvertible 
Securities that are less susceptible to the manipulation that 
Regulation M is designed to prevent because they trade based on their 
yield and creditworthiness, as determined by the current financial 
condition of the issuer. However, given that issuers and selling 
security holders have the greatest interest in an offering's 
outcome,\176\ regardless of the credit quality of the security,\177\ it 
would not be appropriate for the exception to permit those parties to 
make their own probability of default determinations (by their own or a 
third party calculation) in order to meet the conditions of the 
exception.
---------------------------------------------------------------------------

    \175\ See supra note 13 and accompanying text.
    \176\ See Regulation M Adopting Release, 62 FR 530.
    \177\ See Proposal, 87 FR 18323.
---------------------------------------------------------------------------

    Therefore, Rule 102(d)(2)(i), as amended, uses the same bright-line 
test for excepting Nonconvertible Securities. For issuers, selling 
security holders, and their affiliated purchasers to use the exception 
in Rule 102(d)(2)(i), as amended, however, they must rely on the 
probability of default determination made by the distribution 
participant acting as the lead manager (or in a similar capacity) \178\ 
of the distribution and documented in writing pursuant to Rule 
101(c)(2)(i), as amended. Rule 102(d)(2)(i), as amended, does not 
permit reliance on the exception if issuers, selling security holders, 
or their affiliated purchasers make the required probability of default 
determinations themselves, or rely on a determination made by a non-
lead manager or any other third party. This condition to the exception 
means that issuers, selling security holders, and their affiliated 
purchasers would not be able to rely on the new Rule 102(d)(2)(i) 
exception when selling securities directly, unless a lead manager is 
involved in the distribution and had made (and documented) the 
qualifying probability of default determination. This condition 
provides for the continued availability of an exception under Rule 102 
for creditworthy Nonconvertible Securities for broker-dealers who are 
affiliated with an issuer but are not serving as an underwriter or 
other participant in connection with the distribution. At the same 
time, this condition is designed to prevent abuse of the exception by 
issuers, selling security holders, and their affiliated purchasers by 
taking into account that these market participants have the greatest 
interest in an offering's outcome and generally do not have the same 
market access needs as underwriters.\179\ In addition, the condition 
regarding lead-manager probability of default determinations in new 
Rule 102(d)(2)(i) is consistent with the condition regarding lead-
manager probability of default determinations in new Rule 101(c)(2)(i).
---------------------------------------------------------------------------

    \178\ See supra note 88.
    \179\ See Regulation M Adopting Release, 62 FR 530.
---------------------------------------------------------------------------

    Even though probability of default determinations made by or 
directly for issuers or selling security holders or affiliated 
purchasers cannot be used in order for such parties to rely on the new 
exception in Rule 102(d)(2)(i) for Nonconvertible Securities, these 
parties would, however, be able to avail themselves of the exception in 
reliance on a probability of default determination made by the 
distribution participant acting as the lead manager (or in a similar 
capacity) of the distribution pursuant to Rule 101(c)(2)(i), as 
amended. Similar to how the lead or managing underwriter in a firm 
commitment offering communicates certain pricing, allocation, and other 
distribution-related information to the issuer or selling security 
holder in connection with that particular distribution, the lead 
managing underwriter's communications regarding its probability of 
default determination may vary based on the parties and their prior 
course of conduct as to the frequency and manner or mode of such 
communication.
    However, Rule 101(c)(2)(i), as amended, does not require that the 
lead manager making the probability of default determination share the 
determination with the issuer, selling security holders, or their 
affiliated purchasers in order for those parties to rely on the 
exception. Therefore, issuers, selling security holders, and their 
affiliated purchasers will not be able to rely on the exception for 
Nonconvertible Securities if the lead manager does not share the 
probability of default determination or there is no distribution 
participant to act as the lead manager for the distribution. With 
regard to the types of distributions covered, as a result of the 
condition related to lead-manager probability of default 
determinations, the exception for Nonconvertible Securities is 
available to a subset of distributions covered under the existing 
Investment Grade Exception \180\ but more distributions than what was 
covered under the Proposal given that the Proposal would have removed, 
and not replaced, the Investment Grade Exception in Rule 102.\181\ The 
estimated costs associated with the condition related to the lead 
manager making the probability of default determination are included 
below, in Part V.C.1.
---------------------------------------------------------------------------

    \180\ See, e.g., Proposal, 87 FR 18330. As discussed above, in 
Part II.A.1, this may be the case in, for example, self-underwritten 
offerings, at-the-market offerings, or other shelf offerings, to the 
extent such an offering meets the definition of a ``distribution'' 
under Rule 100(b) of Regulation M. With regard to shelf offerings, 
each takedown is to be individually examined to determine whether 
such offering constitutes a ``distribution.'' Regulation M Adopting 
Release, 62 FR 526. If, as a result of the amendments, the exception 
for Nonconvertible Securities is no longer available in connection 
with a distribution, and if no other exception is available, Rule 
102's prohibitions would apply. Accordingly, an issuer and all of 
its affiliated purchasers would be subject to the applicable 
restricted period of Rule 102 when sales off a shelf by an issuer, 
or by any affiliated purchaser, constitute a distribution of 
securities. Similarly, when a selling security holder sells off the 
shelf and such sales constitute a distribution, all other shelf 
security holders who are affiliated purchasers of the selling 
security holder would be subject to the applicable restricted period 
of Rule 102. See Regulation M Adopting Release, 62 FR 531.
    \181\ See Proposal, 87 FR 18333.
---------------------------------------------------------------------------

    As discussed below, in Part II.C, broker-dealers who rely on the 
new exception in Rule 102(d)(2)(i) are required to preserve the written 
probability of default determination made pursuant to Rule 
101(c)(2)(i), as amended. This record preservation requirement could 
help facilitate Commission examinations of broker-dealers who rely on 
Rule 102(d)(2)(ii), as amended, and could deter their misuse of the 
exception through relying on determinations that do not meet the 
conditions of the exception or through relying on the exception when no 
determination has been made.
2. Rule 102(d)(2)(ii): Asset-Backed Securities Offered Pursuant to an 
Effective Shelf Registration Statement Filed on Form SF-3
    The Commission is adopting in new Rule 102(d)(2)(ii) an exception 
for asset-backed securities that are offered pursuant to an effective 
shelf registration statement filed on Form SF-3. The Commission stated 
in the Proposal that the incentive for issuers, selling shareholders, 
and their affiliated purchasers to manipulate the market of a 
distributed security exists regardless of the credit quality of the 
security.\182\

[[Page 39977]]

While this incentive may exist, transaction requirements included in 
Form SF-3 allow for shelf offerings of only those asset-backed 
securities that share the qualities and characteristics of the 
investment grade asset-backed securities that meet the Investment Grade 
Exception \183\ and thus are less likely to be subject to the type of 
manipulation that Regulation M seeks to address.
---------------------------------------------------------------------------

    \182\ See Proposal, 87 FR 18323.
    \183\ See supra notes 153-161 and accompanying text.
---------------------------------------------------------------------------

    In addition, in contrast to how Nonconvertible Securities would be 
excepted, whether an asset-backed security is rated investment grade is 
an objective, observable fact, as is whether an asset-backed security 
is offered pursuant to an effective shelf registration statement filed 
on Form SF-3. Reliance on the new exception in Rule 102(d)(2)(i) does 
not require issuers, selling security holders, or their affiliated 
purchasers to make a calculation in determining whether the subject 
asset-backed security meets the conditions of that exception (i.e., 
that the asset-backed security is offered pursuant to an effective 
shelf registration statement filed on Form SF-3), in contrast to how 
reliance on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i) 
for Nonconvertible Securities is conditioned on a calculation 
determining whether the issuer's probability of default meets the 
specified threshold. Because the Investment Grade Exception for asset-
backed securities does not focus on the potential interests of those 
covered persons seeking to rely on the exception but rather the 
particular qualities of the securities themselves (i.e., that the 
asset-backed securities are appropriate to except from Regulation M 
because they trade on the basis of their yield and creditworthiness, 
traditionally measured by credit ratings, and are largely fungible), 
the measure based on the Form SF-3 shelf eligibility requirements, 
which similarly focuses on the particular qualities of the asset-backed 
securities, is an appropriate substitute standard of creditworthiness 
to replace the reference to credit ratings in the existing Investment 
Grade Exception in accordance with section 939A(b)'s requirements, 
without having to restrict or place any further conditions on who may 
rely on the exception under Rule 102(d)(2)(ii), as amended.
    The Commission stated that it determined to include the Investment 
Grade Exception in Rule 102 based, in part, on the rationales indicated 
for an identical exception to Rule 101.\184\ As discussed above, in 
Part II.A.2, the Commission excepted investment grade asset-backed 
securities from Rule 101 because such securities trade primarily on the 
basis of yield and credit rating.\185\ When the Commission adopted the 
Investment Grade Exception in Rule 101, it stated that the principal 
focus of investors in the asset-backed securities market is on the 
structure of a class of securities and the nature of the assets pooled 
to serve as collateral for those securities rather than on the identity 
of a particular issuer.\186\ The Commission also stated that the 
Investment Grade Exception is for securities that are ``primarily 
serviced by the cashflows of a discrete pool of receivables or other 
financial assets, either fixed or revolving, that by their terms 
convert into cash within a finite time period plus any rights or other 
assets designed to assure the servicing or timely distribution of 
proceeds to the security holders.'' \187\
---------------------------------------------------------------------------

    \184\ Regulation M Adopting Release, 62 FR 531.
    \185\ See Regulation M Adopting Release, 62 FR 527.
    \186\ See Regulation M Adopting Release, 62 FR 527.
    \187\ See Regulation M Adopting Release, 62 FR 527 (citations 
omitted).
---------------------------------------------------------------------------

    The standard in new Rule 102(d)(2)(i) that relies on the Form SF-3 
eligibility requirements continues to be derived from the premise that 
certain asset-backed securities are traded based on factors such as 
their yield and creditworthiness.\188\ As discussed above, in Part 
II.A.2, the transaction requirements included in Form SF-3 allow for 
shelf offerings of only those asset-backed securities that share the 
qualities and characteristics of the investment grade asset-backed 
securities that meet the Investment Grade Exception: with respect to 
either set of securities, the principal focus of investors is the 
structure of a class of securities and the nature of the assets pooled 
to serve as collateral for those securities, rather than on the 
identity of a particular issuer.\189\
---------------------------------------------------------------------------

    \188\ See Regulation M Adopting Release, 62 FR 527.
    \189\ See supra note 149.
---------------------------------------------------------------------------

    The application of the transaction requirements included in the 
Commission's Form SF-3, therefore, should result in the offering of 
asset-backed securities that have similar qualities and characteristics 
to the asset-backed securities currently excepted under Rule 102's 
Investment Grade Exception. Because the ability of issuers, selling 
security holders, and their affiliated purchasers to bid up the price 
of an asset-backed security offered pursuant to an effective Form SF-3, 
during a distribution, is limited by a market participant's ability to 
substitute the security with other securities that are similar and of 
comparable creditworthiness,\190\ such a security is less susceptible 
to the types of manipulation that Regulation M seeks to prevent. In 
accordance with section 939A(b), it is appropriate to continue to 
except in Rule 102(d)(2) asset-backed securities that trade on the 
basis of their yield and creditworthiness.
---------------------------------------------------------------------------

    \190\ See Regulation M Adopting Release, 62 FR 527.
---------------------------------------------------------------------------

    For these reasons, the Commission is adopting in new Rule 
102(d)(2)(ii) an exception for asset-backed securities that are offered 
pursuant to an effective shelf registration statement filed on Form SF-
3. The new exception in Rule 102(d)(2)(i) may be relied upon by 
issuers, selling security holders, and their affiliated purchasers if 
all conditions of the exception are met.

C. Exchange Act Rule 17a-4(b)(17): Adding a Record Preservation 
Requirement for Broker-Dealers in Connection With Probability of 
Default Determinations

    The Commission proposed a new record preservation \191\ requirement 
that broker-dealers who are distribution participants or affiliated 
purchasers must preserve certain records pursuant to Rule 17a-4 under 
the Exchange Act, the Commission's broker-dealer record retention rule. 
Proposed paragraph (b)(17) of Rule 17a-4 would have required broker-
dealers relying on the exception for Nonconvertible Securities to 
preserve the written probability of default determination made pursuant 
to proposed paragraph (c)(2)(i) of Rule 101. Accordingly, those broker-
dealers would be required to preserve for a period of not less than 
three years, the first two years in an easily accessible place, the 
written probability of default determination made pursuant to proposed 
paragraph (c)(2)(i) of Rule 101.\192\
---------------------------------------------------------------------------

    \191\ In the Proposal, this requirement was described as a 
``recordkeeping'' requirement. For clarity and consistency with the 
title of Rule 17a-4 (``Records to be preserved by certain exchange 
members, brokers, and dealers''), this requirement is referred to 
throughout this release as a record preservation requirement.
    \192\ Proposal, 87 FR 18324.
---------------------------------------------------------------------------

    One commenter stated that the proposed record preservation 
requirement for certain broker-dealers is ``plainly appropriate as a 
means of facilitating the Commission in its examination and oversight 
of broker-dealers who rely on the exception in Rule 101 and would be 
required to

[[Page 39978]]

conduct the new probability of default determination.'' \193\
---------------------------------------------------------------------------

    \193\ Better Markets Letter, at 4.
---------------------------------------------------------------------------

    Another commenter stated that the Commission should not adopt the 
proposed record preservation requirement for broker-dealers relying on 
new Rule 101(c)(2)(i)'s exception for Nonconvertible Securities because 
``firms are already subject to extensive recordkeeping requirements 
[and] should continue to have flexibility in determining the precise 
nature and types of records they make and retain for such purpose, just 
as they do for purposes of the various other exceptions to Rule 101 of 
Regulation M.'' \194\
---------------------------------------------------------------------------

    \194\ SIFMA Letter 1, at 12; see also SIFMA Letter 2, at 2.
---------------------------------------------------------------------------

    However, unlike the other securities-based exceptions in Rule 101, 
which apply to ``actively-traded securities,'' \195\ ``exempted 
securities,'' \196\ and ``face-amount certificates or securities issued 
by an open-end management investment company or unit investment 
trust,'' \197\ and which are based on standards that rely on the use of 
publicly available information that can be verified, this exception is 
subject to the specific requirements in section 939A(b) to use 
``standards of credit-worthiness.'' As discussed above, in Parts II.A.1 
and B.1, the probability of default, as derived from structural credit 
risk models, is an appropriate substitute standard of creditworthiness 
to replace the reference to credit ratings in the existing Investment 
Grade Exceptions in accordance with section 939A(b)'s requirements. Due 
to the number of variations among structural credit risk models and 
their estimated inputs, the probability of default estimates may be 
subjective to some extent.\198\ As discussed below, in Part V.A.2, 
creditworthiness is an appropriate standard to reflect manipulation 
risk because securities issued by firms with sound creditworthiness 
trade primarily on yield and creditworthiness (traditionally measured 
by credit ratings) and have low pricing uncertainty and manipulation 
risk. Reliance on the new exception in Rule 101(c)(2)(i) or Rule 
102(d)(2)(i) for issuers of Nonconvertible Securities is conditioned on 
the use of a written probability of default calculation that has been 
determined and documented, in writing, by the distribution participant 
acting as the lead manager. This exception is in contrast to the other 
Regulation M exceptions that require the use of publicly available 
information that can be verified. Accordingly, requiring a record of 
the written probability of default determination to be preserved will 
help facilitate the Commission's examinations of broker-dealers relying 
on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i). The 
record preservation requirement, therefore, is appropriate to help 
deter improper adjusting of the estimation to meet the conditions of 
either of the exceptions. Further, because probability of default 
estimates may be subjective to some extent and not comparable across 
different issuers or for the same issuer across different issues if 
estimates are based on different models, or done by different 
researchers or vendors, the requirement associated with reliance on new 
Rule 101(c)(2)(i) to preserve written probability of default 
determinations is designed to facilitate the Commission's examinations 
of broker-dealers.
---------------------------------------------------------------------------

    \195\ 17 CFR 242.101(c)(1).
    \196\ 17 CFR 242.101(c)(3).
    \197\ 17 CFR 242.101(c)(4).
    \198\ See infra Part V.C.3; see also Proposal, 87 FR 18334.
---------------------------------------------------------------------------

    Another commenter stated that the Commission could publish, or 
require publication of, point estimates at a particular time and as 
rolling averages of default probabilities during a specified period, 
default probability estimates that market participants derive from 
various models, along with default probabilities implied by both market 
prices and credit default swap spreads (to the extent those are traded 
for a particular issuance).\199\ While access to such information could 
be informative for certain market participants and investors, the 
record preservation requirement set forth in new Rule 17a-4(b)(17), as 
stated in the Proposal, was designed to aid the Commission in its 
examinations of broker-dealers relying on the exception in Rule 
101(c)(2)(i), as amended, by requiring such broker-dealers to retain 
the written probability of default determination supporting their 
reliance on the exception.\200\ As such, a requirement for these 
entities to publish the information from the commenter's suggestion 
would not serve this purpose because such a requirement may not 
necessarily involve the type of information needed to meet the 
conditions new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i) \201\ and, 
therefore, would not facilitate the Commission's examinations of 
broker-dealers relying on those exceptions. In addition, the cost 
burden of doing so on a regular basis could be disproportionate to the 
infrequent usage of the exception, as these entities could incur other 
burdens associated with disclosing such information.\202\ Accordingly, 
it could discourage some entities from participating in certain issues, 
which could increase the costs of the affected issues.\203\ Similarly, 
for the Commission to publish this information, such that parties could 
rely on the information, would also not be appropriate because this 
approach would not facilitate its examinations of broker-dealers 
relying on the exception.
---------------------------------------------------------------------------

    \199\ IILF Letter, at 8.
    \200\ Proposal, 87 FR 18324.
    \201\ See, e.g., 17 CFR 242.101(c)(2)(i), as amended, 
242.102(d)(2)(i), as amended. Both of the new exceptions for 
Nonconvertible Securities in Rules 101 and 102 require that the 
issuer's probability of default be documented and determined, in 
writing, without necessarily requiring the other information 
included in the commenter's suggestion.
    \202\ See Proposal, 87 FR 18329; infra Part V.A.2.
    \203\ See, e.g., infra Part V.C.
---------------------------------------------------------------------------

    After reviewing the comments, the Commission is adopting Rule 17a-
4(b) under the Exchange Act largely as proposed, by adding new 
paragraph (17), which requires broker-dealers to preserve the written 
probability of default determination, relied upon pursuant to the new 
exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as applicable. As 
discussed above in Part II.B.1, the new exception in Rule 102(d)(2)(i) 
was not originally proposed. However, proposed Rule 17a-4(b)(17) was 
intended to capture ``broker-dealers relying on the exception for 
Nonconvertible Securities'' \204\ and, therefore, the Commission 
believes it is appropriate to apply it to both exceptions for 
Nonconvertible Securities that are being adopted. Moreover, broker-
dealers relying on either exception should be in a position to comply 
with the requirements of new Rule 17a-4(b)(17) because both exceptions 
require that the lead manager of a distribution make and document in 
writing the probability of default determination pursuant to Rule 
101(c)(2)(i), as amended.\205\ Accordingly, the final rule adds ``or 
Sec.  242.102(d)(2)(i) . . . or Rule 102 . . . as applicable'' in light 
of the addition of the new exception for Nonconvertible Securities in 
Rule 102(d)(2)(i) and that broker-dealers may be relying on the new 
exception either in Rule 101(c)(2)(i) or in Rule 102(d)(2)(i), 
depending on whether they are a covered person under Rule 101 or Rule 
102. In addition, the final rule adds the text ``, relied upon by such 
broker-dealer,'' to clarify that the written probability of default 
determination must be preserved in connection with a broker-dealer's

[[Page 39979]]

reliance on the new exception in Rule 101(c)(2)(i) or in Rule 
102(d)(2)(i), as applicable.
---------------------------------------------------------------------------

    \204\ Proposal, 87 FR 18324.
    \205\ New Rule 17a-4(b)(17) requires broker-dealers to preserve 
the written probability of default determination, relied upon by 
such broker-dealer, pursuant to Sec.  242.101(c)(2)(i) or Sec.  
242.102(d)(2)(i) (Rule 101 or Rule 102 of Regulation M), as 
applicable.
---------------------------------------------------------------------------

    New paragraph (b)(17) of Rule 17a-4 would affect the existing 
practices of broker-dealers by imposing new record preservation 
requirements when relying on the exception in new Rule 101(c)(2)(i) or 
new Rule 102(d)(2)(i). A broker-dealer who is a distribution 
participant acting as the lead manager (or in a similar capacity) of a 
distribution and uses a vendor to determine the probability of default 
could satisfy this record preservation requirement by maintaining 
documentation of the assumptions used in the vendor model, as well as 
the output provided by the vendor supporting the probability of default 
determination. Such a broker-dealer calculating the probability of 
default on its own could satisfy the record preservation requirement by 
maintaining documentation of the value of each variable in deriving the 
probability of default, along with a record identifying the specific 
source(s) of such information for each variable. Other broker-dealers, 
namely those that rely on the written probability of default 
determination of another broker-dealer acting as the lead manager (or 
in a similar capacity), could satisfy the record preservation 
requirement by maintaining a copy of the documentation described above, 
or by retaining a written notice it received of the probability of 
default determination.
    The requirement to preserve, pursuant to Rule 17a-4(b), the written 
probability of default determination is consistent with other record 
retention obligations that Exchange Act rules impose on broker-
dealers.\206\ Exchange members and broker-dealers currently are 
required to comply with the three-year preservation period in Rule 17a-
4(b) for other records and should have in place procedures to satisfy 
such preservation requirements.\207\
---------------------------------------------------------------------------

    \206\ See Proposal, 87 FR 18325.
    \207\ 17 CFR 240.17a-4(b).
---------------------------------------------------------------------------

III. Other Issues

    Certain commenters urged the Commission to take additional or 
different regulatory and non-regulatory actions than the approaches 
that were proposed, including actions that the Commission did not 
propose. These suggestions covered a variety of areas, including use of 
the term ``investors,'' \208\ SEC enforcement actions,\209\ other 
provisions of Regulation M,\210\ insurance company ratings,\211\ 
individual securities,\212\ credit ratings industry reforms,\213\ 
agency operations,\214\ and nondisclosures.\215\ These issues are 
outside the scope of the Proposal and that the final amendments to 
Rules 100(b), 101(c)(2), 102(d)(2), and 17a-4(b) appropriately further 
the Commission's objectives of promoting investor protection, enhancing 
market efficiency, and facilitating capital formation by implementing 
the requirements of section 939A(b) of the Dodd-Frank Act and 
facilitating the Commission during examinations of broker-dealers.
---------------------------------------------------------------------------

    \208\ Letter from Brian (Mar. 25, 2022).
    \209\ Letter from Patrick Lawson (Mar. 26, 2022).
    \210\ SIFMA Letter 2, at 4-5.
    \211\ Letter from Jason Wallace (May 19, 2022).
    \212\ See, e.g., Letter from Anthony Frattin (May 19, 2022); 
Kern Letter; Wang Letter; Ferguson Letter; Navari Letter.
    \213\ See Better Markets Letter, at 2.
    \214\ See Letter from Senator Thom Tillis to Vanessa Countryman, 
Sec'y, SEC (Nov. 4, 2022).
    \215\ Letter from The Delois Albert Brassell Estate and the 
Robert James Brassell Estate (June 1, 2022).
---------------------------------------------------------------------------

IV. Other Matters

    If any of the provisions of these rules, or the application thereof 
to any person or circumstance, is held to be invalid, such invalidity 
shall not affect other provisions or application of such provisions to 
other persons or circumstances that can be given effect without the 
invalid provision or application.
    Pursuant to the Congressional Review Act,\216\ the Office of 
Information and Regulatory Affairs has designated these rules as not a 
``major'' rule as defined by 5 U.S.C. 804(2).
---------------------------------------------------------------------------

    \216\ 5 U.S.C. 801 et seq.
---------------------------------------------------------------------------

V. Economic Analysis

    The Commission is sensitive to the economic consequences and 
effects, including costs and benefits, of its rules. Some of these 
costs and benefits stem from statutory mandates, while others are 
affected by the discretion exercised in implementing the mandates. 
Section 3(f) of the Exchange Act \217\ provides that whenever the 
Commission is engaged in rulemaking pursuant to the Exchange Act and is 
required to consider or determine whether an action is necessary or 
appropriate in the public interest, the Commission shall also consider, 
in addition to the protection of investors, whether the action will 
promote efficiency, competition, and capital formation. Additionally, 
section 23(a)(2) of the Exchange Act \218\ requires the Commission, 
when making rules under the Exchange Act, to consider the impact such 
rules would have on competition. Section 23(a)(2) also provides that 
the Commission shall not adopt any rule which would impose a burden on 
competition that is not necessary or appropriate in furtherance of the 
purposes of the Exchange Act.
---------------------------------------------------------------------------

    \217\ 15 U.S.C. 78c(f).
    \218\ 15 U.S.C. 78w(a)(2).
---------------------------------------------------------------------------

    The analysis below addresses the likely economic effects of the 
amendments, including the anticipated benefits and costs of the 
amendments, and their likely effects on efficiency, competition, and 
capital formation. The Commission also discusses the potential economic 
effects of certain alternatives to the approach taken by these 
amendments. Some of the benefits and costs discussed below are 
impracticable to quantify. For example, sticky offerings are generally 
not identified in the available data and may be difficult to trace in 
the appropriate records of the distribution participants. Therefore, 
much of the discussion of economic effects is qualitative.

A. Baseline

1. The Investment Grade Fixed Income Market
    To assess the economic effects of the amendments, the Commission is 
using as the baseline the nonconvertible debt, nonconvertible 
preferred, and asset-backed securities markets as they exist at the 
time of this release, including applicable rules that the Commission 
has already adopted.
    The affected parties include Nonconvertible Securities and asset-
backed securities (collectively ``fixed-income securities'') \219\ 
distribution and other market participants, such as issuers, selling 
security holders, underwriters, banks, broker-dealers, and their 
affiliated purchasers; fixed-income security investors, such as retail 
investors, mutual funds, exchange traded funds, and separate investment 
accounts; vendors of the relevant market data; and nationally 
recognized statistical rating organizations (``NRSROs''). Currently a 
majority of the distribution participants in the relevant markets are 
subscribed to a major vendor of the market data necessary to evaluate 
various aspects of the distribution. Further, a rating by an NRSRO is 
necessary in order for distribution participants to rely on the 
Investment Grade Exception. Today there are ten credit rating agencies 
registered with the Commission as

[[Page 39980]]

NRSROs.\220\ Three large NRSROs (S&P Global Ratings, Moody's Investors 
Service, Inc., and Fitch Ratings, Inc.) have historically accounted for 
most of the market share in this market. As of December 31, 2021, these 
three market participants accounted for 94.4% of all of the NRSRO 
credit ratings outstanding.\221\
---------------------------------------------------------------------------

    \219\ The term ``fixed-income securities'' in the Economic 
Analysis section refers to nonconvertible debt securities, 
nonconvertible preferred securities, and asset-backed securities.
    \220\ See Current NRSROs, U.S. Sec. & Exch. Comm'n, available at 
https://www.sec.gov/about/divisions-offices/office-credit-ratings/current-nrsros.
    \221\ See Staff Report on Nationally Recognized Statistical 
Rating Organizations (Feb. 2023) at 23, available at https://www.sec.gov/files/2023-ocr-staff-report.pdf.
---------------------------------------------------------------------------

    The affected securities are nonconvertible debt, nonconvertible 
preferred, and asset-backed securities. In 2021, there were 33,798 
issues of nonconvertible debt securities,\222\ with 687 issuers and 301 
agents involved (266 reported as participating underwriters, of which 
201 were the lead underwriters; 39--as trustees, and 10--as fiscal 
agents).\223\ Additionally, in 2021, there were 114 filed prospectuses 
for public offerings of asset-backed securities.\224\
---------------------------------------------------------------------------

    \222\ The nonconvertible debt securities also include preferred 
securities.
    \223\ The statistics are based on the data from Mergent. Some 
agents were reported as performing two or more functions, for 
example as an underwriter and as a lead underwriter.
    \224\ The information is based on EDGAR data for public 
offerings of asset-backed securities. It should be noted that 
prospectuses may contain multiple tranches, including non-offered 
tranches excluded from the public offering of asset-backed 
securities.
---------------------------------------------------------------------------

2. The Investment Grade Exception
    Regulation M is designed to prevent manipulative activities that 
could artificially influence the demand and pricing of covered 
securities.\225\ In particular, Rules 101 and 102 of Regulation M 
prohibit distribution and certain other market participants from 
bidding for or purchasing a covered security, in connection with a 
distribution of securities unless an exception, such as the Investment 
Grade Exception, applies.\226\ At the time the exception was included, 
the investment grade securities, that is securities characterized by 
sound creditworthiness, as measured by credit rating, were considered 
to be traded primarily on yield and credit ratings, and to be largely 
fungible.\227\ Therefore, sound creditworthiness was considered to be a 
good proxy for low manipulation risk. Investment Grade issues were 
presumed to have low probability of default and were thus considered to 
have low pricing uncertainty and low manipulation risk, which formed 
the basis for the exception. For purposes of these amendments sound 
creditworthiness is a good proxy for low manipulation risk since 
securities issued by firms with sound creditworthiness trade primarily 
on yield and creditworthiness (traditionally measured by credit 
ratings).\228\ Further, none of the commenters on the Proposal raised 
concern that creditworthiness would not be an adequate proxy for 
manipulation risk.
---------------------------------------------------------------------------

    \225\ See supra Part I.
    \226\ See 17 CFR 242.101(a), 242.102(a); see, e.g., 17 CFR 
242.101(c)(2), 242.102(d)(2).
    \227\ See Regulation M Adopting Release, 62 FR 527.
    \228\ There are other metrics that could serve as a proxy for 
manipulation risk in Rules 101 and 102, such as security public 
float or visibility to other market participants. One commenter for 
instance proposed Form S-3 and F-3 standard and a WKSI-based 
standard to measure manipulation risk (SIFMA Letter 1). However, 
unlike measures of creditworthiness, such criteria fail to capture 
the pricing point where the security is trading solely based on its 
yield and maturity and thus has low pricing uncertainty and low 
manipulation risk. Therefore, measures of creditworthiness are a 
better proxy for manipulation risk.
---------------------------------------------------------------------------

    The application of the Investment Grade Exception to Rules 101 and 
102 is primarily limited to two cases: re-openings (an offering of an 
additional principal amount of securities that are identical to the 
securities already outstanding, for example, when an issuer wishes to 
make a series of offerings via a re-opening to match its funding needs 
or when some foreign sovereign issuers conduct a re-opening for public 
finance purposes \229\) and sticky offerings (an offering where a lack 
of demand results in an underwriter being unable to sell all of the 
securities in a distribution, for example, when an investor failed to 
honor a previously expressed indication of interest; also, as stated in 
the Proposal, another example a commenter provided is in a best-efforts 
offering \230\).\231\ Re-openings are used infrequently and constitute 
about 0.3% of the relevant securities' markets' issuance volume.\232\ 
Sticky offerings are not identified in the relevant databases, making 
it difficult to assess their relative magnitude.
---------------------------------------------------------------------------

    \229\ See Proposal, 87 FR 18316. Note, however, that not every 
foreign sovereign issue is conducted in the form of re-opening.
    \230\ See Proposal, 87 FR 18316. In a `best-effort' offering, 
the underwriters are not required to sell any specific number or 
dollar amount of securities but will use their best efforts to sell 
the securities offered. See Plain English Disclosure, Release No. 
34-38164, (Jan. 14, 1997) [62 FR 3152 (Jan. 21, 1997)]. Note, 
however, that not every best-effort offering will become sticky, 
where the underwriter is unable to sell all of the securities in the 
distribution.
    \231\ See Proposal, 87 FR 18329. Note that the Commission 
received no comments on the types of issues that typically rely on 
the exception.
    \232\ The estimate is obtained using Mergent data for relevant 
securities during 2021.
---------------------------------------------------------------------------

    Re-openings are used in situations when such financing method 
offers the benefit of cost-effectiveness. For example, it may be 
cheaper for an issuer to offer a series of small offerings as opposed 
to one large offering, as the latter could result in a lower offering 
price due to the supply pressure. Further, since a re-opening issue is 
fungible with securities already in circulation and can be traded 
interchangeably with these securities in the secondary market, it 
provides additional liquidity benefits to the investors.\233\
---------------------------------------------------------------------------

    \233\ See Letter from Kenneth E. Bensten, Jr., Executive Vice 
President, Public Policy and Advocacy, SIFMA to Elizabeth M. Murphy, 
Secretary (July 5, 2011) at 6; John Berkery & Remmelt Reigersman, 
Re-openings: Issuing Additional Debt Securities of an Outstanding 
Series, Mayer Brown 1-2 (2020), available at https://www.mayerbrown.com/-/media/files/perspectives-events/publications/2020/05/reopenings_-issuing-additional-debt-securities-of-an-outstanding-series.pdf. See also Proposal, 87 FR 18329.
---------------------------------------------------------------------------

    As discussed above, sticky offerings typically result when a large 
investor fails to fulfill its expressed purchase interest in the issue, 
which could be due to a negative factor that transpired about the issue 
or issuer.\234\ Any offering of the relevant security thus can become a 
sticky offering. In such cases it may become challenging to trade the 
issue based solely on the yield and maturity (otherwise it would have 
become possible to find another purchaser in a timely manner). This may 
give rise in some cases to a heightened risk of manipulation in 
connection with a distribution of securities even if the security is 
rated as investment grade.\235\
---------------------------------------------------------------------------

    \234\ See Proposal, 87 FR 18329.
    \235\ We note, however, that not all sticky offerings are issued 
by an issuer with a low creditworthiness and have a high 
manipulation risk. It is thus important to have a standard of 
creditworthiness that is able to capture most recent available 
information on issuer creditworthiness, such as probability of 
default, and account for the cases of possible sudden declines in 
creditworthiness.
---------------------------------------------------------------------------

    Rule 102 provides that, in connection with a distribution of 
securities effected by or on behalf of an issuer or selling security 
holder, it shall be unlawful for such person, or any affiliated 
purchaser of such person, directly or indirectly, to bid for, purchase, 
or attempt to induce any person to bid for or purchase, a covered 
security during the applicable restricted period.\236\ Issuers and 
selling security holders generally do not have the same market access 
needs as underwriters and are not expected to buy the securities they 
are issuing. However, as pointed out by one of the commenters, their 
affiliated broker-dealers, which do not serve as an underwriter, may 
seek to rely on Rule 102 exception.\237\
---------------------------------------------------------------------------

    \236\ See supra Part I for a relevant discussion.
    \237\ SIFMA Letter 1, at 12-13.

---------------------------------------------------------------------------

[[Page 39981]]

    The Investment Grade Exception was included in Regulation M as it 
was considered a good proxy for the likelihood of manipulation 
risk.\238\ However, the reference to credit ratings in the Commission's 
rules may encourage investors to place undue reliance on the credit 
ratings. Credit ratings themselves are potentially imprecise and often 
lagging indicators of creditworthiness.\239\
---------------------------------------------------------------------------

    \238\ See supra Part I.
    \239\ We note that the SEC staff took a similar position in the 
COVID-19 Market Monitoring Group, Credit Ratings, Procyclicality and 
Related Financial Stability Issues: Select Observations, SEC Staff 
(July 15, 2020) (``Cost of debt capital is driven by a wide range of 
financial and non-financial factors and forces; ratings downgrades 
are generally lagging indicators of cost of debt capital.''), 
available at https://www.sec.gov/news/public-statement/covid-19-monitoring-group-2020-07-15.
---------------------------------------------------------------------------

B. Benefits of the Amendments

    As mentioned above, section 939A(b) of the Dodd-Frank Act requires 
the Commission to ``remove any reference to or requirement of reliance 
on credit ratings, and to substitute in such regulations such standard 
of creditworthiness as the Commission determines to be appropriate.'' 
\240\ In this amendment, the Commission will require distribution 
participants, issuers, selling security holders and affiliated 
purchasers, in order to avail themselves of these exceptions from 
Regulation M, to rely upon the structural credit risk models as a 
measure of creditworthiness.\241\ These models have been used to 
estimate the probability of default of an issuer.\242\
---------------------------------------------------------------------------

    \240\ Public Law 111-203, sec. 939A(a). The Commission has 
issued several releases concerning the removal of references to 
credit ratings: Security Ratings, Release No. 34-64975 (July 27, 
2011) [76 FR 46603 (Aug. 3, 2011)]; Removal of Certain References to 
Credit Ratings Under the Securities Exchange Act of 1934, Release 
No. 34-71194 (Dec. 27, 2013) [79 FR 1522 (Jan. 8, 2014)]; Removal of 
Certain References to Credit Ratings under the Investment Company 
Act, Release No. IC-30847 (Dec. 27, 2013) [79 FR 1316 (Jan. 8, 
2014)]; Asset-Backed Securities Disclosure and Registration, Release 
No. 34-72982 (Sept. 4, 2014) [79 FR 57184 (Sept. 24, 2014)]; Removal 
of Certain References to Credit Ratings and Amendment to the Issuer 
Diversification Requirement in the Money Market Fund Rule, Release 
No. IC-31828 (Sept. 16, 2015) [80 FR 58124 (Sept. 25, 2015)].
    \241\ See, e.g., the seminal model by Robert C. Merton, On the 
Pricing of Corporate Debt: The Risk Structure of Interest Rates, 29 
Journal of Finance 449, 449-70 (1974), along with related successive 
refinement models such as Fischer Black & John C. Cox, Valuing 
Corporate Securities: Some Effects of Bond Indenture Provisions, 31 
J. Fin. 351, 351-67 (1976); Robert Geske, The Valuation of Corporate 
Liabilities as Compound Options, 12 J. Fin. & Quantitative Analysis 
541, 541-52 (1977); and Oldrich A. Vasicek, Credit Valuation, KMV 
(Mar. 22, 1984), among others.
    \242\ For example, the Merton (1974) Model and the Successor 
Models are included in the curriculum for such credentials as the 
Chartered Financial Analyst. See, e.g., Credit Analysis Models, >CFA 
Inst. (2022), available at https://www.cfainstitute.org/en/membership/professional-development/refresher-readings/credit-analysis-models. One commenter, however, suggested that ``most of 
our member firms do not use them [the credit risk models] for other 
purposes either, to the extent such models are used at all, they 
serve merely as a supplement to member firms' own proprietary credit 
analysis as part of their decision making on whether to extend a 
loan or other credit.'' (SIFMA Letter 1 at 5). See also Part II.A.1 
for a relevant discussion.
---------------------------------------------------------------------------

    Structural credit risk models typically take the issuer balance 
sheet measures of debt obligations as given and estimate a probability 
of default based on the market value and volatility of the firm's 
equity. The value of equity is viewed in these models as the value of a 
call option on firm assets where the strike price is the total notional 
value of debt. Since the market value of equity, the volatility of 
equity, and the notional value of debt can be calculated from the 
market trading and balance sheet data, under the structural credit risk 
models the volatility of the value of the assets and the market value 
of assets, which are not observable, can be estimated. The probability 
of default can be calculated as the probability that the call option 
will expire out-of-the-money, which occurs when the value of the 
company falls below the book value of the debt.
    As discussed above, structural credit risk models are based on the 
structure of the balance sheet.\243\ Since the future value of the firm 
is unknown, a structural credit risk model must make assumptions about 
the probability distribution of possible firm values in different 
scenarios, some of which may trigger default. These assumptions include 
the current firm value and the volatility of firm value, for which the 
observed market value of equity and the volatility of equity is often 
an input. Some models include assumptions over the firm's dividend 
policy.
---------------------------------------------------------------------------

    \243\ An alternative set of models used to derive probability of 
default are `reduced-form models'. The reduced-form models rely on 
statistical analysis rather than the balance sheet to determine a 
firm's creditworthiness. However, compared to structural credit risk 
models, they lack in rigorous theoretical justification as well as 
economic interpretation of the resulted relationships between the 
model inputs. See, e.g., Edward Altman, Andrea Resti, & Andrea 
Sironi, Default Recovery Rates in Credit Risk Modeling: A Review of 
the Literature and Empirical Evidence, 33 Econ. Notes 183 (2004) 
(discussing the competing models), available at https://onlinelibrary.wiley.com/doi/10.1111/j.0391-5026.2004.00129.x.
---------------------------------------------------------------------------

    For purposes of these amendments, the probability of default 
derived from the structural credit risk models is an appropriate proxy 
for creditworthiness. As discussed previously in Part V.A.2, 
creditworthiness is an appropriate standard to reflect manipulation 
risk since securities issued by firms with sound creditworthiness trade 
primarily on yield and creditworthiness (traditionally measured by 
credit ratings) and have low pricing uncertainty and manipulation risk. 
The Commission received several comments supporting the probability of 
default as a standard for Rules 101 and 102 exception.\244\ However, 
one commenter opposed this option and suggested a standard based on 
Forms S-3 and F-3 or on WKSI standard.\245\ However, these alternatives 
are not good measures of sound creditworthiness as compared to 
probability of default because they fail to reflect the pricing point 
where a security is traded solely on its yield and maturity. Thus, the 
probability of default based on structural credit risk models is a more 
appropriate proxy for creditworthiness, and thereby for manipulation 
risk.
---------------------------------------------------------------------------

    \244\ See, e.g., IILF Letter, Bloomberg L.P. Letter, and Better 
Markets Letter.
    \245\ See SIFMA Letters 1 and 2 and the relevant discussion in 
Part II.A.1.
---------------------------------------------------------------------------

    Consistent with the Proposal, the Commission is adopting a 0.055% 
probability of default threshold. The Commission requested and received 
comments on this proposed threshold level. Specifically, one commenter 
expressed support of the proposed threshold and also noted that at any 
given date, the composition and population of any selected sample 
meeting the threshold could change; \246\ as such some variation of the 
estimated percentages of the captured universe of securities eligible 
for the existing Investment Grade Exception is to be expected. Another 
commenter expressed that the threshold should be increased to 0.5% 
because it believes the exceptions as amended should be crafted to 
capture as many of the securities covered under the existing investment 
grade exceptions as possible; this commenter did not address the 
corresponding increase in the percentage of currently ineligible 
securities or the costs of that increase.\247\ No other commenters 
suggested a different or lower threshold and, overall, the commenters 
did not identify any economic effects of the proposed threshold level 
that were not considered in the Proposal.
---------------------------------------------------------------------------

    \246\ See Bloomberg L.P. Letter, at 2.
    \247\ See SIFMA Letter, at 10.
---------------------------------------------------------------------------

    The Commission acknowledges that the percentage of investment grade 
securities that would be captured under a specific threshold fluctuates 
over time and as conditions change that affects the various inputs into 
the models. As of March 2023, the 0.055% probability of default 
threshold captured

[[Page 39982]]

approximately 76% of the investment grade securities in the final 
sample of nonconvertible Fixed-Income Securities used (1996 distinct 
investment grade issues with probability of default below 0.055% out of 
2637 total investment grade rated issues in the sample).\248\ This 
threshold also captured approximately 24% of non-investment grade 
issues (64 out of 269 non-investment grade issues in the sample).
---------------------------------------------------------------------------

    \248\ The investment grade status for nonconvertible securities 
issued between 2018 and 2023 was obtained from Mergent (as of the 
last available Mergent update through Mar. 2023) while the 
probability of default estimates were obtained for a cross-section 
of securities available in Bloomberg (as of Mar. 28, 2023). Please 
refer to Mario Bondioli, Martin Goldberg, Nan Hu, Chengrui Li, Olfa 
Maalaoui Chun, & Harvey J. Stein, The Bloomberg Corporate Default 
Risk Model (DRSK) for Public Firms (working paper Aug. 28, 2021), 
available at https://ssrn.com/abstract=3911300 (retrieved from SSRN 
Elsevier database), for methodology description of Bloomberg 
probability of default measure.
---------------------------------------------------------------------------

    This estimation differs from that in the Proposal. In the Proposal, 
we observed, using data from October 2021, that the 0.055% threshold 
captured about 90% of investment grade securities (2436 out of 2710 
issues) and about 37% of non-investment grade issues (125 of 341 non-
investment grade issues).\249\ Overall, at the time of the analysis of 
data as of March 2023, 2060 issues met the proposed exception as 
compared with the 2637 issues under the current exception.
---------------------------------------------------------------------------

    \249\ See Proposal, 87 FR 18330.
---------------------------------------------------------------------------

    Given the reaction of commenters to the proposed 0.055% threshold 
\250\ and that there is an unavoidable trade-off between capturing 
securities that are ineligible for the existing Investment Grade 
Exception and leaving out some securities that are currently eligible, 
the proposed threshold is intended to strike a reasonable balance 
between these two statistical realities over time.\251\
---------------------------------------------------------------------------

    \250\ As discussed above, one commenter expressed general 
support of the proposed 0.055% threshold (see Bloomberg L.P. Letter, 
at 2) while another commenter suggested increasing the level to 0.5% 
(See SIFMA Letter 1, at 10).
    \251\ As pointed out by one commenter, some variation of the 
estimates is unavoidable, and ``this highlights the importance of 
selecting an objective, data driven model that is consistently 
applied over time and documented by the distribution participant.'' 
See Bloomberg Letter, at 2.
---------------------------------------------------------------------------

    Nonconvertible debt securities and nonconvertible preferred 
securities of issuers for which the probability of default, estimated 
as of the sixth business day immediately preceding the determination of 
the offering price and over the horizon of 12 full calendar months from 
such day, is 0.055% or less, as determined and documented, in writing, 
by the distribution participant acting as the lead manager (or in a 
similar capacity) of a distribution, as derived from a structural 
credit risk model are to be excepted from Rules 101 and 102.
    An advantage of using probabilities of default implied by 
structural credit risk models instead of NRSRO credit ratings is that 
these model-implied probabilities of default generally use current 
estimates of equity valuation and volatility based on the recent 
trading activity, and hence incorporate more recent news affecting the 
valuation and perceived volatility of the firm. In contrast, credit 
rating agencies are generally slower than the market in updating credit 
ratings and outlooks and thus may reflect less up-to-date 
information.\252\
---------------------------------------------------------------------------

    \252\ We note that the SEC staff took a similar position in the 
COVID-19 Market Monitoring Group, Credit Ratings, Procyclicality and 
Related Financial Stability Issues: Select Observations, SEC Staff 
(July 15, 2020) (``Cost of debt capital is driven by a wide range of 
financial and non-financial factors and forces; ratings downgrades 
are generally lagging indicators of cost of debt capital.''), 
available at https://www.sec.gov/news/public-statement/covid-19-monitoring-group-2020-07-15. Some academic studies find evidence 
that structural credit risk models may be able to respond to 
aggregate and firm specific news faster than credit ratings. Also, 
such models are able pick up on differences in default risk within a 
credit rating bucket. However, credit ratings do not necessarily 
imply probabilities of default and thus may not be directly 
comparable to probability of default estimated using a structural 
credit risk model. See Jing-zhi Huang & Hao Zhou, Specification 
Analysis of Structural Credit Risk Models (Fed. Res. Bd., Fin. & 
Econ. Discussion Series, 2008-552008), available at https://www.federalreserve.gov/pubs/feds/2008/200855/200855pap.pdf; Moody's 
Analytics, EDF Overview (2011) (outlining the approach by Moody's 
KMV), available at https://www.moodysanalytics.com/-/media/products/EDF-Expected-Default-Frequency-Overview.pdf; Giuseppe Montesi & 
Giovanni Papiro, Risk Analysis Probability of Default: A Stochastic 
Simulation Model, 10 J. Credit Risk 29 (2014).
---------------------------------------------------------------------------

    The Proposal did not limit which distribution participants are 
allowed to produce probability of default estimations for the purposes 
of the exception. In order to ensure consistency and reliability of the 
estimates within any particular distribution and reduce the potential 
subjectivity and non-uniformity of the estimates the amendments specify 
that only lead managers are responsible for estimating the probability 
of default for a given distribution.\253\ Lead managers would have 
flexibility of either calculating the probability of default internally 
using structural credit risk models, given the wide availability of 
software products available on the market that perform such 
calculations, or obtaining an estimate from a vendor. One of the 
benefits of the amendment is that the lead managers will have the 
flexibility of selecting the model they find most appropriate to assess 
the creditworthiness of issuers for the purposes of using the 
exception.\254\ This means the lead managers will not have to rely on a 
credit rating for the issue in order to determine its eligibility for 
Rules 101 and 102 exception and will no longer have to rely on an 
NRSRO's choice of the model for such purposes.\255\ Furthermore, 
multiple vendors currently provide estimates of the probability of 
default based upon structural credit risk models as a part of default 
packages that include various market data and metrics.\256\
---------------------------------------------------------------------------

    \253\ Some of the costs associated with this option are 
discussed in the Costs Section of the Economic Analysis.
    \254\ However, this will not be the case for other distribution 
participants who are not considered the lead manager of the 
distribution, which may deter such participants from relying on the 
exception. Further, this may result in lead managers' selecting a 
model that allows them to rely on the exception but is not 
necessarily the best model of the securities' creditworthiness and 
manipulation risk. These issues are discussed in more detail in 
infra Part V.C.
    \255\ Even though the lead manager would have to use a 
structural credit risk model, there are many versions of such models 
available, and the specific model parameters can be selected as 
well, providing considerable flexibility of the estimates as 
compared to the specific choices used in the assessments by NRSROs.
    \256\ Vendors offer a number of commercial applications based on 
structural credit risk models. The probability of default calculated 
by structural credit risk models, such as the Merton (1974) Model 
and the Successor Models, can also be calculated by lead managers 
without the use of a vendor. One commenter, however, suggested that 
currently firms seldom use probability of default models in 
connection with issuances of the relevant securities. See SIFMA 
Letter 1 at 5.
---------------------------------------------------------------------------

    Removing and replacing the references to credit ratings from Rules 
101 and 102 of Regulation M may also have a benefit of expanding the 
number of options available to lead managers compared to what they 
would have under the requirements of the Investment Grade Exception. 
Specifically, the exceptions' requirement will no longer rely on a 
limited number of vendors providing credit ratings, which may reduce 
possible negative consequences from limited competition. Structural 
credit risk models as a measure for creditworthiness could therefore 
serve as a better proxy for manipulation risk than credit ratings 
because, by prescribing a methodology rather than a metric generated by 
only a certain category of regulated vendors (that is, NRSROs), 
distribution lead managers may have more options for either using a 
vendor-supplied structural credit risk model or using their own 
proprietary version of a publicly available structural credit risk 
model.
    Under the final rule amendments, the structural credit risk models 
cannot be applied to asset-backed securities due to

[[Page 39983]]

the complexity of the structure of such instruments.\257\ Even though 
one commenter suggested that probability of default can be estimated 
for asset-backed securities \258\ such estimation based on structural 
credit risk models is not routinely used due to the complexity of the 
structure of these securities and the corresponding complex application 
of such models. Further, another commenter supported proposed Form SF-3 
standard for the Investment Grade Exception with respect to asset-
backed securities.\259\ The final amendments provide that securities 
that are offered pursuant to an effective shelf registration statement 
filed on Form SF-3 should also be excepted from Rules 101 and 102. The 
Form SF-3 shelf eligibility requirements provide objective criteria 
that can also ensure that the securities are consistent with the 
Commission's original basis for the Investment Grade Exceptions. Asset-
backed securities that are offered pursuant to an effective shelf 
registration statement filed on Form SF-3 are less at risk of the 
manipulation that Regulation M addresses. Specifically, the Form SF-3 
shelf eligibility requirements limit the number of nonperforming assets 
in the asset-backed security pool, require review of the pool assets if 
certain conditions are met, and require certification by the chief 
executive officer, among other things.
---------------------------------------------------------------------------

    \257\ See a relevant discussion in supra Part II.B.2.
    \258\ See IILF Letter, at 6.
    \259\ SIFMA Letter 1, at 11.
---------------------------------------------------------------------------

    As the Commission noted when adopting Form SF-3, the Form 
incentivizes sponsors to carefully review and disclose the underlying 
assets' characteristics, reducing the overall uncertainty about the 
asset-backed security \260\ and, with respect to these final 
amendments, the risk of manipulation. The Commission received no 
comments that suggest otherwise. Asset-backed securities that are 
offered pursuant to an effective shelf registration statement filed on 
Form SF-3 have similar qualities and characteristics to the investment-
grade asset-backed securities currently excepted in Rule 
101(c)(2).\261\ A review of recent EDGAR database filings confirms that 
almost all asset-backed securities issued pursuant to an effective 
shelf registration statement filed on Form SF-3 have investment grade 
ratings.\262\
---------------------------------------------------------------------------

    \260\ See supra notes 121-125 and accompanying text.
    \261\ One commenter opposed use of SF-3 standard for asset-
backed securities and suggested relying on the probability of 
default instead (IILF Letter, at 7). However, probability of default 
calculations based on a structural credit risk model are complex for 
this type of securities due to their complex structure and are not 
routinely used. Another commenter in fact expressed support of using 
SF-3 standard for asset-backed securities (SIFMA Letter 1, at 11).
    \262\ Based on EDGAR database filings from 2022.
---------------------------------------------------------------------------

C. Costs of the Amendments

    The Commission recognizes that some of the affected underwriters, 
their affiliates, as well as issuers, selling security holders and 
affiliated purchasers may bear costs from the amendments. The 
amendments may alter the universe of securities that are eligible for 
the new exceptions. If some distribution participants decide not to 
participate in certain issues because of the rule amendments, the costs 
of the affected issues may increase. For example, when fewer banks or 
broker-dealers are available, the underwriters may be able to charge 
higher fees. Additionally, as the result of the amendments, fewer 
issues may take place or issuers may rely more on private markets,\263\ 
potentially limiting issuers' ability to raise capital and affecting 
investors in the relevant securities as the available security 
selection and liquidity may be reduced.
---------------------------------------------------------------------------

    \263\ See SIFMA letter 1, at 5.
---------------------------------------------------------------------------

    There are several types of costs that could arise: (1) costs 
associated with calculations or obtaining the probability of default 
estimate; (2) costs associated with preserving records related to the 
probability of default estimation; (3) costs due to the probability of 
default being an imperfect proxy for creditworthiness, (4) asset-backed 
securities' costs associated with the amendments, (5) indirect and 
other costs of the amendments. We discuss these costs in detail below.
1. Costs Associated With Obtaining the Estimate of the Probability of 
Default
    Lead managers may incur costs related to determining the 
probability of default. Consistent with the Paperwork Reduction Act 
(``PRA'') section,\264\ the Commission estimates that it will take a 
lead manager 3 hours to establish a system to gather the data serving 
as the inputs and then perform the analysis necessary to calculate the 
probability of default of the issuer whose securities are the subject 
of the distribution, for an aggregate cost of $218,889 \265\ Consistent 
with the PRA section,\266\ the Commission also estimates that it will 
take a lead manager one hour to gather the inputs required to calculate 
probability of default each time it participates in a distribution of 
Nonconvertible Securities. There were 33,798 offerings of 
Nonconvertible Securities in 2021. Therefore, it is estimated that 
annually lead managers will spend maximum of $12,268,674 \267\ in the 
aggregate complying with this requirement if all lead managers choose 
to estimate the probability of default internally.
---------------------------------------------------------------------------

    \264\ See infra Part VI.
    \265\ The Commission estimates the wage rate based on salary 
information for the securities industry compiled by SIFMA. See 
Management & Professional Earnings in the Securities Industry--2013, 
SIFMA (Oct. 7, 2013). These estimates are modified by the Commission 
staff to account for an 1800 hour work-year and multiplied by 5.35 
(professionals) or 2.93 (office) to account for bonuses, firm size, 
employee benefits and overhead. These figures have been adjusted for 
inflation through Jan. 2023 using data published by the Bureau of 
Labor Statistics' Consumer Price Index inflation calculator, 
available at https://www.bls.gov/data/inflation_calculator.htm. 201 
lead managers x 3 hours x $363 hour for a compliance manager = 
$218,889.
    \266\ See infra Part VI.C.1.
    \267\ Cost estimated is based on the sum of 33,798 offerings 
multiplied by 1 burden hour multiplied by $363, for a compliance 
manager. See Management & Professional Earnings in the Securities 
Industry--2013, SIFMA (Oct. 7, 2013). These estimates are modified 
by the Commission staff to account for an 1800 hour work-year and 
multiplied by 5.35 (professionals) or 2.93 (office) to account for 
bonuses, firm size, employee benefits and overhead. These figures 
have been adjusted for inflation through Jan. 2023 using data 
published by the Bureau of Labor Statistics' Consumer Price Index 
inflation calculator, available at https://www.bls.gov/data/inflation_calculator.htm.
---------------------------------------------------------------------------

    However, some lead managers may rely on third party vendors rather 
than internally calculate the probability of default. Any costs 
associated with using a vendor to obtain probability of default 
estimate, however, should be small, as the vendors typically already 
have subscriptions available to provide calculations regarding the 
probability of default based on structural credit risk models.\268\ 
Furthermore, lead managers, in particular those that choose to 
determine the probability of default estimate internally, are likely to 
already have the computational resources necessary to conduct such 
analysis internally. Therefore, the total costs for the lead managers 
of complying with the requirement should be below $12,268,674.
---------------------------------------------------------------------------

    \268\ See infra note 256. One commenter suggested that firms 
rarely use probability of default models in connection with 
issuances of the relevant securities. However, probability of 
default estimates are typically provided by the vendors in a package 
with other data firms are often subscribed to.
---------------------------------------------------------------------------

    Further, since the rule amendments specify that only the lead 
manager can supply the estimate of the probability of default for the 
purposes of relying on the exception, some issues where there is no 
distribution participant to act as the lead manager for the 
distribution, such as with self-underwritten offerings, at-the-market 
offerings, or other shelf offerings, may not be able to rely on the

[[Page 39984]]

exception. These issues may therefore be subject to Regulation M 
restrictions and may have to rely on private markets and may face 
potentially higher issuing costs or not take place.\269\
---------------------------------------------------------------------------

    \269\ Such costs, however, cannot be quantified due to lack of 
available data.
---------------------------------------------------------------------------

2. Costs Associated With Maintaining Records Related to the Probability 
of Default Estimation
    Broker-dealers relying on the new exception in Rule 101(c)(2)(i) or 
Rule 102(d)(2)(i) must preserve the written probability of default 
determination made pursuant to Rule 101(c)(2)(i), as amended. 
Consistent with the PRA section,\270\ the Commission estimates that it 
will take a distribution participant 25 hours to update the applicable 
policies and systems required to account for capturing the records made 
pursuant to new Rule 101(c)(2)(i), for an aggregate cost of 
$2,731,575.\271\ Consistent with the PRA section,\272\ the Commission 
also estimates that it will take a distribution participant 10 hours to 
maintain such records as well as to make additional updates to the 
applicable record preservation policies and systems to account for the 
rules. Therefore, it is estimated that annually broker-dealers will 
spend $1,092,630 \273\ in the aggregate complying with this 
requirement.
---------------------------------------------------------------------------

    \270\ See supra Part VI.C.2.
    \271\ 301 distribution participants x 25 hours x $363 hour for a 
compliance manager = $2,731,575. See Management & Professional 
Earnings in the Securities Industry--2013, SIFMA (Oct. 7, 2013). 
These estimates are modified by the Commission staff to account for 
an 1800 hour work-year and multiplied by 5.35 (professionals) or 
2.93 (office) to account for bonuses, firm size, employee benefits 
and overhead. These figures have been adjusted for inflation through 
Jan. 2023 using data published by the Bureau of Labor Statistics' 
Consumer Price Index inflation calculator, available at https://www.bls.gov/data/inflation_calculator.htm.
    \272\ See supra Part VI.C.2.
    \273\ Cost estimated based on the sum of 301 distribution 
participants multiplied by 10 burden hours multiplied by $363, for a 
compliance manager. See Management & Professional Earnings in the 
Securities Industry--2013, SIFMA (Oct. 7, 2013). These estimates are 
modified by the Commission staff to account for an 1800-hour work-
year and multiplied by 5.35 (professionals) or 2.93 (office) to 
account for bonuses, firm size, employee benefits and overhead. 
These figures have been adjusted for inflation through Jan. 2023 
using data published by the Bureau of Labor Statistics' Consumer 
Price Index inflation calculator, available at https://www.bls.gov/data/inflation_calculator.htm.
---------------------------------------------------------------------------

3. Costs Associated With Structural Credit Risk Model Based Probability 
of Default Being an Imperfect Proxy for Creditworthiness
    As discussed previously, the structural credit risk models are 
designed to measure creditworthiness, and creditworthiness itself is 
considered a good measure of manipulation risk. There are costs that 
are currently present in the relevant markets associated with 
creditworthiness being an imperfect proxy for manipulation risk. 
However, in the absence of a better proxy for manipulation risk, 
creditworthiness has continued to successfully serve the purpose of 
measuring such risk for many years. This is also supported by the 
comments stating that the investment grade standard has been 
successfully used in Rules 101 and 102 exception.\274\ The final rule 
amendments are not expected to alter those costs and the discussion 
that follows focuses instead on the costs associated with the 
structural credit risk models as a proxy for creditworthiness.
---------------------------------------------------------------------------

    \274\ See, e.g., Rothwell, at 2 and ABA Letter, at 15 -17.
---------------------------------------------------------------------------

    The use of any model to estimate creditworthiness necessarily 
provides an imperfect measure. Structural credit risk models are no 
exception. We note, however, that models such as structural credit risk 
models often are a part of the analysis involved in obtaining a credit 
rating.\275\
---------------------------------------------------------------------------

    \275\ See, e.g., John Y. Campbell, Jens Hilscher, & Jan 
Szilagyi, In Search of Distress Risk, 63 J. Fin. 2899 (2008), 
available at https://scholar.harvard.edu/files/campbell/files/campbellhilscherszilagyi_jf2008.pdf.
---------------------------------------------------------------------------

    Some ways to implement structural credit risk models make use of 
historical trading data to produce a reliable estimate of the model 
input parameters. These data may not be available for certain 
infrequently traded securities. In some circumstances, the market for a 
security has not yet been established and sufficient trading data are 
unavailable, making it difficult to apply the exception.
    Additionally, structural credit risk models rely on a number of 
parameter estimates such as firm market value and volatility, which 
could be difficult to assess as these values change with market 
conditions and business fluctuations. A changing term structure of 
interest rates and noise trading in the market can further distort the 
probability of default estimates. Incorrect parameter estimates may 
result in the incorrect estimates of default probability and allow 
distribution participants to rely on the exception for risky issues or 
prevent distribution participants from relying on the exception for 
safe issues. Implied probabilities of default are sensitive to market 
prices and estimates of market volatility and consequently tend to be 
counter cyclical, increasing during market downturns, which are often 
also periods of increased uncertainty. A constant threshold which is 
not time-varying will potentially result in fewer firms qualifying for 
the exception during market downturns, which may result in more 
issuances during this period not qualifying or firms choosing not to 
issue, hence increasing their cost of capital or limiting their access 
to capital.
    While credit rating downgrades are also countercyclical occurring 
more frequently during market downturns, they tend to be slow in 
incorporating updates.\276\ Thus, the impact of the counter cyclicality 
of default probabilities implied by structural credit risk models could 
be stronger relative to using credit ratings: during periods of 
distress, using these probabilities of default will likely result in 
fewer firms with an investment grade credit rating falling below the 
threshold, and thus fewer firms qualifying for the exception relative 
to using credit ratings. Lead managers who make probability of default 
determinations pursuant to new Rule 101(c)(2)(i) could make reasonable 
adjustments to model parameters and inputs to recalculate the 
probability of default as market conditions change, mitigating the 
costs discussed above.
---------------------------------------------------------------------------

    \276\ We note that the SEC staff took a similar position the 
COVID-19 Market Monitoring Group, Credit Ratings, Procyclicality and 
Related Financial Stability Issues: Select Observations, SEC (July 
15, 2020) (``Cost of debt capital is driven by a wide range of 
financial and non-financial factors and forces; ratings downgrades 
are generally lagging indicators of cost of debt capital.''), 
available at https://www.sec.gov/news/public-statement/covid-19-monitoring-group-2020-07-15.
---------------------------------------------------------------------------

    Due to the number of variations among structural credit risk models 
and their estimated inputs, the probability of default estimates may be 
subjective to some extent and not comparable across different issuers 
or for the same issuer across different issues if estimates are based 
on different models, or done by different researchers or vendors. The 
latter may affect market participants' ability to effectively rely on 
the estimates to make comparative assessments across multiple 
securities. However, this is also true of the credit ratings that often 
rely on similar models, which mitigates these costs of the amendments 
relative to the market baseline.
    Further, as a result of the Rules 101 and 102 amendments, all 
underwriters as well as issuers, selling security holders and 
affiliated purchasers will rely on the lead manager's assessment of the 
probability of default in order to use the exception.\277\ This should 
mitigate the subjectivity and non-uniformity of the estimation concerns 
for the same

[[Page 39985]]

issue and to some degree across issues for the same issuer to the 
extent the same parties are engaged by the issuer for different issues. 
This requirement allows the lead manager to perform estimations which 
determine if the resulted probability of default falls below the 
threshold for all the distribution participants and their affiliates 
and thus the availability of the exception. Some of these participants 
may decide to withdraw if the exception is not available. However, the 
lead manager is interested in the best outcome of the distribution and 
therefore has strong incentives to encourage the participation of these 
entities in the distribution, mitigating the above concern. This may, 
on the other hand, incentivize lead managers to select models and 
estimation specifics in such a way to ensure the resulted estimates are 
below the threshold, potentially allowing issues of issuers with low 
creditworthiness and high manipulation risk to rely on the exception. 
The public availability of alternative probability of default estimates 
available for the investors through multiple vendors, however, should 
mitigate this concern.
---------------------------------------------------------------------------

    \277\ See supra Part II.B.1.
---------------------------------------------------------------------------

    In addition, as discussed above in reference to the selected 
threshold, the proposed amendment may expand the universe of issuers of 
nonconvertible securities that qualify for the exception and include 
issuers that did not receive an investment grade credit rating, but 
have a structural credit model implied probability of default that 
falls below the threshold. The debt prices of these firms may be prone 
to manipulation if the price of their debt is relatively more sensitive 
to the idiosyncratic risks of the issuers.
    Additionally, this amendment may create potential opportunities for 
new products offered by the vendors designed specifically for a given 
issue or issuer. A custom designed estimate paid for by a party with an 
interest in the outcome of the distribution may lead to potential 
conflicts of interest since the vendor is incentivized in this case to 
produce an estimate which will allow the issuer, their affiliates and 
selling security holders, and other distribution participants to rely 
on the exception. However, the existing major vendors supplying 
probability of default estimates have numerous clients currently using 
this information for business purposes other than the Rules 101 and 102 
exception. Therefore, given the reputational concerns it is unlikely 
that these vendors will produce a product to cater specifically to the 
use of these estimates for purposes of relying on the Rules 101 and 102 
exception.
    Additionally, the model input estimates or assumptions may be 
selected by the lead manager in such a way as to produce the desired 
estimation result if the model is estimated internally and may result 
in lead managers' selecting the models so as to be able to rely on the 
exception.\278\ This may result in an additional cost of adding some 
manipulation risk to the relevant markets if manipulation prone issues 
are allowed to rely on the exception as a result.
---------------------------------------------------------------------------

    \278\ The definition of structural credit risk models for 
purposes of new Rule 101(c)(2)(i) is limited to commercially or 
publicly available models, which would limit a distribution 
participant's ability to develop its own models to achieve favorable 
results.
---------------------------------------------------------------------------

    Finally, the threshold of 0.055% for the exception is based on 
model assumptions and available data. Some commenters expressed support 
for the proposed threshold level selection,\279\ while one commenter 
suggested a higher level.\280\ Future market evolution may result in 
this threshold becoming either too large or too small, allowing risky 
issues to rely on the exception or preventing less risky issues from 
using it. One commenter expressed a similar concern about a set-level 
threshold specification in the rules.\281\ The threshold may vary by 
industry, with the threshold being more restrictive in some industries 
relative to the original NRSRO investment grade designation. Moreover, 
probabilities of default as implied by structural credit risk models 
tend to be counter-cyclical and can spike in periods of crisis due to 
decreases in market valuation and increases in equity volatility. 
Consequently, during such periods, fewer investment grade firms 
generally fall below the threshold. Credit ratings by NRSROs are also 
countercyclical but tend to be slow-moving, since credit rating changes 
often lag updates to firm conditions that will impact cost of 
capital.\282\
---------------------------------------------------------------------------

    \279\ See Bloomberg L.P. Letter. at 2, which provides analysis 
supporting the proposed probability of default threshold. 
Additionally, IILF Letter, at 6 suggests that the proposed threshold 
is in a reasonable range.
    \280\ See SIFMA Letter 1 at 10.
    \281\ See IILF Letter, at 6.
    \282\ We note that the SEC staff took a similar position in the 
COVID-19 Market Monitoring Group, Credit Ratings, Procyclicality and 
Related Financial Stability Issues: Select Observations, SEC Staff 
(July 15, 2020) (``Cost of debt capital is driven by a wide range of 
financial and non-financial factors and forces; ratings downgrades 
are generally lagging indicators of cost of debt capital.''), 
available at https://www.sec.gov/news/public-statement/covid-19-monitoring-group-2020-07-15.
---------------------------------------------------------------------------

4. Costs Associated With Asset-Backed Securities' Amendments
    The amendments may render some asset-backed securities ineligible 
to rely on the exception from the Regulation M. This may increase 
issuance costs for the underwriters as well as issuers, selling 
security holders and affiliated purchasers. For instance, broker-
dealers may reduce an offering's size or increase fees if the exception 
to Regulation M is no longer available.\283\ Additionally, issuers may 
need to establish new business relationships due to Regulation M 
restrictions. Furthermore, some issuers may decide not to issue the 
affected securities if the exceptions to Regulation M are no longer 
available. As a result, some asset-backed securities' issues may not 
take place, which could affect issuers' ability to raise capital and 
could affect investors in the relevant markets by potentially reducing 
the selection of the available asset-backed securities.
---------------------------------------------------------------------------

    \283\ Such changes in fees or changes in size cannot be 
reasonably quantified due to lack of available data on the 
respective changes (before and after an occurrence) in the relevant 
values.
---------------------------------------------------------------------------

5. Indirect and Other Costs of the Amendments
    Besides the direct effects on the distribution participants and 
affected securities discussed above the final rule amendments may also 
generate indirect effects including on investors in these securities 
and NRSROs. For instance, distribution participants other than lead 
managers may want to verify the estimates provided by the lead manager 
by either obtaining the estimate from a vendor or making the 
calculations internally, which will result in additional costs for 
these participants.\284\
---------------------------------------------------------------------------

    \284\ These costs are estimated as $363 per participant per 
distribution if estimates are obtained internally. Consistent with 
the PRA, the Commission estimates that it would take one hour per 
issue to calculate probability of default. Cost estimated is based 
on 1 burden hour multiplied by $363, for a compliance manager. See 
Management & Professional Earnings in the Securities Industry--2013, 
SIFMA (Oct. 7, 2013). These estimates are modified by the Commission 
staff to account for an 1800 hour work-year and multiplied by 5.35 
(professionals) or 2.93 (office) to account for bonuses, firm size, 
employee benefits and overhead. These figures have been adjusted for 
inflation through Jan. 2023 using data published by the Bureau of 
Labor Statistics' Consumer Price Index inflation calculator, 
available at https://www.bls.gov/data/inflation_calculator.htm.
---------------------------------------------------------------------------

    Additionally, the lead managers, although not required, may need to 
expend resources in terms of their staff time and resources in order to 
notify other distribution participants, their affiliated purchasers, 
issuers, selling security holders, and their affiliated purchasers of 
their probability of default

[[Page 39986]]

determinations that were estimated pursuant to Rule 101(c)(2)(i).
    The Commission estimates that it will take 0.25 hours per lead 
manager per issue (8,450 hours annually) \285\ to notify other 
distribution participants of the probability of default estimates. 
Therefore, the total estimated cost for the lead managers associated 
with notifying other distribution participants is estimated as 
$1,732,250.\286\
---------------------------------------------------------------------------

    \285\ 33,798 issues times 0.25 hours.
    \286\ 8,450 hours * $205 hour for a junior business analyst wage 
= $1,732,250 See Management & Professional Earnings in the 
Securities Industry--2013, SIFMA (Oct. 7, 2013). These estimates are 
modified by the Commission staff to account for an 1800 hour work-
year and multiplied by 5.35 (professionals) or 2.93 (office) to 
account for bonuses, firm size, employee benefits and overhead. 
These figures have been adjusted for inflation through January 2023 
using data published by the Bureau of Labor Statistics' Consumer 
Price Index inflation calculator, available at https://www.bls.gov/data/inflation_calculator.htm.
---------------------------------------------------------------------------

    Further, if issuer participation in the relevant security issues, 
for example in the case of re-openings or issues that are more likely 
to become sticky offerings, becomes limited, some issues may not take 
place that otherwise would. Investors may additionally face a more 
limited choice of investment instruments as a result. This may also 
affect liquidity of their portfolios in the case of re-openings, since 
re-openings can offer additional liquidity benefits as the securities 
offered in re-openings are interchangeable with the existing issues. 
However, as already discussed in the case of re-openings, these costs 
are expected to be minimal as re-openings are used infrequently.
    The rule amendments do not rely on an NRSRO rating in order to 
determine if an issue is eligible for the exception. This may diminish 
NRSROs' clientele to the extent NRSROs choose not to provide structural 
credit risk model-based estimates of the probability of default for 
their existing clients opting to rely on the exception. However, the 
amendment may increase the clientele of the vendors that supply 
relevant data and metrics to the lead managers or other distribution 
participants who wish to verify the lead manager estimates, if such 
vendors already supply probability of default estimates or choose to 
offer this estimate as a part of their services. In addition, if firms 
do not solicit credit rating services from NRSROs beyond the estimate 
of a probability of default implied by a structural credit risk model, 
investors will not be able to benefit from the information provided by 
a credit rating report and ongoing coverage of the firm that otherwise 
will be provided through the distribution participant.

D. Efficiency, Competition, and Capital Formation

    As discussed previously, lead managers will have flexibility in 
selecting the structural credit risk model to access creditworthiness 
as a measure of manipulation risk for the business. This may encourage 
issuers to issue securities in relevant markets, as well as 
participation of other distribution participants, such as selling 
security holders and affiliated purchasers. As a result, this could 
improve competition between issuers for investors as well as 
competition between lead managers for underwriting business.
    Further, widely available estimates of the probability of default 
as well as an option of internal model estimation could lead to a more 
competitive environment in the provision of models as the requirement 
to rely on proprietary credit risk models of a small number of NRSROs 
is removed. The improved competition, market participation and 
efficiency ultimately should lead to more efficient capital formation 
as the access to and functioning of the relevant fixed income markets 
improves.
    However, it is possible that a new business model could emerge in 
the relevant markets that leads to conflicts of interest and 
neutralizes the effects discussed above. For instance, lead managers 
could contract with a vendor or a credit rating agency directly to 
create a custom estimate of the probability of default. This could 
result in a business model where an interested party pays for the 
supplied estimate and where vendors may be incentivized to produce an 
estimate designed to fit the desired estimation result. Thus issuers 
that otherwise will not be able to rely on the exception could end up 
being excepted potentially increasing the manipulation risk in the 
relevant markets, which in turn could negatively affect competition and 
capital formation. The reputational concerns, however, would generally 
prevent vendors from generating estimates specifically designed for the 
needs of a small number or a single customer.
    Additionally, the positive effects discussed above could be offset 
by the fact that only lead managers can obtain an estimate of the 
probability of default for the distribution. Some issues where there is 
no distribution participant to act as the lead manager for the 
distribution, such as with self-underwritten offerings, at-the-market 
offerings, or other shelf offerings'' may not have the exceptions 
available.\287\ This may deter participants from such distributions and 
in some cases result in securities being issued in private markets or 
issues not taking place. This may negatively affect the competition and 
capital formation in the relevant market.
---------------------------------------------------------------------------

    \287\ See supra Part V.C.1.
---------------------------------------------------------------------------

    Some issuers may also face higher costs or no longer be able to use 
the exception, for example, due to imperfect model estimates because of 
market fluctuations or changing market. High costs of issuance or 
inability to rely on the exception may deter participants from issuing 
the affected securities, which could affect competition and capital 
formation in the relevant markets. Further, potential negative effects 
of non-uniform estimates and subjectivity additionally reduce these 
benefits. As discussed previously, variations in model assumptions, 
parameters, or data sample used necessarily introduce an element of 
subjectivity in the final estimates and leads to differences in the 
estimates across different issues or issuers. Finally, potentially 
increased issuance costs due to some asset-backed securities being 
ineligible for the exception may also negatively affect market 
participation and competition of the relevant markets.

E. Reasonable Alternatives

    Alternative 1 discussed below deals with the probability of default 
threshold, alternatives 2-4 discuss alternative approaches to using 
structural credit risk models as a standard of creditworthiness to 
measure manipulation risk. Alternative 5 discusses elimination of the 
exception from Rule 101, alternative 6 deals with asset-backed 
securities, alternative 7 discusses Rule 102 options, while the last 
alternative discussed the record preservation requirement.
1. Alternative Threshold for Probability of Default
    The Probability of Default threshold of 0.055% was chosen in an 
effort to maximize investment grade securities captured and minimize 
the non-investment grade securities captured. However, a different 
threshold could be used in the Rule exception, which would capture 
different proportions of investment and non-investment grade 
securities. For example, based on data as of March 2023, a higher 
threshold of 0.5% is estimated to capture about 97% of investment grade 
securities (2550 out of 2637 investment grade issues) and about 70% of 
non-investment grade issues (188 out of 269 non-investment grade 
issues). A lower threshold of 0.03% is estimated to capture about 64% 
of investment grade securities (1675 out of 2637 investment grade

[[Page 39987]]

issues) and 11% of non-investment grade issues (29 out of 269 non-
investment grade issues).
    The advantage of a higher threshold is that it captures a larger 
set of investment grade securities, but at the expense of also 
capturing an additional set of non-investment grade securities, which 
could be prone to manipulation risk. Increasing the threshold would 
allow more investment grade securities to rely on the exception at 
expense of a potentially higher manipulation risk; on the other hand, 
decreasing the threshold would limit the ability of some of the 
investment grade securities to use the exception, but would also limit 
the number of non-investment grade securities allowed to rely on the 
exception and, as a result, also limit manipulation risk.
    The Commission proposed 0.055% threshold level, which scoped in 
about 90% of investment grade issues and about 37% of non-investment 
grade issues.\288\ One of the commenters suggested increasing the 
threshold in order to capture a larger percentage of the previously 
eligible investment grade issues,\289\ another commenter suggested that 
the proposed threshold level is appropriate,\290\ while none of the 
commenters suggested decreasing the threshold. Furthermore, at any 
given date, the proportion of currently eligible securities that would 
be captured varies. Manipulation risk remains the primary concern of 
Regulation M. Because the originally proposed threshold of 0.055% 
remains appropriate for these purposes, and acknowledging the variation 
in eligible securities that would be captured over time, increasing (or 
decreasing) this threshold for the primary aim of capturing more (or 
fewer) of currently eligible securities does not justify changing this 
threshold.
---------------------------------------------------------------------------

    \288\ Based on the data as of Oct. 2021. Based on Mar. 2023 
data, 0.055% threshold scopes in about 76% of investment grade 
issues (1996 out of 2637 issues) and about 24% or non-investment 
grade issues (64 out of 269 issues).
    \289\ SIFMA Letter 1, at 7 and 10, see also a relevant 
discussion in Part II.B.1.
    \290\ See Bloomberg L.P. Letter, at 2.
---------------------------------------------------------------------------

    Rather than providing a specific number as a threshold, a method 
for distribution participants to use in calculating such a threshold 
could be specified instead. For example, such method could involve 
calculating a set of probability of default estimates for a sample of 
Nonconvertible Securities with characteristics such as yield and 
maturity similar to the distribution participant's securities issued 
over a specified time interval and comparing it to a specified standard 
of creditworthiness. A longer time interval of the data sample would 
capture more issues and improve statistical accuracy at expense of 
having market conditions potentially changing and generating incorrect 
estimates. A shorter time interval of the sample ensures the market 
conditions have not changed but includes fewer issues resulting in a 
smaller sample and lower statistical accuracy. One of the commenters 
expressed similar ideas advocating for an estimation method rather a 
fixed threshold level, which would result in a more flexible threshold 
level.\291\
---------------------------------------------------------------------------

    \291\ See IILF Letter, at 7.
---------------------------------------------------------------------------

    The main advantage of specifying a method as opposed to a number 
for the threshold is its flexibility with respect to changing market 
conditions. The main disadvantage of this alternative is subjectivity 
of the analysis involved, which may lead to non-uniform application of 
the Regulation M exceptions across issues or issuers if the estimated 
threshold differs considerably across issues or issuers; or incentivize 
market participants to adjust the threshold estimation to be able to 
rely on the exception. Some commenters expressed a concern for the 
estimates' subjectivity and non-uniformity as discussed previously. 
This alternative could introduce additional subjectivity and non-
uniformity and thus is sub-par to the originally proposed option.\292\
---------------------------------------------------------------------------

    \292\ See SIFMA Letter 1, at 5, Better Markets Letter, at 4, and 
the relevant discussion in Part II.D.
---------------------------------------------------------------------------

2. Exception Based on Security Characteristics
    As an alternative replacement for the reference to investment grade 
securities, the Commission considered analysis that could be based on 
security characteristics, such as (1) total amount of issue outstanding 
(public float); (2) yield to maturity of the security during a past 
trading period; or (3) empirical duration.\293\ Other relevant security 
characteristics that could be used are outlined in the 2011 
Proposal.\294\ Such analysis could be performed internally or 
externally and could be additionally verified by a third party. All of 
these alternatives were included in the Proposal and the Commission 
received no comments in regards to these alternatives. Below we discuss 
public float, yield to maturity and empirical duration criteria in more 
detail.
---------------------------------------------------------------------------

    \293\ Empirical duration is bond duration calculated based on 
historical data rather than a formula. Typically, it is estimated 
using a regression analysis of the relationship between market bond 
prices and Treasury yields.
    \294\ 2011 Proposing Release, 76 FR 26557-64.
---------------------------------------------------------------------------

     Exception Based on the Total Amount of Issue Outstanding 
(Public Float).
    To the extent that it is more difficult to manipulate price of a 
larger issue, public float could be used as an alternative criterion to 
reflect manipulation risk. This criterion has the advantage of being 
straightforward and easy to evaluate. Due to its simplicity, it lacks 
the estimation issues associated with other measures such as the 
probability of default. However, determination of a threshold for 
public float to select securities for the exception is complicated due 
to its considerable variation across issuers or industries. A specific 
threshold selection could potentially disadvantage smaller issuers--
especially during periods of market downturns when valuations are low. 
Additionally, public float is not inherently an indication of low 
credit risk since a distressed firm can have a large amount of debt.
     Exception Based on Yield to Maturity.
    Securities that are traded primarily on yield and maturity have low 
manipulation risk, as discussed before, since their pricing does not 
reflect issuer specific risks. Yield to maturity, therefore, can be 
used as an alternative criterion to evaluate manipulation risk. 
However, using yield to maturity as a criterion for securities eligible 
for the exception is also problematic. Even though this criterion is 
similarly easy to obtain and lacks any major estimation issues, 
selecting a threshold is not straightforward. For instance, yield to 
maturity differs considerably by industry. Selecting a fixed threshold 
may result in some industries being under-represented and others over-
represented in the pool of eligible issues. Moreover, yield to maturity 
often moves with risk-free rates; thus fewer firms would be excepted 
during periods of high interest rates. The default-free component of 
yield to maturity makes this measure a very noisy proxy of credit 
worthiness.
     Exception Based on Empirical Duration.
    Empirical duration is another alternative proxy that could be used 
to evaluate Nonconvertible Securities for an exception from Regulation 
M. Negative empirical duration might be an indication that a 
Nonconvertible Security or its issuer is of low creditworthiness. A 
Nonconvertible Security with negative empirical duration is less 
affected by changes in interest rates than Nonconvertible Securities of 
creditworthy issuers and trades similar to equity securities. Although 
negative empirical duration may demonstrate that a particular issuer or 
security is not creditworthy, it has some limitations that affect the 
viability

[[Page 39988]]

of negative empirical duration as a substitute for the reference to 
credit ratings in the Investment Grade Exception. In particular, this 
measure relies heavily on statistical analysis, requires the 
Nonconvertible Security to be traded, and may lack intuitive 
interpretation, which renders empirical duration a poor proxy for the 
type of manipulation that Regulation M is designed to prevent.
3. Exception Based on Issuer Characteristics
    The Commission also considered an exception based on issuer 
characteristics, for example, the interest coverage ratio, the WKSI 
standard, as suggested in the 2008 Proposing Release,\295\ a Form S-3/
F-3-based standard, or a criterion based on a reduced-form credit risk 
model, as an alternative to the structural credit risk models. We 
discuss these alternatives below.
---------------------------------------------------------------------------

    \295\ 2008 Proposing Release, 73 FR 40095-97.
---------------------------------------------------------------------------

     Exception Based on the WKSI Standard.
    The Commission could adopt a standard based on the amount of the 
issuer's total securities outstanding or based on the WKSI standard as 
a criterion to determine eligibility for the exception. The issuers 
that fall under the WKSI definition or with sufficient amounts of total 
securities issued or outstanding are large and established firms that 
typically have sound creditworthiness. The Commission included this 
alternative in the Proposal. One commenter expressed some support for 
this alternative.\296\ The advantage of this characteristic is its 
simplicity, uniformity, and the lack of subjectivity of the analysis. 
However, the WKSI standard as discussed in the 2008 Proposing Release, 
for example, was heavily criticized for allowing risky high-yield 
issues to be eligible for the exception and preventing issues by 
smaller but otherwise creditworthy issuers from relying on the 
exception, which remains a considerable concern.\297\ Even though one 
of the commenters suggested a standard based on the WKSI standard due 
to its simplicity, uniformity and lack of subjectivity,\298\ such a 
standard would fail to capture the pricing point where securities trade 
solely based on their yields and maturity and not on the issuer 
characteristics, where pricing uncertainty and manipulation risk are at 
their minimum. Thus, such a standard would be a sub-par measure of 
manipulation risk as compared to the probability of default.
---------------------------------------------------------------------------

    \296\ See SIFMA Letter 1 at 9.
    \297\ ABA Letter, at 15-17 and Letter from Deborah A. Cunningham 
and Boyce I. Greer, Co-chairs, Securities Industry and Financial 
Markets Association (``SIFMA'') Credit Rating Agency Task Force, to 
Florence E. Harmon, Acting Secretary (Sep. 4, 2008) at 13.
    \298\ See SIFMA Letter 1 at 9.
---------------------------------------------------------------------------

     Exception Based on Forms S-3 and F-3.
    One commenter stated that the complexity of the proposed 
probability of default calculations would impose additional regulatory 
burdens that could be avoided if the exception, instead, relied on a 
standard based on readily verifiable and publicly available 
information.\299\ This commenter proposed using Form S-3 or Form F-3 as 
a standard for the exception given the uniformity, simplicity and a 
lack of subjectivity of such a standard.\300\ The Form S-3 or Form F-3 
eligibility criteria are intended to access whether an issuer is widely 
followed,\301\ rather than an issuer's creditworthiness. A widely 
followed issuer may be more likely to have a low manipulations risk, 
making this a reasonable alternative criterion to consider for the 
Investment Grade Exception. However, such a standard does not 
differentiate securities that are traded solely on their yield and 
creditworthiness from securities that trade also on the issuer identity 
and thus have a high manipulation risk. Therefore, probability of 
default is a preferred standard to rely upon in the assessment of 
manipulation risk for the purposes of the Investment Grade Exception.
---------------------------------------------------------------------------

    \299\ See SIFMA Letter 1 at 5.
    \300\ See SIFMA Letter 1, at 5-8 as well as the related 
discussion in Part II.B.1.
    \301\ Form S-3 and Form F-3 Release, 76 FR 46607.
---------------------------------------------------------------------------

     Exception Based on the Interest Coverage Ratio.
    Another possible issuer-based criterion for exception eligibility 
is the interest coverage ratio. This alternative was included in the 
Proposing Release and no commenters expressed a view on this option. A 
high interest coverage ratio typically indicates the issuer's ability 
to repay debt and can be used as a criterion to reflect 
creditworthiness. It has the advantage of being a simple and easy to 
calculate value. However, the interest coverage ratio is an accounting 
measure that can result in inconsistent outcomes as it is based on the 
reported earnings rather than cash flows. Reported earnings may differ 
based on accounting practices of the firm. Structural credit risk 
models have an advantage over interest coverage ratio since they are 
not dependent on reported earnings, which are heavily influenced by 
accounting practices.
     Exception Based on Reduced-Form Credit Risk Model.
    An alternative to using structural credit risk models is reduced-
form credit risk models.\302\ The latter models could be a good measure 
of creditworthiness and of manipulation risk to the extent that 
creditworthiness is a good proxy for manipulation risk. This 
alternative was discussed in the Proposal. One of the commenters 
proposed a similar alternative relying on debt security prices, yields, 
or credit spreads instead of using a structural credit risk model for 
the probability of default estimation.\303\ Unlike structural models, 
reduced-form models do not assume default occurs when firm value falls 
below a threshold. The default is instead assumed to follow an 
unobserved process and the default model can be fitted to the market 
data. The advantage of these models is they do away with some of the 
unrealistic requirements of structural credit risk models, for example 
when the firm value, its volatility or other required parameters are 
unobserved.
---------------------------------------------------------------------------

    \302\ The reduced-form credit risk models are discussed, for 
example, in Robert Litterman & Thomas Iben, Corporate Bond Valuation 
and the Term Structure of Credit Spreads, 17 (3) Fin. Analysts J. 
52, 52-64 (1991); Robert A. Jarrow & Stuart M. Turnbull, Pricing 
Derivatives on Financial Securities Subject to Default Risk, 50 J. 
Fin. 53, 53-86 (1995); Robert A. Jarrow, David Lando, & Stuart M. 
Turnbull, A Markov Model for the Term Structure of Credit Risk 
Spreads, 10 Rev. Fin. Stud. 481, 481-523 (1997); Darrell Duffie & 
Kenneth J. Singleton, Modeling the Term Structures of Defaultable 
Bonds, 12 Rev. Fin. Stud. 687, 687-720 (1999).
    \303\ IILF Letter, at 2.
---------------------------------------------------------------------------

    Even though such models can be considered more flexible and may 
provide better fit for the observed default events, their ability to 
predict future defaults may not necessarily exceed that of the 
structural models. In addition, unlike structural models, they suffer 
from a lack of theoretical background of the assumed relationships, or 
the intuitive interpretation of the model dependencies and why the 
defaults occur. Unrestricted use of these models might also provide 
more opportunity to choose a reduced-form model specification which 
enables use of the exception. Further, some commenters expressed a 
concern for a lack of consistency and uniformity across issues or 
issuers in using probability of default standard for the 
exception.\304\ Since reduced-form models are more flexible and less 
structured than structural credit risk models, such concerns would be 
more pronounced in

[[Page 39989]]

a standard that is based on the reduced-form models.
---------------------------------------------------------------------------

    \304\ See SIFMA Letter 1, at 5, Better Markets Letter, at 4.
---------------------------------------------------------------------------

4. Exception Based on Issuer and Issue Characteristics
    The Commission considered, as another alternative, an analysis 
based on both security and issuer characteristics; for example, 
characteristics outlined in Exchange Act Rule 15c3-1. This alternative 
was discussed in the Proposal and the Commission received no comments 
in regard to this option. Rule 15c3-1 specifies a set of factors to 
determine a minimum amount of credit risk broker-dealers can use to 
determine if a security can qualify for lower haircuts: (1) credit 
spreads; (2) securities-related research; (3) internal or external 
credit assessments; (4) default statistics; (5) inclusion in an index; 
(6) enhancements and priorities; (7) price, yield and/or volume; or (8) 
asset-class specific factors.\305\ Some of these factors, such as 
default statistics or credit assessments, measure issuer 
creditworthiness, while others, such as price, yield, or volume, 
measure the manipulation risk present in each specific issue, providing 
a good overall assessment of manipulation risk.
---------------------------------------------------------------------------

    \305\ See Removal of Certain References to Credit Ratings Under 
the Securities Exchange Act of 1934, Release No. 34-71194 (Dec. 27, 
2013) [79 FR 1522, 1527-28 (Jan. 8, 2014)].
---------------------------------------------------------------------------

    The advantage of this alternative is that it would align the 
exception with already existing standards that broker-dealers might 
apply to determine whether a security has a minimal amount of credit 
risk. The standard in Rule 15c3-1 was adopted in 2013 as a replacement 
for a reference to investment grade securities pursuant to section 939A 
of the Dodd-Frank Act. Such test could have minimum additional costs 
for broker-dealers who already have all the necessary procedures in 
place for its application.
    The Rule 15c3-1 standard is commonly used for seasoned securities 
and, therefore, includes a longer time period to obtain information 
about issues that may not be available for the new issuances and for 
seasoned (actively traded) distributions that may have only a one-day 
restricted period also subject to Regulation M. Moreover, the Rule 
15c3-1's minimal credit risk standard is based on a set of eight 
different factors, some of which include price or volume, with respect 
to each specific issue. Depending on these other participants' systems 
and regulatory obligations, it may be costly for them to replace the 
investment grade standard with the minimal credit risk standard. This 
could result in a situation where different distribution participants 
are facing different costs,\306\ possibly deterring some market 
participants.
---------------------------------------------------------------------------

    \306\ This is unlike the structural credit risk model based 
probability of default that would imply the same costs for all the 
participants who obtain the estimated values.
---------------------------------------------------------------------------

5. Elimination of the Investment Grade Exception From Rule 101
    The Commission also considered eliminating the Investment Grade 
Exception for Fixed-Income Securities from Rule 101. Elimination of the 
exception was discussed as an alternative in the Proposal and the 
Commission did not receive any direct comments on this option. However, 
as discussed in Part II, commenters broadly supported the Commission's 
efforts to find an alternative standard of creditworthiness in place of 
the references to credit ratings in Rule 101's Investment Grade 
Exception (as opposed to removing the Investment Grade Exception, 
without a replacement).\307\ The advantage of this alternative is 
eliminating the situations when manipulation-prone securities fall 
under the exception due to limitations of proxies used to select the 
securities to be excepted. For instance, as discussed above, there are 
various limitations of the structural credit risk models' applications, 
which may limit the ability of certain issuers to rely on the exception 
or allow issuers with a higher risk of having their securities 
manipulated to avoid Regulation M. If the exception is eliminated, any 
limitations of such a proxy for manipulation risk are eliminated as 
well. In addition, this approach could ultimately relieve lead managers 
from the need to spend time or costs to implement, understand, and 
calibrate any standard such as a structural credit risk model.
---------------------------------------------------------------------------

    \307\ See, e.g., SIFMA Letter 1, at 2; Bloomberg L.P. Letter, at 
1.
---------------------------------------------------------------------------

    However, this approach raises a number of concerns. Specifically, 
eliminating the exception could make some offerings in the excepted 
securities considerably more costly. For example, with respect to re-
openings, broker-dealers who might otherwise elect to re-open a bond 
offering may determine not to do so to avoid restrictions of Regulation 
M that could arise during such a re-opening if it becomes a sticky 
offering. This could increase the cost of the issue that has to rely on 
the next-best alternative structure. Further, an alternative 
transaction structure, if selected, may decrease the liquidity of the 
securities being issued because they would not be fungible with the 
previously issued securities. This may also result in some distribution 
participants, such as broker-dealers, deciding not to participate. This 
could limit the number of available broker-dealers, potentially 
increasing fees faced by the issuers. Further, if certain issues do not 
take place under the amendments, it could reduce the selection of 
available securities for the investors in the relevant markets and may 
limit issuers' ability to raise capital.
    However, these costs might be mitigated because a party subject to 
the prohibitions of Rule 101 could structure its buying activity before 
or after the applicable restricted period so as not to incur any costs 
associated with relying on the exception.
    The above arguments apply to all currently excepted investment 
grade securities because any such issue can become a sticky offering 
and the underwriters have to account and adjust for this possibility 
ex-ante. In a scenario where an underwriter is unable to sell its 
allotted securities to the public on or promptly after the pricing 
date, there is no exception on which to rely, the underwriter/broker-
dealer would likely ex-ante adjust the cost of issuance to reflect this 
added risk. Broker-dealers could be more cautious in structuring 
potentially sticky offerings if they know they will be required to 
comply with Rule 101 (and have no exceptions available), by reducing an 
offering's size or increasing fees as a risk premium. This could 
potentially raise the cost of investment grade offerings. However, this 
could also decrease the probability of an offering to become sticky, 
potentially reducing manipulation risk in the relevant markets.
    The removal, without replacement, of the Investment Grade Exception 
could also affect the liquidity of the Fixed-Income Securities if re-
openings of issues already in circulation are more costly, potentially 
reducing issuers' reliance on this financing structure, which 
negatively affects the investors in the relevant markets.
    This alternative could also disrupt some established business 
relationships. In certain circumstances new relationships may need to 
be established. For example, if an offering becomes sticky, absent 
Investment Grade Exception to rely on some broker-dealers may be 
limited in their ability to trade relevant securities and decide to 
withdraw, in which case the issuer may need to seek a different broker-
dealer. This would increase costs of the affected security offerings, 
including the new broker dealer fees or the search costs, especially 
when the market has a limited number of available broker-dealers.

[[Page 39990]]

6. Alternative for Asset-Backed Securities
    As an alternative for asset-backed securities the Commission 
considered using a standard based on the value at risk. This 
alternative was included in the Proposal and no commenter expressed any 
view on this standard. Value at risk measures the percentage loss of 
the security in the worst case scenarios over a specified time period. 
It can be estimated by performing a simulation over the underlying 
securities' pool and determining the cash flows available to the asset-
backed security in each scenario. A number of commercially available 
options can be used to perform this analysis. Value at risk can be a 
good indicator of manipulation risk since low value at risk indicates 
that the majority of the cash flows are sufficiently assured. The price 
of the asset-backed security in this case is more certain and is less 
subject to manipulation risk.
    However, value at risk is by construction estimated for a specified 
time period and thus only accounts for the potential losses during such 
period, while losses may also occur after this time period. In this 
case the price of the asset-backed security may depend on issue-
specific factors and be prone to manipulation despite the estimated 
value at risk over the specified time period being low. This may allow 
securities with high manipulation risk to rely on the exception.
    One of the commenters proposed as an alternative to use probability 
of default-based standard for asset-backed securities calculated using 
prices and credit spreads.\308\ However, probability of default is 
typically not used for these securities due to the complexity of their 
structure and corresponding complexity of the calculations. Further, 
another commenter supported proposed Form SF-3 standard to use for the 
asset-backed securities due to its uniformity and simplicity.\309\
---------------------------------------------------------------------------

    \308\ See IILF Letter, at 6.
    \309\ See SIFMA Letter 1, at 11.
---------------------------------------------------------------------------

7. Alternatives for Rule 102 Exception
    The Commission also considered and proposed eliminating, without 
replacing, the Investment Grade Exception in Rule 102. Disruption to 
the trading market may be limited because distribution participants 
will still be able to rely on the exception from Rule 101 if they meet 
the requirements of the proposed rules. However, one of the commenters 
pointed out that eliminating the exception from Rule 102 may affect 
issuer-affiliated broker-dealers that do not act as an underwriter and 
may need to rely on the Rule 102 exception.\310\ Eliminating the 
exception from Rule 102 may increase issuance costs or deter market 
participants from issuing such securities. Therefore, elimination of 
the exception from Rule 102 was not the best option in comparison to 
the alternative selected.
---------------------------------------------------------------------------

    \310\ See SIFMA Letter 1, at 12.
---------------------------------------------------------------------------

8. Alternative for the Record Preservation Requirement
    The Commission considered not adding the record preservation 
requirement. The option of not adding the record preservation 
requirement for broker-dealers was suggested by one of the commenters 
due to the additional burdens it creates for the broker-dealers.\311\ 
However, the record preservation requirement may help ensure an 
estimate of the probability of default is produced for all the 
distribution participants to rely upon to determine eligibility of the 
issue for the exception. The record preservation requirement also helps 
address concerns about the existence of some subjectivity involved in 
the selection of a particular structural credit risk model and data 
sample specifics by the lead managers, and the possibility of lead 
managers selecting these specifics so as to generate a probability of 
default estimate below the threshold level. The potential consequences 
of not including a record preservation requirement, therefore, could be 
that issues with high manipulation risk are allowed to rely on the 
exceptions from Regulation M. It is also intended to aid the Commission 
staff in examinations of the broker-dealers in evaluation of the 
specific model and data used to determine the probability of default of 
the issue in addition to exception eligibility.
---------------------------------------------------------------------------

    \311\ See SIFMA Letter 1, at 11.
---------------------------------------------------------------------------

VI. Paperwork Reduction Act

    Certain provisions of the final amendments contain ``collection of 
information'' requirements within the meaning of the Paperwork 
Reduction Act of 1995 (``PRA'').\312\ The hours and costs associated 
with determining whether a Nonconvertible Security qualifies for the 
new exception in Rule 101(c)(2)(i) and preserving the corresponding 
records under Rule 17a-4(b)(17) constitute PRA burdens.
---------------------------------------------------------------------------

    \312\ 44 U.S.C. 3501 et seq. The burdens associated with the 
information collection requirements are referred to as ``PRA 
burdens.''
---------------------------------------------------------------------------

    In accordance with the PRA, the Commission is submitting the final 
amendments to the rules to the Office of Management and Budget 
(``OMB'') for review.\313\ The Commission published a notice requesting 
comment on these collections of information requirements in the 
Proposal and submitted these requirements to the OMB for review in 
accordance with the PRA. An agency may not conduct or sponsor, and a 
person is not required to respond to, a collection of information 
unless it displays a currently valid control number. The titles and 
control numbers for these collections of information are as follows:
---------------------------------------------------------------------------

    \313\ See 44 U.S.C. 3507; 5 CFR 1320.11.

------------------------------------------------------------------------
                                                            OMB control
              Rule                        Title                 no.
------------------------------------------------------------------------
Rule 101.......................  Rule 101, 17 CFR              3235-0464
                                  242.101 (Activities by
                                  Distribution
                                  Participants).
Rule 17a-4.....................  Records to be Preserved       3235-0806
                                  by Certain Brokers and
                                  Dealers.
------------------------------------------------------------------------

    These PRA burdens are distinct from the existing OMB-approved 
collection of information burden estimates under Rules 101, 102, and 
17a-4 because the Commission has not estimated that respondents incur 
PRA burdens when determining whether a security qualifies for the 
Investment Grade Exception, nor did Rule 17a-4 include a recordkeeping 
requirement in connection with reliance on the Investment Grade 
Exception.\314\
---------------------------------------------------------------------------

    \314\ In the Proposal, the Commission proposed to eliminate the 
Investment Grade Exception under Rule 102 of Regulation M, without 
proposing an alternative standard in its place. However, as 
discussed above, in Part II.C, the Commission is adopting an 
exception that is based on an issuer's probability of default in 
both Rule 102 and Rule 101.
---------------------------------------------------------------------------

    The Commission did not receive any comments on the Proposal's PRA 
analysis. While one commenter did reference the potential burden of the

[[Page 39991]]

proposed amendments generally,\315\ no commenters specifically 
addressed the Commission's estimates of burdens and costs in the 
Proposal's PRA analysis. In addition, the Commission's estimates of the 
collection of information for the amendments, as adopted, have been 
updated from the estimates included in the Proposal, as appropriate, 
with the updated estimates based on the modifications in the adopted 
rule and based on more recent data.
---------------------------------------------------------------------------

    \315\ SIFMA Letter 1, at 5.
---------------------------------------------------------------------------

    The Commission is adopting in Rules 101 and 102 the proposed 
exception that is based on an issuer's probability of default, as 
described above in Parts II.A and B, to replace the Investment Grade 
Exceptions. The Commission is also adopting a corresponding record 
preservation requirement in Rule 17a-4(b), which requires broker-
dealers to preserve the written probability of default determination, 
relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 
102(d)(2)(i), as applicable.
    As discussed above, Regulation M is designed to preserve the 
integrity of the securities trading market as an independent pricing 
mechanism by prohibiting activities that could artificially influence 
the market for an offered security. Subject to exceptions, Rule 101 
prohibits distribution participants and their affiliated 
purchasers,\316\ and Rule 102 prohibits issuers, selling security 
holders, and their affiliated purchasers, from directly or indirectly 
bidding for, purchasing, or attempting to induce another person to bid 
for or purchase a covered security during a restricted period.\317\ 
Rule 17a-4 requires a broker-dealer to preserve certain records if it 
makes or receives them.\318\
---------------------------------------------------------------------------

    \316\ 17 CFR 242.101.
    \317\ 17 CFR 242.102.
    \318\ 17 CFR 240.17a-4.
---------------------------------------------------------------------------

    In accordance with the requirements of section 939A(b), the 
Commission is adopting amendments to Rules 101 and 102 of Regulation M 
that remove the Investment Grade Exceptions and add, in their place, 
new exceptions for Nonconvertible Securities for which the issuer's 
probability of default, estimated as of the sixth business day 
immediately preceding the determination of the offering price and over 
the horizon of 12 full calendar months from such day, is 0.055% or 
less, as determined and documented, in writing, by the distribution 
participant acting as the lead manager (or in a similar capacity) of a 
distribution, as derived from a structural credit risk model.\319\ The 
Commission is also adopting Rule 17a-4(b)(17), which requires broker-
dealers to preserve the written probability of default determination, 
relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 
102(d)(2)(i), as applicable, for a period of not less than three years, 
the first two years in an easily accessible place.\320\
---------------------------------------------------------------------------

    \319\ 17 CFR 242.101(c)(2)(i), as amended, 242.102(d)(2)(i), as 
amended.
    \320\ 17 CFR 240.17a-4(b)(17), as amended.
---------------------------------------------------------------------------

    The Commission is also adopting identical new exceptions in Rules 
101(c)(2)(ii) and 102(d)(2)(ii) for asset-backed securities that are 
offered pursuant to an effective shelf registration statement filed on 
Form SF-3.\321\ The discussion of estimates that follows is limited to 
the new information collection requirements that result from the final 
amendments related to the probability of default determinations in Rule 
101(c)(2)(i), as amended, as well as the record preservation thereof in 
reliance on the new exceptions provided in Rule 101(c)(2)(i) or Rule 
102(d)(2)(i) pursuant to Rule 17a-4(b)(17). The Commission is not 
estimating that the new exception for asset-backed securities that are 
offered pursuant to an effective shelf registration statement filed on 
Form SF-3 in Rules 101(c)(2)(ii) and 102(d)(2)(ii) will increase or 
decrease the existing approved information collections because whether 
an asset-backed security is offered pursuant to an effective shelf 
registration statement filed on Form SF-3 is an objective, observable 
fact that would not incur any PRA burden.
---------------------------------------------------------------------------

    \321\ 17 CFR 242.101(c)(2)(ii), as amended, 242.102(d)(2)(ii), 
as amended.
---------------------------------------------------------------------------

A. Respondents

    The respondents under the amended rules are lead managers who 
choose to make a probability of default determination in order to rely 
on the exception for Nonconvertible Securities and other broker-dealers 
who use the lead manager's probability of default determination in 
relying on an exception for Nonconvertible Securities. As noted in Part 
V.A.1, there were 201 lead managing underwriters and 100 other non-lead 
manager broker-dealers of Nonconvertible Securities in 2021.\322\ The 
Commission assumes that, on balance, these numbers will remain 
consistent given the capital, expertise, and relationships needed to 
serve as the lead underwriter of a Nonconvertible Securities offering. 
The Commission, therefore, is estimating that 301 respondents will be 
subject to PRA burdens under the amendments. The respondents under the 
amendments to Rule 101(c)(2)(i) are lead managers who make probability 
of default determinations. The Commission, therefore, is estimating 
that 201 respondents will be subject to PRA burdens under Rule 
101(c)(2)(i), as amended. The respondents under the amendments to Rule 
17a-4(b)(17) are broker-dealers who rely on the new exception in Rule 
101(c)(2)(i) or Rule 102(d)(2)(i). The Commission, therefore, is 
estimating that 301 respondents will be subject to PRA burdens under 
new Rule 17a-4(b)(17).\323\
---------------------------------------------------------------------------

    \322\ See supra Part V.A.1.
    \323\ [201 lead manager broker-dealers] + [100 non-lead manager 
broker-dealers] = 301 respondents under new Rule 17a-4(b)(17).
---------------------------------------------------------------------------

B. Use of Information

    The information collected under the amendments ensures that the 
Nonconvertible Securities that are least likely to be subject to the 
type of manipulation that Regulation M seeks to address are excepted 
from Rules 101 and 102. Further, the Commission believes that the 
information contained in the records required to be preserved pursuant 
to Rule 17a-4(b)(17) will facilitate the Commission in conducting 
examinations of broker-dealers who rely on the new exceptions in Rule 
101(c)(2)(i) or Rule 102(d)(2)(i).

C. Collection of Information

    As discussed below, the Commission believes that respondents will 
incur PRA burdens under the amendments to Rule 101(c)(2)(i) because 
distribution participants who are acting as the lead manager (or in a 
similar capacity) of a distribution and make a probability of default 
determination are required for each distribution of Nonconvertible 
Securities to determine the subject issuer's probability of default in 
order to rely on the exception. These respondents may also incur PRA 
burdens in their probability of default determinations. Respondents who 
are broker-dealers and rely on the new exception in Rule 101(c)(2)(i) 
or Rule 102(d)(2)(i) will incur PRA burdens under the requirements set 
forth in new Rule 17a-4(b)(17) because they are required to preserve 
records of the written probability of default determination.
1. Burden and Cost Estimates Related to the Rule 101 Amendments
    Rule 101(c)(2)(i), as amended, permits lead managers to gather the 
data serving as the inputs and then perform the analysis necessary to 
calculate the probability of default of the issuer

[[Page 39992]]

whose securities are the subject of the distribution to meet the 
conditions of the exception.\324\ This requirement will result in 
respondents incurring a PRA recordkeeping burden. This process will 
likely be highly automated, and that respondents will initially comply 
with this requirement by reprograming systems to create a means to 
calculate electronically the probability of default based on manually 
gathered and entered inputs for financial modeling. The respondents who 
make probability of default determinations will be broker-dealers 
serving as lead managers and are likely to have experience in using 
their own proprietary version of a publicly available structural credit 
risk model. Accordingly, the initial configuration of systems will be 
handled internally and take 3 hours per respondent. The Commission also 
assumes that broker-dealers serving as lead managers already have the 
software and systems in place required to make the calculations.\325\ 
The Commission therefore estimates that the total industry-wide initial 
burden for configuring systems to make and probability of default 
estimates is 603 hours.\326\
---------------------------------------------------------------------------

    \324\ The Commission recognizes that some respondents may choose 
to utilize the probability of default estimates that are calculated 
and made available by a third-party vendor rather than make the 
determination themselves. In the Proposal, the Commission noted that 
the Commission's burden estimates for the adopted amendments to Rule 
101 are based on respondents gathering the required data and 
calculating the probability of default, internally, without the use 
of third-party vendors, because the Commission lacks granular 
information from which to base an estimate of the proportion of 
respondents that would use vendors. The Commission requested comment 
on the extent to which respondents may use third-party vendors, as 
well as the costs and time burdens of using such services. See 
Proposal, 87 FR 18326 n.129. However, the Commission did not receive 
comments in response to this request. For purposes of estimating the 
PRA burdens under the final rules as amended, the Commission 
continues to assume that all respondents will make the probability 
of default determination internally with data they have gathered, 
rather than use third party vendors. As discussed above, in Part 
II.A.1, there may be distributions with more than one distribution 
participant acting as the lead manager (or in a similar capacity), 
but only one of the distribution participants acting as the lead 
manager would be permitted to make the probability of default 
determination for the particular distribution. See supra note 91. 
Therefore, for purposes of the PRA estimations in this release, only 
one lead manager on any distribution for purposes of these 
calculations is assumed.
    \325\ Further, respondents who choose to utilize probability of 
default estimates that are calculated and made available by a third-
party vendor will already have access to the vendor's software and 
systems containing these estimates, typically as part of an existing 
subscription, so they will not need to procure further services or 
subscriptions from these vendors to access any such determinations. 
However, as noted above, for purposes of estimating these PRA 
burdens, the Commission assumes all respondents would make their own 
calculations and not use third party vendors. This assumption is 
being made to provide an estimate reflecting for the more costly of 
the two approaches.
    \326\ [201 lead managers] x [3 hours] = 603 hours. The Proposal 
included 237 respondents, which was taken from available data from 
2020. The number included herein reflects the number from the 
available data from 2021, as discussed above, in Part V.A.1. In 
addition, under the Proposal, the 237 figure included non-lead 
manager broker-dealers who would have been eligible, under the 
proposed Rule 101(c)(2)(i), to make probability of default 
determinations in order to meet the Nonconvertible Securities 
exception's conditions.
---------------------------------------------------------------------------

    An issuer's probability of default is forward-looking and changes 
over time, so the Commission believes that respondents will manually 
gather the inputs required to calculate an issuer's probability of 
default each time it participates in a distribution of Nonconvertible 
Securities. There were 33,798 offerings of Nonconvertible Securities in 
2021.\327\ Because financial modeling generally, and the probability of 
default calculation more specifically, is well-known by industry 
participants, the Commission believes that respondents have employees 
who are familiar with how to gather the required model inputs. The 
Commission, therefore, estimates that it will take lead-manager 
respondents roughly one hour per distribution of Nonconvertible 
Securities to determine and document, in writing, the probability of 
default determinations. Accordingly, the Commission estimates that 
calculating the probability of default pursuant to Rule 101(c)(2)(i), 
as amended, will result in an aggregate annual ongoing industry-wide 
burden of 33,798 hours. The Commission estimates that the total PRA 
burden resulting from the final amendments to Rule 101 is 34,401 hours 
in the first year \328\ and 33,798 hours thereafter.
---------------------------------------------------------------------------

    \327\ This number was obtained from Mergent, a financial data 
provider.
    \328\ [603 hours (initial burden)] + [33,798 hours (ongoing 
annual burden)] = 34,401 hours.
---------------------------------------------------------------------------

    The Commission does not believe that the amendments to Rule 
101(c)(2)(ii) excepting asset-backed securities that are offered 
pursuant to an effective shelf registration statement filed on Form SF-
3 will result in respondents incurring PRA burdens because whether an 
asset-backed security has an effective shelf registration statement 
filed on Form SF-3 is an objective, observable fact.\329\ Further, 
there is no corresponding record preservation requirement for 
respondents documenting reliance on the exception for asset-backed 
securities under Rule 101(c)(2)(ii), as amended.
---------------------------------------------------------------------------

    \329\ See 17 CFR 239.45.
---------------------------------------------------------------------------

2. Burden and Cost Estimates Related to the Rule 17a-4 Amendments
    New Rule 17a-4(b)(17) requires broker-dealers to preserve the 
written probability of default determination, relied upon pursuant to 
the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as 
applicable, for a period of not less than three years, the first two 
years in an easily accessible place.
    The Commission estimates that this record preservation requirement 
imposes an initial burden of 25 hours per respondent for updating the 
applicable policies and systems required to account for preserving the 
records made pursuant to Rule 101(c)(2)(i), as amended. Accordingly, 
the Commission estimates that the total industry-wide initial burden 
for this requirement is 7,525 hours.\330\ The Commission also estimates 
that respondents will incur an ongoing annual burden of 10 hours per 
firm for maintaining such records, as well as to make additional 
updates to the applicable record preservation policies and systems to 
account for preserving the records pursuant to new Rule 17a-4(b)(17), 
resulting in a total ongoing industry-wide burden of 3,010 hours.\331\ 
The Commission, therefore, estimates that the total PRA burden 
resulting from the amendment to Rule 17a-4 is 10,535 hours in the first 
year \332\ and 3,010 hours per year thereafter.
---------------------------------------------------------------------------

    \330\ [301 respondents] x [25 hours] = 7,525 hours.
    \331\ [301 respondents] x [10 hours] = 3,010 hours.
    \332\ [7,525 hours (initial burden)] + [3,010 hours (ongoing 
annual burden)] = 10,535 hours.

                                                PRA Summary Table
----------------------------------------------------------------------------------------------------------------
                                                                                  Ongoing annual
                                                                  Initial burden   burden hours/     Total PRA
            Industry-wide burden due to amendments to                  hours        year (after    burden hours
                                                                                    first year)    in first year
----------------------------------------------------------------------------------------------------------------
Rule 101........................................................             603          33,798          34,401

[[Page 39993]]

 
Rule 17a-4......................................................           7,525           3,010          10,535
----------------------------------------------------------------------------------------------------------------

D. Collection of Information Is Mandatory

    The information collections for making probability of default 
determinations under the amendments to Rule 101 are mandatory for 
reliance on exceptions in Rule 101(c)(2)(i) or Rule 102(d)(2)(i). In 
addition, the information collections involving the preservation of 
written probability of default determinations under the amendments to 
Rule 17a-4 are mandatory if a broker-dealer relies on the new exception 
in Rule 101(c)(2)(i) or Rule 102(d)(2)(i).

E. Confidentiality of Responses to Collection of Information

    The Commission would not typically receive confidential information 
as a result of these collections of information. To the extent that the 
Commission receives--through its examination and oversight program, 
through an investigation, or by some other means--records or 
disclosures from a distribution participant regarding the probability 
of default determination, such information would be kept confidential, 
subject to the provisions of applicable law.

F. Retention Period for Record Preservation Requirement

    Pursuant to new Rule 17a-4(b)(17), a broker-dealer is required to 
preserve the written probability of default determination, relied upon 
pursuant to the new exception in Rule 101(c)(2)(i) or Rule 
102(d)(2)(i), as applicable, for a period of not less than three years, 
the first two years in an easily accessible place.

VII. Regulatory Flexibility Act Certification

    The Regulatory Flexibility Act (``RFA'') \333\ requires Federal 
agencies, in promulgating rules, to consider the impact of those rules 
on ``small entities,'' \334\ a term that includes ``small businesses.'' 
\335\ Section 603(a) \336\ of the Administrative Procedure Act,\337\ as 
amended by the section 604(a) of the RFA requires the Commission to 
undertake a final regulatory flexibility analysis of rules it is 
adopting, unless the Commission certifies that the rules would not have 
a significant impact on a substantial number of small entities.\338\
---------------------------------------------------------------------------

    \333\ 5 U.S.C. 601 et seq.
    \334\ 5 U.S.C. 605(b).
    \335\ Although section 601(b) of the RFA defines the term 
``small business,'' the statute permits agencies to formulate their 
own definitions. The Commission has adopted definitions for the term 
``small business'' for the purposes of Commission rulemaking in 
accordance with the RFA. Those definitions, as relevant to this 
rulemaking, are set forth in 17 CFR 240.0-10 (``Rule 0-10''). Rule 
0-10 also provides that the Commission may, if warranted by the 
circumstances, use a different definition for particular 
rulemakings. See 17 CFR 240.0-10.
    \336\ 5 U.S.C. 603(a).
    \337\ 5 U.S.C. 551 et seq.
    \338\ 5 U.S.C. 605(b).
---------------------------------------------------------------------------

    Small entities include broker-dealers with total capital (net worth 
plus subordinated liabilities) of less than $500,000 on the date in the 
prior fiscal year as of which its audited financial statements were 
prepared pursuant to Rule 17a-5(d) under the Exchange Act,\339\ or, if 
not required to file such statements, a broker-dealer who had total 
capital (net worth plus subordinated liabilities) of less than $500,000 
on the last day of the preceding fiscal year (or in the time it has 
been in business, if shorter), and is not affiliated with any person 
(other than a natural person) who is not a small business or small 
organization.\340\ A small business or small organization, for purposes 
of ``issuers'' or ``person'' other than an investment company, is 
defined as a person who, on the last day of its most recent fiscal 
year, had total assets of $5 million or less.\341\ In the Proposal, the 
Commission certified, pursuant to section 605(b) of the RFA, that the 
proposed amendments to Rules 101 and 102 would not have a significant 
economic impact on a substantial number of small entities.\342\ The 
Commission requested but did not receive any comments on the 
certification as it related to the entities impacted by Rule 101 or 
Rule 102 of Regulation M, or by Rule 17a-4 under the Exchange Act.
---------------------------------------------------------------------------

    \339\ See 17 CFR 240.17a-5(d).
    \340\ See 17 CFR 240.0-10(c).
    \341\ 17 CFR 242.0-10(a).
    \342\ See Proposal, 87 FR 18337.
---------------------------------------------------------------------------

    Based on the Commission's analysis of the existing information 
relating to broker-dealers who are subject to Rules 101, 102,\343\ and 
17a-4, it is unlikely that any broker-dealer categorized as a ``small 
business'' or ``small organization'' under Rule 0-10 could serve as an 
underwriter or other distribution participant, as they would almost 
certainly have insufficient capital to participate in underwriting 
activities. In addition, the Commission continues to believe that none 
of the various persons affected by the amendments would qualify as a 
small entity under the Rule 0-10 definition as it is unlikely that any 
issuer of that size had investment grade securities that were eligible 
for the Investment Grade Exception. Accordingly, the Commission 
believes it is unlikely that, in the future, a small entity may become 
impacted by the amendments because broker-dealers who enter this market 
are likely to have at least $500,000 in total capital, as described 
above, or to be affiliated with a person who is not a small business or 
small organization as defined under Rule 0-10, and because issuers of 
securities that are eligible for the new exceptions provided in Rules 
101(c)(2) and 102(d)(2) are likely to have total assets greater than $5 
million.
---------------------------------------------------------------------------

    \343\ As discussed above, in Part II.B, broker-dealers who are 
affiliated with the issuer and do not meet the definition of 
``distribution participant'' under Rule 100(b) of Regulation M may 
be covered persons under Rule 102. Even if those broker-dealers had 
net capital over $500,000, they would not be small entities under 
Rule 0-10 because they are affiliated with an issuer (of investment 
grade securities) that is not a small entity.
---------------------------------------------------------------------------

    For the foregoing reasons, the Commission certifies, pursuant to 
section 605(b) of Title 5 of the U.S. Code, that the amendments to 
Rules 100, 101, 102, and 17a-4 will not have a significant economic 
impact on a substantial number of small entities.

Statutory Authority

    The final amendments contained in this release are being adopted 
under the authority set forth in sections 939 and 939A of the Dodd-
Frank Act and

[[Page 39994]]

sections 3(b), 15, 23(a), and 36 of the Exchange Act.

List of Subjects in 17 CFR Parts 240 and 242

    Broker-dealers, Fraud, Issuers, Reporting and recordkeeping 
requirements, Securities.

Text of Rule Amendments

    For the reasons set out in the preamble, the Commission is amending 
title 17, chapter II of the Code of Federal Regulations as follows:

PART 240--GENERAL RULES AND REGULATIONS, SECURITIES EXCHANGE ACT OF 
1934

0
1. The authority citation for part 240 continues to read, in part, as 
follows:

    Authority:  15 U.S.C. 77c, 77d, 77g, 77j, 77s, 77z-2, 77z-3, 
77eee, 77ggg, 77nnn, 77sss, 77ttt, 78c, 78c-3, 78c-5, 78d, 78e, 78f, 
78g, 78i, 78j, 78j-1, 78j-4, 78k, 78k-1, 78l, 78m, 78n, 78n-1, 78o, 
78o-4, 78o-10, 78p, 78q, 78q-1, 78s, 78u-5, 78w, 78x, 78dd, 78ll, 
78mm, 80a-20, 80a-23, 80a-29, 80a-37, 80b-3, 80b-4, 80b-11, 7201 et 
seq., and 8302; 7 U.S.C. 2(c)(2)(E); 12 U.S.C. 5221(e)(3); 18 U.S.C. 
1350; and Pub. L. 111-203, 939A, 124 Stat. 1376 (2010); and Pub. L. 
112-106, sec. 503 and 602, 126 Stat. 326 (2012), unless otherwise 
noted.
* * * * *
    Section 240.17a-4 also issued under secs. 2, 17, 23(a), 48 Stat. 
897, as amended; 15 U.S.C. 78a, 78d-1, 78d-2; sec. 14, Pub. L. 94-
29, 89 Stat. 137 (15 U.S.C. 78a); sec. 18, Pub. L. 94-29, 89 Stat. 
155 (15 U.S.C. 78w);
* * * * *

0
2. Amend Sec.  240.17a-4 by adding paragraph (b)(17) to read as 
follows:


Sec.  240.17a-4  Records to be preserved by certain exchange members, 
brokers and dealers.

* * * * *
    (b) * * *
    (17) The written probability of default determination, relied upon 
by such broker or dealer, pursuant to Sec.  242.101(c)(2)(i) or Sec.  
242.102(d)(2)(i) of this chapter (Rule 101 or Rule 102 of Regulation 
M), as applicable.
* * * * *

PART 242--REGULATIONS M, SHO, ATS, AC, NMS, AND SBSR AND CUSTOMER 
MARGIN REQUIREMENTS FOR SECURITY FUTURES

0
3. The authority citation for part 242 continues to read as follows:

    Authority: 15 U.S.C. 77g, 77q(a), 77s(a), 78b, 78c, 78g(c)(2), 
78i(a), 78j, 78k-1(c), 78l, 78m, 78n, 78o(b), 78o(c), 78o(g), 
78q(a), 78q(b), 78q(h), 78w(a), 78dd-1, 78mm, 80a-23, 80a-29, and 
80a-37.


0
4. Amend Sec.  242.100 in paragraph (b) by adding in alphabetical order 
a definition for ``Structural credit risk model'' to read as follows:


Sec.  242.100  Preliminary note; definitions.

* * * * *
    (b) * * *
    Structural credit risk model means any commercially or publicly 
available model that calculates, based on an issuer's balance sheet, 
the probability that the value of the issuer will fall below the 
threshold at which the issuer would fail to make scheduled debt 
payments, at or by the expiration of a defined period.
* * * * *

0
5. Amend Sec.  242.101 by revising paragraph (c)(2) to read as follows:


Sec.  242.101  Activities by distribution participants.

* * * * *
    (c) * * *
    (2) Certain nonconvertible and asset-backed securities. (i) 
Nonconvertible debt securities and nonconvertible preferred securities 
of issuers for which the probability of default, estimated as of the 
sixth business day immediately preceding the determination of the 
offering price and over the horizon of 12 full calendar months from 
such day, is 0.055% or less, as determined and documented, in writing, 
by the distribution participant acting as the lead manager (or in a 
similar capacity) of a distribution, as derived from a structural 
credit risk model; or
    (ii) Asset-backed securities that are offered pursuant to an 
effective shelf registration statement filed on Form SF-3 (Sec.  239.45 
of this chapter); or
* * * * *

0
6. Amend Sec.  242.102 by revising paragraph (d)(2) to read as follows:


Sec.  242.102  Activities by issuers and selling security holders 
during a distribution.

* * * * *
    (d) * * *
    (2) Certain nonconvertible and asset-backed securities. (i) 
Nonconvertible debt securities and nonconvertible preferred securities 
of issuers for which the probability of default, estimated as of the 
sixth business day immediately preceding the determination of the 
offering price and over the horizon of 12 full calendar months from 
such day, is 0.055% or less, as determined and documented, in writing, 
by the distribution participant acting as the lead manager (or in a 
similar capacity) of a distribution, as derived from a structural 
credit risk model, pursuant to Sec.  242.101(c)(2)(i); or
    (ii) Asset-backed securities that are offered pursuant to an 
effective shelf registration statement filed on Form SF-3 (Sec.  239.45 
of this chapter); or
* * * * *

    By the Commission.

    Dated: June 7, 2023.
Vanessa A. Countryman,
Secretary.
[FR Doc. 2023-12591 Filed 6-16-23; 8:45 am]
BILLING CODE 8011-01-P