Sources: Col. (1) and (2), Arthur D. Little, Inc.; Col. (3), Norgaard and Schick, op. cit., p. 8. NS-9 contained in the Norgaard and Schick data. It does illustrate, however, that the distortions uncovered in the reviewers' averages appear to permeate their individual company analysis. This We must now isolate the cause of the reviewers' distortion. can be accomplished only be examining in detail an individual company's data. The only stock insurance company for which we have the Norgaard and Schick data is the Continental Casualty Company. There is nothing atypical about the actions or performance of Continental during the period studied its actions and returns were like most other property and liability insurance companies. We use its data solely because it is the only stock company for which we have the necessary detailed data to duplicate the reviewers' calculations. To understand the implications of the Norgaard and Schick approach to accounting, we will examine in detail the profit accounting which they report for Continental Casualty Company in the year 1966. As can be seen from Dr. Norgaard's worksheet (reproduced as Appendix B), in 1966 Continental Casualty experienced a $5-million increase in assets. It paid $11.9 million in dividends and bought $800,000 of its treasury stock. Professor Norgaard indicates that these numbers imply a "return" of $17.7 million. Relating this return to an asset base of $1006 million, Professor Norgaard reports a "rate of return" of 1.7%--not an overly profitable year; but, nevertheless, a year in which we are told Continental Casualty Company experienced a $17.7-million profit. Let us now turn to the financial statement filed by Continental Casualty reported in the authoritative A.M. Best's Insurance Reports and Moody's Bank and Finance Manual. We note that in 1966 the company sustained a loss of each of its major activities. In its insurance business it sustained a statutory underwriting loss of $10 million; on the investment side, combining interest and dividends received with realized capital gains and unrealized capital gains, Continental experienced a net loss of $3 million. In round numbers, the total loss was therefore $13 million. On an accrual basis (i.e., adjusting earnings for increases in equity appearing in the unearned premium reserves), the loss is $9 million. These income statement results are well reflected in the company's balance sheet. In the year 1966, the net worth of policyholders' surplus of Continental declined by $28 million. Part of this decline was attributable to the payment of $11.9 million in dividends to stockholders. The decline in policyholders' surplus, excluding the payment of dividends, was thus $16.1 million. It is standard accounting and financial theory that one of the best single measures of a company's profitability is its change in net worth. It would seem awkward, if not mischievous, to characterize a company experiencing a $16.1-million decline in net worth (excluding dividends) as having a profitable year. Yet Norgaard and Schick report a $17.7-million profit for Continental Casualty in 1966. These numbers are typical of the distortion the Norgaard and Schick method forces into their insurance analyses. It is therefore not surprising that their figures bear no relation to ADL's results or to those of any other investigator of the insurance industry. An extensive search of the recent literature finds only Professor Hofflander accepting the Norgaard and Schick paper, and he expresses serious doubt about the method used by Norgaard and Schick to calculate profits. 11 One unacceptable defense of the peculiar Norgaard and Schick approach is the contention that while it may cause distortion in any one year this distortion will "wash out" over the 15 years on which they base their study. This contention is untrue. To illustrate the sustained effect of the Norgaard and Schick profit measure, we again turn to the Continental Casualty Company. 12 Norgaard and Schick report (p. 8, Table 2) that Continental Casualty had a 15-year rate of return of 16.2%. From Professor Norgaard's worksheet (see Appendix B, p. NS-30) of this report, we see that this number is derived by taking an unweighted average of the individual yearly rates of return. However, taking the same 15 years and measuring profit by the method described above (i.e., the sum of net income plus the incomes attributable to investment activities), we see in Table 2 that the Continental Casualty Company experienced only a 3.67% return for the period. 13 It should be emphasized that the basis of the profit figures for this calculation was not Arthur D. Little, Inc.'s, but was drawn from one of the standard reference manuals for financial analysis, the authoritative Moody's Bank and Finance Manual. Further, it is Moody who presents and titles the two income streams as "net income" and "gain in value of investments (net)." It is clear from the Moody statement that these two are the primary income figures for insurance companies. The asset base 11Dr. Hofflander, as quoted in the November 4, 1968, issue of the National Underwriter, "is unhappy with their [Norgaard and Schick's] definition of income. 12Dr. Hofflander, as quoted in the November 4, 1968, issue of the National Underwriter, "obviously if a company has a premium volume growing fairly rapidly then the rate of return would be overstated for that time period. It is the author's [Norgaard and Schick's] contention that since they used a 15-year period of time this particular item washes out, but I'm not convinced that this is true." 13We believe that weighted averages are appropriate for this type of measurement and have used them in our work. The differences between the Norgaard and Schick and ADL results cannot be attributed to this minor influence. For example, the unweighted average of the Continental Casualty Company's 15 yearly rates of return is 4.6%. NS-11 eighted average is 4.55%. nweighted average. ces: Cols. (1) and (2), Moody's Bank & Financial Manual, April 1968, dends to policyholders. Col. (4), Best's Reports, Consolidated Balance Sheet using the Norgaard and Schick definition of assets as exemplified in Norgaard's worksheet. Col. (6), Norgaard's worksheet and the Norgaard and Schick state ment. figures, by which the income streams were divided to obtain the rate of return, are taken from Best's Reports using the Norgaard and Schick definition of assets as exemplified in their worksheet. It is instructive to compare Columns 5 and 6 of Table 2. In Column 5 we report the rate of return which conforms to standard accounting, finance, and economic theory. It is an all-inclusive rate of return in that it considers all possible sources of income, be they operating income or investment income (interest and dividends, realized capital gains, and unrealized capital gains). For the 15-year period, the average rate of return was 3.67%, typical of the entire industry for that period. 14 Column 6 reports the Norgaard and Schick rate of return for Continental Casualty. In every instance it is significantly higher than the rate given in Column 5. In one case, 1956, it reaches 70.4%--more than 16 times higher than the true rate of return. That rate of return alone should alert any scholar that there is, perhaps, some faultiness in the definition. However, it seems not to concern Professors Norgaard and Schick, for they accept this extraordinary return with equanimity. They reported an average rate of return for the 15-year period of 16.2%, or 450 percent greater than the all-inclusive rate of return reported in Column 5. It is clear, then, that the distortion caused by measuring profits primarily as growth in assets (the reviewers' approach) does not "wash out" over a 15-year period. The Norgaard and Schick rate of return bears no relationship to the rate of return as measured by any acceptable or useful standard of profitability for insurance companies. Nor does it bear any relationship to any definition of rate of return that has been put forward by the accounting, legal, financial, or economics professions. In the hypothetical examples given above we saw that the Norgaard and Schick approach had the potentiality for a gross distortion of the rate of return. A hypothetical firm generally held to be losing a great deal of money can turn up on the reviewers' scale as earning a high rate of In 14It is instructive to note how in 1956 the Continental Casualty Company registered a 70.4% return or the reviewers' scale of profitability. 1956 the Continental acquired financial control of the National Fire Insurance Co. of Hartford by an exchange of 1-1/4 shares for each National Fire share. The consolidated balance sheet used by Norgaard and Schick showed assets increasing from $274 million to $453 million between 1955 and 1956. This increase of 65.3% the reviewers' formula counted as part of the rate of return. When added to dividends and treasury stock, it gave the 70.4% rate of return Norgaard and Schick report. Typically, however, their measure did not consider changes in liabilities. In this case the growth rate of Continental's liabilities was 71.1% or greater than its asset growth rate. Considering the asset growth due to consolidation as profit (the reviewers approach) implies that Continental could sell off the merged company's assets and not honor its liabilities. NS-13 |