New York Codes, Rules and Regulations
Title 11 - INSURANCE
Chapter V - Rates And Rating Organizations
Subchapter F - Treatment Of Excess Profits In Motor Vehicle Insurance
Part 166 - Treatment Of Excess Profits In Motor Vehicle Insurance
Subpart 166-2 - Discussion Of General Rules
Section 166-2.1 - Discussion of reasonable rate of return

(a) A reasonable rate of return has been established by utilizing the average annual aftertax rate of return earned by other industries experiencing substantially the same variation in their rates of return. Professor Williams offers two choices for the reasonable rate of return, both based on standard statistical techniques. One measures risk using the variance in insurers' annual rates of return in New York; the other approach uses the standard deviation in these annual rates of return. Using data from 1962 through 1978, the department's consultant calculates a reasonable return of 21.6 percent (using variance) and 17.9 percent (using standard deviation).[FN3] The department has recalculated these figures using data through 1980 and has obtained returns of 22.7 percent (variance) and 18.3 percent (standard deviation).

(b) The consultant states that both the variance measure and the standard deviation measure can be statistically and logically justified. He chooses the standard deviation measure (rounded to 18 percent), in part because "the indicated reasonable rate of return is closer to what most of the other methods [discussed in the report] appear to suggest."[FN4] Although the 18- percent reasonable profit point exceeds the return actually earned by the industry, historical results reflect two factors: interest rates far lower than current levels, and the disastrous underwriting losses of the mid-1970's. Although lower investment yields affected countrywide results as well, less severe underwriting results have produced national rates of return considerably closer to the 18-percent level. As insurers' long-term portfolios (containing investments at extremely low yields) are gradually reinvested in higher yielding securities, it is expected that the return actually earned by the industry should more closely approximate 18 percent. The department therefore concurs with Professor Williams' choice of 18 percent.

(c) Several speakers at the October 5, 1982 public hearing contended that the choice of 18 percent was too high, in view of the falling interest rates of the last few months, and the historical failure of the industry to earn such a return over an extended period. The department has rejected the suggestion to reduce the "reasonable" point at this time, however, because it is undesirable to modify the results of a long-term analysis as a result of recent events which may not persist.

(d) Before selecting the Williams approach for estimating the reasonable rate of return, the department had reviewed other methods of measuring profitability. Two of these deserve particular mention. The Capital Asset Pricing Model (CAP-M) has been used in other states, and is often employed in financial analysis. Professor Williams expresses several reservations about CAP-M, including the fact that there are serious technical difficulties involved in estimating the parameters of the model which would be needed to apply CAP-M retrospectively to one line of business in one state. Additionally, this department observed that CAP-M is highly sensitive to volatile interest rates (since one of its parameters is the currently avaiable risk-free rate of return), even though these rates often do not reflect what insurers actually earn. While this method may prove useful at some future date in a more stable financial market, it would be an unwise selection at this time.

(e) Use of the discounted cash flow method has also been suggested to estimate a reasonable rate of return. It is extremely difficult to obtain representative data for this analysis; data submitted by the Consumer Protection Board in support of this approach (which has been used in public utility regulation) were based on nine stock insurers, two of which do not write automobile business in New York. The method analyzes insurers' dividend patterns and retained earnings countrywide for all lines in order to estimate a reasonable rate of return. No reflection of the underwriting results of New York's automobile business is included. This deficiency appears to invalidate the approach, since one intent of the statute is to provide a safeguard in the event that unusual occurrences, such as a no-fault law or other legislative enactment, would produce windfall profits in New York. The department notes that much of the difficulty of implementing section 677(5) (now section 2329) arises from the requirement that a "by line, by state" analysis be made, although many of the components are, by their nature, companywide quantities. Additionally, there is no methodology for selecting an excess profit above the reasonable point, as the law requires. While there is much to admire in the work of public utility regulators, their approach is not generally applicable to the task before the department, largely because of the differing competitive nature of the industries.

(f) An additional consideration in the department's choice of its consultant's approach, and of the 18-percent reasonable rate of return, is the department's own survey of money managers and investment advisers. The survey, mentioned in Williams (at page 17), asked the respondents to indicate what rate of return insurers specializing in automobile insurance should realize. The median rate of return specified by these managers and advisers was 18 percent.

(g) The department recognizes that the 18-percent figure may be inappropriate for future years, as economic conditions and actual insurer results may change. An important part of the department's ongoing monitoring of profitability will be periodic reevaluation of this figure.

[FN3] Williams, pp. 46-47.

[FN4] Williams, p. 48.