Determining value in the life assurance business
While scanning the weekend press, we stumbled upon a Business Times article titled “Assurance co’s on life support.” It provides a snapshot of the long-term insurance industry at 5 December 2008 by comparing the embedded values of six locally listed assurance companies to their share prices. And it’s not a pretty picture. The article reminds us that many JSE-listed companies trade at a discount to their book value at present – and that “the JSE’s life insurance board is the most extreme example, with five of its six listings going for less than their liquidation value.”
A quick look at Old Mutual confirms the group’s embedded value (reported at 30 June 2008) is R22.50 per share against its latest close of R6.90 per share, rating the counter at an apparently cheap price-to-book ratio of 30.7%. What has led to such a massive discrepancy between Old Mutual’s calculated value (embedded value) and its price?
What is embedded value?
Part of the difference can be ascribed to the appalling market conditions over the last six to 12 months and part to the management blunders at head office. But to get a better understanding of the problem we need to unpack the concept of embedded value.
Valuing a life assurance company is more of an art than a perfect science. We find the reason for this in the nature of the income earned by insurance companies. It’s not like selling fast food where your costs and revenues are offset at the transaction date and your gross profit margin is easily identifiable. Instead you’re selling long-dated policies that generate annuity income for years to come. During the first two years of the policy the life company reports a significant loss as marketing costs, up-front commissions and other policy administration expenses dwarf the premium earned. But net income rises steadily in ensuing years as costs diminish and premium income and the value of assets under management increase. How much is each policy worth?
Life insurance companies must estimate the present value of the future profit from all the policies on their book. It’s a complex actuarial exercise that involves estimates on the return on investment, mortality rates and policy lapses and surrenders over a number of years. The estimation of future profits is the cornerstone of embedded value… A tool used by insurers to give their shareholders an idea of what their underlying business is worth.
Too many assumptions could spoil the valuation party
Without getting too technical we’ll take quick look at some of the embedded value terminology that’s commonly used by the life insurance giants. We need to be able to distinguish between Embedded Value Life Insurance (EVLI) and Total Embedded Value (TEV). Though these calculations will vary slightly from company to company (and country to country, the basic premise is the same.
ELVI is the estimated present value of future profits expected over time from existing life insurance policies – an estimate of the worth of the insurance business only. The insurance company looks at all the policies on its book and calculates (actuarially) what all that income is worth today! It then adds the assets it has to set aside to meet solvency requirements and any undistributed profits before subtracting the cost of holding this solvency capital... As we mentioned earlier, a number of assumptions enter this calculation, including mortality rates, lapses and surrenders etc. This hasn’t deterred life companies and industry analysts from relying on the technique to determine the value of insurance businesses. TEV is simply a measure of the insurer’s total underlying worth, calculated by adding the value of its non-life activities to ELVI, and subtracting any debt.
The life insurance company includes its calculation method with the results statement. Sanlam, for example, defines Group Equity Value as follows: The embedded value of covered business, being the life insurance businesses of the group, which comprises the required capital supporting these operations and the net present value of their in-force books of business (VIF); the fair value of other Group operations, which includes the investment management, capital markets, credit, short-term insurance and the non-covered wealth management operations of the Group; and the fair value of discretionary and other capital.
Embedded value and the king of value investing
At first glance this value estimation exercise is quite similar to that used by the king of value investing, Warren Buffett. He says that value is nothing more than the present value of all future earnings. But the problem lies in how accurately the future earnings can be predicted. If the ‘variables’ used in the calculation are too far off the mark the consequences could be drastic.
We can demonstrate this phenomenon by looking at Sanlam’s trading update for the 10 months to October 2008. The group reduced its embedded value to R20.50 per share (from R22.50 per share at June). At the latest price of R16.60 per share Sanlam trades on a price-to-book ratio of 82%. And that suggests the group’s embedded value calculation is shifting to reflect the massive reduction in assets under management. When Old Mutual publishes its next results we expect its embedded value to re-rate sharply downward too. In other words, the embedded value will drop to reflect poor market conditions rather than the share price rising to narrow the ‘apparent’ valuation gap!
Editor’s thoughts:
There are many investment analysts who suggest buying shares when they offer a discount to book value. It works for value investors who want to buy ‘clear-cut’ industrial companies at a good discount. But with insurance companies the earnings streams can be difficult to determine. And you have to trust the insurer’s valuation of its own business. Would you buy Old Mutual (or one of the other life insurers) based on their embedded value – or do you trust the market to value these shares? Add your comments below, or send them to [email protected]
