Page images
PDF
EPUB

pany a take-over target, many insurance company variables were examined. From this large number of variables, four were found to be most significant:

1. AVERAGE EXPENSE RATIO. The expense ratio is the total of all costs connected with selling insurance policies divided by premiums written. This ratio is an indication of how efficiently a firm is operating. It includes the costs of home office operations as well as commissions to salesmen. This ratio ranged from 39.42 for Trinity Universal to 13.30 for Government Employees.

2. STOCK AS A PERCENTAGE OF PORTFOLIO. This ratio indicates how aggressively a firm manages its portfolio. Apparently the higher the stock percentage the more aggressively management is managing the portfolio and the less likely it is that outsiders could improve the performance. This value ranged from 76.33% for Continental Insurance to 7.4% for American Motorist.

3. RETURN ON PORTFOLIO. This figure represents the total of interest, dividends and market appreciation as a percentage of portfolio at market. It measures portfolio performance. It is, of course, greatly affected by the period selected for measurement. It was assumed in the case of this variable that the better the performance the less attractive was the firm as a take-over target. The values ranged from 078 for American Home to .021 for Republic Insurance. 4. PRICE TO Liquidating valUE. Liquidating value is the supposed value of the insurance company if sold in parts. It includes the market value of securities plus unrealized gain in unearned premiums less liabilities. This figure compared the average market price for each firm's stock to its liquidating value. A company selling for a price substantially below liquidating value could be purchased with its own assets, thus making it quite attractive. The values ranged from 1.31 for American Home to .48 for Crum-Forster (U. S. Fire).

The results of the initial compilation of variables indicated that some combination of them could distinguish targets from non-targets. The technique used for making this determination was linear-discriminant analysis.

Linear-discriminant analysis (LDA) is a tool often used to evaluate credit risks," and has application to many other financial problems. LDA is particularly well suited for the type of problem described in this article. The requirements are: 1) Data points (in this case, insurance companies) which can be divided into two groups depending on the presence or absence of a single attribute (good credit risk, bad credit risk; target, non-target). 2) A number of readily quantifiable at

tributes, values of which can be associated with each data point. The technique differs from the well-known linear regression-correlation analysis in that the dependent variable in regression analysis is known; whereas, in LDA it is unknown and to be determined.

The procedure, which is normally performed by a computer using a "canned" program, determines the one linear combination of the attributes (i.e., the discriminant function) which best classifies the two groups according to the value of a number called the "Z" value. This number is used to determine into which of the two categories a particular data point falls. Various tests are available to determine if the equation is reliable—that is, how often the discriminant function selected by the computer program misclassifies a data point for which the true classification is known.

In this case we selected six firms which had been acquired and arrayed them with six firms which had been neither acquired nor publicly tendered for. While this was a small sample, the total number of companies available was also small. Working from this sample, the LDA program formed a discriminant equation which

was:

Z = 14.016X, - .735X — 3.485X, + .463X ̧ where:

X1 = Average expense ratio

X= Stock in portfolio at market/total assets X1 = Investment gain (dividends, interest, appreciation)/portfolio

X, Price of company's common stock/liquidating

value of stock

The variables were calculated for each company on the basis of the average for the period 1965-1967. The reason for using this period was that it preceded the merger period (1968) and allowed for some yearly variation. The procedure indicated that a “Z” value of 4.640 best divides the targets from the non-targets. That is, firms with values greater than 4.640 are potential targets. Using the basic equation, a "Z" can be determined for each company.

An example may help to show the procedure. Let us take Continental Insurance and Hartford Fire.

[blocks in formation]

Using the number 4.64 to separate targets and nontargets we see that Hartford is a target and Continental is not. If we continue this procedure for all companies for which data are available and whose stock is nationally traded we get an array as shown in Table 1. Table 1 indicates that one company, Home, should not have been a target, given its "Z" value, but it was. Chubb & Son is also a non-target and although a merger was announced it has since been abandoned.

The unusually high values for reinsurance companies indicate these companies are excellent candidates for take-over; however, their high scores are associated with high expense ratios. It is doubtful that our technique can be used to predict whether or not they are merger candidates. In order to apply the discriminant function to companies for which liquidation values were unknown, we substituted the average liquidation values for all companies.

An examination of the basic formula brings to light one unexpected result. A high liquidation value relative to market price apparently is not a desirable trait so far as conglomerates are concerned. One would expect conglomerates to be attracted to a situation where an

acquired firm's assets could be used in its own acquisition. But apparently this was not the case. New Developments

Since the mergers of 1968 a new development has occurred in the insurance field-the insurance holding company. An important purpose of this new organization is to ward off the threat of future mergers by using the insurance company assets for expansion into noninsurance areas. This is expected to improve the priceearnings ratio by giving the insurance companies an aura of growth while at the same time allowing them to shift assets from insurance into areas of higher earnings potential. Table 2 shows the firms among the top 20 who have formed holding companies.

The insurance holding company differs from the insurance group (several insurance companies combined) in that its charter is sufficiently general that it is not restricted to insurance. The effect of the holding company structure on the earnings and price-earnings ratio, however, should not be felt immediately. Consequently, it probably will not appreciably change our analysis. On the other hand, the use of the holding company signals

TABLE 1

"Z" NUMBERS FOR SELECTED PROPERTY & LIABILITY INSURANCE COMPANIES,* 1965-1967

[blocks in formation]
[blocks in formation]

In 1968, conglomerates began acquiring large insurance companies and in the process brought great changes to insurance operations. The industry's response was to form insurance holding companies. The present confrontation between insurance companies and conglomerates will undoubtedly continue with the more vulnerable companies having to struggle mightily to stay independent. Several factors have made insurance companies particularly desirable for conglomerates. The most important of these is a ready supply of liquid assets.

This study shows that the most vulnerable companies are those with high operating ratios indicating, in part, high cost of operations. Companies which have large stock portfolios and which earn relatively high returns on their portfolios are less desirable to conglomerates. Finally, conglomerates seem to prefer companies with a a relatively high ratio of liquidation on values to market.

Although there is no clear evidence as to the effect of the present acquisition movement, it seems reasonable to hypothesize that conglomerates will use the acquired insurance companies as a source of funds for other ventures while they make some attempts to correct management deficiencies. Whether they can succeed in the latter seems problematic.

FOOTNOTES

1. See, "Can Fifty Money Managers All Be Wrong?", Forbes, March 1, 1969.

2. The attitude of Continental Corporation Chairman Vie tor Herd would seem to bear this out. In a posunt inter view he explained Continental's three per cent return on investments by explaining: "Our number one objec. tive is liquidity." Ibid.

8. H. Myers and E. W. Forgy, "Development of Numerical Credit Evaluation Systems," Journal of American Statistical Association, Sept. 1963. For a clear and simple explanation of the method, see W. Beranek, Analysis of Financial Decisions, Homewood, Richard D. Irwin, 1963, pp. 327-334.

4. A recent and outstanding example of the use of discriminant analysis in finance can be found in Edward L. Altman, "Financial Ratios, Discriminant Analysis, and the Prediction of Corporate Bankruptcy," Journal of Finance, Sept. 1968. Professor Altman explains much of the statistical background and analysis which has been excluded in our explanation.

5. Linear-discriminant analysis (LDA) is a process designed to determine one linear function for two or more groups of data. The data are composed of an unknown dependent variable (the discriminant = Z) and two or more independent variables. An appropriate form of the equation is:

Z1 = a1X12X21 ± â ̧X3; ± â«X«±â¡Xμ

The independent variables X„ are combine in a linear combination (where j = variable and ico.pany). If the two groups of data are sufficiently dissimilar than the coefficients, which are determined from the original linear fit, they can be used to determine in which group a specific company belongs. The process has many characteristics in common with linear multiple regression analysis. LDA has become more popular with the availability of good computer programs. The analysis in this paper used the Biomedical Computer Program, BMD04M, "Discriminant Analysis for Two Groups."

For the more statistically minded, the following data based on the results may be helpful.

[merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

From the National Underwriter, Jan. 24, 1970

CORPORATE TRENDS-ANNUAL REVIEW: INSURER DIVERSIFICATION
DOMINATES TRENDS IN 1969

By Gaetano Staffa

This brief examination of 1969 financial developments places recent trends in proper perspective, and draws a number of inferences from the statistical data presented.

During the past two years, these developments have been reported in detail in the "Corporate Trends" column that appears monthly in both the life-health and fire-casualty editions of The National Underwriter. Many corporate moves have been dramatic and far-reaching and their causes and effects have become the subject of ively and spirited debate.

The conclusion of the 1968 annual review summed up many of the problems that the industry then faced, and that have since intensified. It said, in part: "Narrowing profit margins, excessive inflation, rising loss ratios, regulatory problems and consumer discontent explain the movement of property-liability insurers toward adopting a more flexible corporate structure and diversifying into related and unrelated areas of finance and industry.

"Excessive inflation also explains the loss of consumer dollars from fixeddollar life and annuity contracts to equity investments which theoretically offer a hedge against inflation. The response from life companies, of course, has been to offer a broader range of financial services, particular y equities, to the pub ic." These observations appear as valid now as they did a year ago during the height of the holding company and merger movements. In 1968, many of the industry's giants began to move decisively in the areas of acquisition and diversification and a host of smaller companies followed their lead.

In years past, the insurance industry has been characterized as unimaginative and lacking in innovative spirit. However, during the past several years, the industry has been the wellspring of innovation, and has attempted to meet its commitment to the underwriting function, while striving to improve its profit position.

WON'T BE LEFT BEHIND

Many insurers have taken long, hard looks at their balance sheets, income statements, markets and competitive positions and have decided that they do not intend to be left behind in the race to provide total corporate and personal financial services for the projected lucrative market of the 1970s.

The property-liability industry is being increasingly pressured by legislative, regulatory and consumer groups to broaden its underwriting capacity in the face of inadequate rates, rising losses and spiraling inflation. If the porpertyliability industry capitulates to this pressure without receiving the necessary rate relief, it will place itself in a perilous situation.

The hostile attitudes of Federal and state governments toward the conglomerate invasion of the insurance and other industries were chiefly responsible for the curtailment of forced takeovers of many vulnerable insurers. Many insurers have used this moratorium to strengthen their positions and initiate expansion programs of their own in order to improve their lackluster earnings record.

MERGERS

Merger announcements during 1969 were somewhat below the 1968 level primarily because of the absence of significant activity among the larger insurers. This was due to warnings from the Justice Department and the Federal Trade Commission that they would carefully scrutinize such combinations and would use the anti-trust laws to de-escalate this activity, if necessary.

39-280 0-71-Pt. 17A-70

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][graphic][merged small][merged small][merged small]

A total of 105 merger announcements were reported in 1969, compared with 135 announcements the prior year. Most of the merger activity was concentrated in the area of smaller-to-medium life insurance companies. A number of merger transactions involving at least one large insurer or insurance holding company during 1969 are listed below:

American General-Fidelity & Deposit

CNA Financial Corp.-Canadian Premier Life 'Continental Casualty-Los Angeles Ins. Co.

Crum & Forster-Constitution Ins. Co.

"Franklin Life-Great International Life "Home Ins. Co.-Federal Life & Casualty John Hancock-Maritime Life

Kemperco-North American Life

U.S.F.&G.-Thomas Jefferson Life

It would appear from the amount of merger and acquisition activity that occurred during the year that only two areas have experienced marked downward trends: Mergers involving larger insurers and mergers between insurance and non-insurance corporations. It seems safe to conclude that if merger activity between these two groups had continued at levels comparable with 1968, total activity would have equaled or exceeded 1968 levels.

NON-INSURERS

The limited number of transactions in which non-insurers acquired interests in insurers during the past year (excluding bank-insurer affiliations) involved the following companies:

1 Continuental Casualty is the principle property-liability subsidiary of CNA Financial Corp. Franklin Life has changed the name of Great International Life of New York to United Franklin Life.

3 Federal Life & Casualty was acquired by American General as part of its acquisition of Channing Financial Corp., and was, in turn, sold to Home.

« PreviousContinue »