Friday, December 22, 2006

Two value strategies

I believe that all good investing involves buying something below it real value. Up until now I have been concentrating on what may be called a Buffett strategy. This is essentially investing in good business at prices below their value. For example you can buy companies with high return on equity (ROE) and low P/E. You can make guesses about their probable future growth and so estimate their fair value P/E and buy if the market P/E is below that.

There is another way where you buy companies at low price to book value, P/B, regardless of whether they are good or bad. These strategies are connected because of the identity

P/B = P/E * ROE

This is true simply because ROE = E/B. So you can't really have low P/E , low P/B AND high ROE. You need to choose whether you want to go with high ROE and low P/E or low ROE and low P/B. With good companies like Coke or Walmart your going to get stable and predictable earnings, high ROE and high P/B. Generally you care about getting them at low P/E.

The other strategy is buying companies with low P/B. In the extreme you want to get things where the Current Assets - Liabilities is less than the market cap. This is tangible book value and is what you would get (hopefully) if the company was liquidated right away.

The main difference is that for the growth companies, you are valuing the companies based on the earnings that will come in the future. For the low P/B companies you are simply looking at what is there already. In some ways these low P/B companies are a more conservative bet. You don't need to project far in the future. You just look at what is there. The trick for these is to figure out whether your actually going to be able to extract that value.

Things can go wrong. The management can piss away stockholder value by trying to save the company with hair-brained schemes or outright fraud. When tangible book value is substantially more than market cap, the stockholders would usually be happy if the company simply liquidated and paid out a final dividend. However often this doesn't happen. If the CEO is getting a high salary he is often happy bleeding the company for years and simply collecting salary. He need to look for management with interests alligned with shareholders. You also need to look out for other liabilities that may arise such as pending lawsuits or other fishy things on the balance sheet or 10Qs.

Combining these two strategies is probably a good idea since they should be uncorrelated. In fact the low P/B stocks themselves should have practically no correlation and so is good for diversification. The diversification thing is more than just reducing volatility. Value investors shouldn't fear that. It has more to do with the fact that some of these companies will blow up and could cost you everything you put in. However most will return above book value. Some of these will more than triple in a few years and should make up for the occasional loss. It is not short term volatility that matters but rather volatility of long term portfolio growth. If you lose it all your done. This relates to the Kelly formula from gambling theory. Never bet your whole payroll on one bet. Figure out the right amount to bet on each based in your edge and the odds your getting.

Also with these low P/B companies you have low expectations of earnings growth. Generally you just want them to quit losing money, stablize their business and generate postive earnings with below average ROE. Then they return to above book value and you sell them. This could be considered more conservative than expecting an already good company to get even better and grow even faster. It is these low expectations that make them a reasonably good investment.