I was in a meeting today with a great company and with great people. We were discussing financial measures for a life insurer, including Embedded Value, when one of the well respected actuaries asks what the Return on Equity is (rather than Return on Embedded Value).

Now there are many people who don’t like Embedded Values, and they certainly have some good reasons for being wary of it. Embedded Values are a measure of shareholder value in a life insurer that adds the shareholder valued embedded in prudent regulatory reserves to the net worth of the company with an adjustment for the cost of holding capital. It is needed because current accounting measurement of insurance liabilities is also typically overly prudent and thus book value on an accounting basis is a really bad measure of shareholder value in a life insurer.

Embedded Values are flawed due to imperfect allowances for risk, lack of comparability across countries and even sometimes companies and a methodology that can make value magically appear out of taking on mismatch risk.  Worst of all, in a practical sense, is that life insurers typically trade at a discount to Embedded Value in the market (Embedded Value is greater than Market Value) even though theory suggests Embedded Value should be less than Market Value since we haven’t included any allowance for future new business/goodwill/franchise value.

None of this was the basis for the preference for Return on Equity over Embedded Value.

You can’t eat Embedded Value.

That was the criticism. You can’t eat Embedded Value. This is true.  Of course, you can’t eat Equity either.  You can’t eat Book Value.  You can’t eat earnings.  While we’re at it, you can’t actually even eat cash, although you are reasonably likely to be able to potatoes directly with the cash (as in a dividend) than with most other measures.

So while you can’t eat Embedded Value, you just as well can’t eat Book Value.  In the life insurance space, it’s really hard to get a handle on “tangible” asset value.  Our liabilities will always be based on an estimate of future cashflows. Embedded Value is in its most basic form just a different value of different future cashflows discounted at a different rate.  It is no less (or more) tangible than Book Value.  Return on Embedded Value is no less tangible than Return on Equity since they both depend on uncertain estimates of future cashflows and experience.

The difference is, we know that Book Value is categorically, definitely undervalued.  The degree of undervaluation is different from one company to another depending on the level of prudence included in the liabilities. In South Africa, it’s also fundamentally undervalued because the liabilities ignore the economic value arising out of optional (but reasonably predictable) future premium increases.

This means that Return on Equity is an overstated measure of return since the measure of Equity is understated because the liabilities are overstated.  Worse, we can’t compare it from one insurer to another since their liability valuations will be different.

So while you can’t eat Embedded Value, you will make more value for your shareholders basing decisions on Embedded Value than on Book Value.

Interestingly, new accounting developments (IFRS4 Phase 2) and regulatory developments (Solvency Assessment and Management, based on Solvency II from Europe) will completely change the way we measure financial performance of insurers. And no, I still won’t be using Return on Equity.