SEPTEMBER 2001
Cover Story
Analyzing Insurance Companies
by Scott D. Horsburgh, CFA, Contributing Editor

Editor's note: Our thanks to Scott D. Horsburgh, CFA, for contributing this article on analyzing insurance companies. Scott is a research analyst and portfolio manager with Seger-Elvekrog Inc., Bloomfield Hills, Mich. and a contributing editor of Better Investing. As always with BITS editorial features, no investment recommendation is intended. Emphasis is totally on learning.

One of my primary areas of focus as an investment analyst is the financial services industry. This is the first of two parts focusing on the challenges of analyzing financial stocks: part one will focus on insurance companies and part two will look at banks.

Financial stocks present challenges for investors because their accounting is much different from a traditional manufacturing company. One of the first pieces of information a prospective investor would want to know about any company is, how is sales growth? That can be a very easy question to answer for a manufacturing company, but what are the sales of a financial services company?

Value Line uses "earned premiums" as a proxy for sales of an insurance company. However, much of an insurer's revenue also comes from investing the premiums it has collected. Most property and casualty insurance companies actually pay out more in claims and expenses than the insurance premiums they take in. Therefore, they would actually be running at a loss if it weren't for investment income. That's why I classify the revenues of an insurance company as "net premiums earned" plus "net investment income" plus other sources of revenue such as fee income and realized capital gains/ losses. For a large insurance company like American International Group (AIG), adding together all these items can be very difficult because it has multiple product lines. The company does, however, provide a total revenue figure in its press release of quarterly earnings.

Another challenge facing insurance company investors is the calculation of profits. Each insurance company most definitely provides a to-the-penny calculation of profits but these figures are in many ways just estimates. A property and casualty company can only estimate the claims its customers will make. Frequently, it may take some time for storm damages or thefts to be reported, or the damages may be more severe than previously thought. Property claims are the easy ones. The hard ones are "casualty" claims such as lawsuits over defective products that take years to litigate. Yet, the insurer is required to estimate the losses and provide for them in its income statement and balance sheet.

Knowing that so much of an insurance company's income depends on estimates, how can one be sure that the reported profits indeed are real? The best way is to look at the insurance company's Form 10-K that is filed with the Securities and Exchange Commission (SEC) no later than 90 days after the close of its fiscal year. Each property and casualty insurance company must report its "Analysis of Consolidated Capital Net Losses and Loss Expense Reserve Development." That is the long way of saying that it needs to show how much it originally set aside for losses and loss adjustment expenses each year, how much it subsequently paid out and how it re-estimated those losses years later. These figures have to go back 10 years and as an investor one can see how reliable past estimates have been.

AIG actually has one of the better loss development histories (see table below). The company had significant loss development deficiencies in the years 1990-93, as the accompanying table shows. (Ed. Note: Download the PDF version of the article to retrieve a legible table showing the loss provision historical profile.) The company simply did not set enough aside for future losses. The numbers improved each year and basically broke even in 1994-95.

AIG has reported "redundant" reserves, that is to say that it actually set aside too much to cover losses, for the years 1995-1999. Note that the company's accuracy has not only gotten better, but the numbers have been more favorable indicating that profits in recent years were actually understated because the company set aside too much for reserves. I'd rather have a company set aside too much than too little. There are no hard and fast rules as to what adequate loss development looks like, but it is reassuring to see some consistency and also to see positive numbers going back as many years as possible.

Another way of measuring the effectiveness of a property and casualty insurance company is its "combined ratio." This is a complicated calculation that adds together the insurer's losses plus its expenses to produce new policies. A superior insurance company is one that produces an "underwriting profit." This is a company whose combined ratio is less than 100 and indicates that its insurance policies are producing a profit in addition to generating investment income. AIG's first quarter combined ratio was 95.89.

Most companies run a negative combined ratio, which means that they are using investment income to subsidize insurance losses. In AIG's case, it generated a profit of $256.4 million on its insurance policies in addition to investment income. This underwriting profit constituted over 11 percent of AIG's first quarter profits. A positive combined ratio or at least one that is in the very low 100s is a sign of a well-run property and casualty insurance company. Life insurance companies do not have such a measure.

Another issue that is unique to insurance companies, particularly life insurance companies, is known as "deferred policy acquisition costs" (DPAC). AIG does not separately break out deferred acquisition costs on its income statement but many other insurers do. Life insurance companies in particular must pay significant up-front premiums to the agents who produce the business. Commissions and other up-front expenses are usually higher than the premiums paid by the customer in the first year.

Life insurers run at a cash flow loss during the first year but then make the money up over time. It is customary to add these commissions together with other policy preparation expenses and then amortize (recognize as expense) these amounts over the expected life of the policy. Like with loss reserves, this introduces a higher degree of estimation than one would see in a manufacturing company. For example, these deferred costs made up 28 percent of AFLAC's first quarter profits and the cumulative value of deferred expenses that are on AFLAC's books constitute 68 percent of shareholders' capital. It is very rare that an insurance company writes off the value of some of these deferred costs but if a company does this to any significant degree, it is a tremendous warning sign to stay away from such a company.

Because of the need to maintain large investment portfolios to pay off losses that have already happened but have not yet been settled, and the need to pay large commissions up front, the insurance industry tends to have a fairly low return on equity when compared to other industries. AIG's return on equity has been running in the 14-15 percent range over the past few years and this is actually above its previous trends. The Stock Selection Guide encourages investors to look at trends rather than simply the raw numbers. It can also be helpful for investors to compare return on equity and profit margins of one insurer to those of other insurers.

Another factor that is unique to the property and casualty industry is its own special cycle that does not relate to overall economic cycles. This "PC cycle" tends to be a long cycle in which high insurance losses drive companies either out of business or to not take on certain risks. The loss of capital in the industry causes prices to rise and profits to expand. This higher profitability attracts additional capital to the industry. The higher capital causes increased competition and falling prices, leading to the next cycle of lower profitability. We may be emerging from a long cycle of low returns and higher insurance losses because more companies have been unwilling to take on unprofitable business in recent years. It is possible that this could represent the long awaited "turn" in the PC cycle.

The last major factor that is unique about financial companies is the relationship between interest rates and the stock prices of financial companies. Financial companies are considered to be "interest-rate sensitive." Lower interest rates are perceived to be beneficial to insurance companies while higher interest rates are thought to hurt. For insurance companies, changes in interest rates can actually be a double-edged sword. Insurers tend to have large portfolios of fixed income investments. The value of these investments falls when interest rates rise, yet the opportunity to invest new money in higher yielding investments can make an insurer's interest income rise when rates go up. However, the impact on their balance sheet is much more severe than the modest improvement in the income statement and so insurance stocks tend to go down when the expectation is for interest rates to go up. The opposite happens when the outlook is for interest rates to fall.

One of the most difficult things to do when investing is to predict interest rates. Investors should not even attempt to do it because it is a waste of time. A good insurance company will still be a good insurance company regardless of interest rates and one must simply stay with the good ones through good times and bad.

While we are on the subject of long-term investing, be sure to be particularly careful when projecting higher earnings growth for an insurance company. The lower returns on capital in this industry keep long-term growth rates to moderate levels. Even though AIG is one of the most effective companies in the world, its long-term growth rate is probably no more than 12-13 percent. P/E ratios in the industry tend to be fairly low because of the limited potential for huge growth. AIG's P/E ratios tend to be very high because it is a blue chip company with consistent, above-average growth.

Although AIG has generated annualized earnings growth of 14.6 percent over the past 10 years, other factors suggest that it may be appropriate to select a lower growth rate for the future. Revenue growth has averaged 11 percent annually and the higher earnings growth reported by AIG comes from rising profit margins. Earnings can't rise much faster than revenues forever and so it is reasonable to select an earnings growth rate that approximates the growth in revenues. Also, return on equity (ROE) has been in the 13-14 percent range and long-term growth shouldn't exceed ROE unless one forecasts ROE to rise. AIG's ROE is already among the highest in the industry.

Financial services companies have traditionally traded at very low P/Es when compared to non-financial companies with similar growth rates. AIG typically features one of the highest P/Es in the industry because of its reputation for consistent growth. When selecting P/Es, one can't just blindly adopt the historical data. AIG's P/E ratios rose along with the market P/Es throughout the late 1990s. The P/E of 34.5 in 1999 is probably the maximum reasonable P/E at almost three times the expected growth rate. I excluded the 42.7 P/E of 2000 because it was driven to irrational heights by investors' desire to find companies able to grow in this difficult environment of 2000-2001. Good luck with your research efforts.

Scott Horsburgh is president as well as research analyst and portfolio manager with Seger-Elvekrog Inc. (www.seger-elvekrog.com), Bloomfield Hills, Mich.