In the first part of this article, we talked about some of the qualitative criteria investors should use to evaluate insurance companies. Now let's get down to some of the quantitative aspects of insurance valuation -- the hard numbers that help indicate insurers' overall health.

Tangible book value
Tangible book value (TBV) is the value of an insurer's assets minus its liabilities, excluding intangibles and goodwill. This could also be a proxy for liquidation value -- an estimate of what the insurance company would be worth if the company closed its doors, paid out claims, and returned excess capital to shareholders. (Hence the term "shareholders' equity.") Obviously, you'd want to pay a lower multiple of tangible book value to increase your margin of safety, assuming you can't get shares at a discount to TBV.

The amount of float
A big part of a bank's value is its amount of core deposits, because a bank can easily earn more from a depositor's money than it pays out in interest. And that doesn't even include the myriad fees it can reap, including overdraft and service charges on those deposits.

Similarly, an insurer pockets any investment income it earns from policyholder float. So the amount of float an insurer generates plays a big role in determining its worth. To calculate float, add together loss and loss adjustment expenses with unearned premiums, then subtract premiums receivable and deferred policy acquisition costs. Using the latest quarterly figures, I calculate Income Investor recommendation Mercury General's (NYSE:MCY) float at roughly $1.5 billion, or about half the company's nearly $3 billion market cap. Meanwhile, I calculate Progressive's (NYSE:PGR) float at about $6.8 billion, or nearly half the company's $14 billion market cap. Clearly, Progressive's larger amount of float makes the company more valuable.

Cost of float
The cost of float varies inversely with its value. The more you have to pay for it, the less it's worth. If I opened up my own bank, and offered depositors a 10% interest rate, I could probably attract deposits very quickly -- until I went broke trying to pay out that interest.

Insurers that generate low-cost float are much, much more valuable than their higher-cost peers. The best way to gauge the cost of float would be to look at a company's long-term combined ratio versus its competitors. If the insurer consistently earns combined ratios of less than 100%, that's very good. It means that the insurer earns an underwriting profit, and has a positive cost of capital. Think of it as a bank whose depositors pay the bank to hold their deposits -- and allow the bank to pocket any investment income it earns.

The two most sustainable ways to generate low-cost float are exercising underwriting discipline and maintaining low overhead and expense ratios. Progressive, and Berkshire's (NYSE:BRK-A) (NYSE:BRK-B) GEICO, particularly fit the bill in those regards. Last year, they posted 13.3% and 11.9% underwriting profit margins, respectively, compared to the auto insurance industry's average of 7%.

Float growth
Float growth, as long as it comes at a reasonable price, increases an insurer's value. For example, through organic growth and acquisitions, Berkshire increased its float from about $20 million in 1967 to $50.9 billion at the end of 2006. That represents a roughly 21%-22% annual increase, almost identical to Berkshire's share price appreciation and the annual growth in its per-share book value.

Investment income
The last piece of the puzzle? How profitably the insurer actually uses its float. Some insurers tend to put all their money in low-yielding investment-grade fixed-income securities. There's nothing wrong with that, but it'd be nice if insurers could earn higher risk-adjusted returns, the way Berkshire, Markel (NYSE:MKL), and White Mountains (NYSE:WTM) do. Flagstone Re (NYSE:FSR) and Greenlight Re (NYSE:GLRE) are just getting started, but they follow the same mold. I'd also advise Fools to look at Enstar Group (NYSE:ESGR).

Add these factors together -- tangible book value; the size, cost, and expected growth of the float; and the expected returns from that float -- and you should be able to estimate an insurer's true worth.

Related Foolishness: