Price on Value
June, 2000

Price Trading Bands:

Adding Cream to Your Quality Stocks

John Price,  Ph.D.

Got an itchy buy/sell finger? Can't wait to press the return key to send your trading orders rushing through a T1 line to your favorite online broker? If so, you probably find long-term, Buffett-style investing too boring. Take heart. Here is a way to be loyal to your fundamentalist principles while satisfying your desire for more action. It might also add a little cream to your profits. 

We start at the same place that I have continually been recommending—quality companies with a high return on equity, not too much debt, and smooth earnings growth. We should also look for companies that have a strong economic moat or sustainable competitive advantage. (See my article "The Three 'Little' Words of Investing." Click here.)

Here are a few examples that meet these criteria.

The sales growth and the earnings per share (EPS) growth is the annual percentage growth over five years. STAEGR^TM is a measure of the stability of the growth in earnings over the past ten years. A measure of 100 percent indicates that the growth is constant over all this period. More details are given in my article "Earnings Forecasts Made Easy." Click here. The stocks were chosen to have high values in these measures.

The final column is the total annualized return over ten years assuming that all the dividends were reinvested. All the returns are excellent. Even the lowest figure of 18.9 percent means that $10,000 converts to $56,470 over ten years. These figures are just reported as they stand meaning that the stocks were not chosen to have high historical returns. Total returns were not part of the screening process.

Of course, trying to determine whether these returns will continue in the future is the central goal of long-term investing. The fundamentals and high STAEGR point to earnings growth continuing to be strong in the future, but you can still pay too much for even the best of companies.

When you examine the histories of stocks like those above you see an overall upward trend superimposed by many troughs and peaks. The problem is how to pick the troughs—to buy the stocks—and the peaks—to sell the stocks. On closer examination you often see another feature, namely that the price to earnings ratio seems to have a fairly constant range over time. For example, as I described in last month's article, the p/e for Gillette is mostly between 30 and 60 whereas the p/e for MGIC is between 10 and 20. I call these limits the p/e band for a given company.

The idea is to make use of this band to determine when a stock is underpriced or overpriced.

Even though the p/e ratio for Gillette usually lies within fixed limits, because its earnings are fairly variable, it is not so easy to make use of this fact. Things are much better with MGIC. With great regularity, the earnings per share for MGIC have grown on average by about 25 percent per year. The forecast growth rate is not as high, but it is still in the double digits.

Price Trading Band

Here is the trading strategy using this information on low and high p/e ratios. Look at the lowest and highest p/e ratios in the previous year. Suppose they are 9 and 15. Use these ratios to form an inner interval of ratios. For example, you could take the middle third of the interval formed by the lowest and highest p/e ratios. In this case the inner interval would be 11 to 13. Now wait until the stock price causes the p/e ratio to drop below 11. This is the buy signal. The sell signal is when the stock price pushes the p/e ratio above 13.

In other words, the interval formed by the lowest and highest p/e ratios for the previous year are used to form what I call a price trading band for the current year. If the price drops below this band, buy the stock. If it rises above it, then sell.

Before implementing this strategy, there are three important provisos: (1) The companies must have had earnings growth in the past that were both consistent and high and that it is likely that these features will continue into the future. (2) The low and high p/e ratios for each year were fairly constant. (3) You would like to add this stock to your long-term portfolio provided you could buy it at a rational price.

There are two main things that can go wrong. (This is apart from the fundamentals changing which is always a possibility when holding stock over a long period.) After buying the stock, the market changes its mind about the company's "natural" p/e range and it shifts to a lower level. If this happens, the sell signal may not get triggered. But if you were already buying the stock as a long-term investment, then this just means that you are left holding a stock that you would have bought anyway.

On the other side, the p/e range may shift upwards from year to year so that you don't find an entry point.

To test the strategy, I considered MGIC from September 1990 to March 2000 using monthly data. Denote the lowest and highest p/e ratios for a given year by x and y. Denote the difference between x and y by w, that is, w = y - x. I formed the price trading band for the next year by considering prices that caused the p/e ratio to be at least x + 0.4 w and at most x + 0.8 w.

For the example of the middle third considered above, the p/e ratios would be x + w/3 and x + 2w/3. I wanted to increase the time that the strategy would indicate that the stock should be held. For this reason I chose an interval that was both higher and wider than the middle third interval.

The following chart illustrates the times at which the strategy gave buy signals and sell signals.

It might look as if the method missed out on a lot of the bull runs for MGIC. After all, just a simple buy and hold strategy for MGIC would have given an annualized return of 20.6 percent. It would have been nice, for instance, to capture more of the profits in the price spike in mid 1998. But who would have guessed that the price was going to run so high that the p/e ratio would rise from around 12-15 to over 27.

The point is, however, that the method of the price trading band only tied up your capital for a period of 40 months. Over the time of the study, an investment of $10,000 in this strategy would have grown to $34,000. Using the fact that this was achieved in 40 months of actual investing gives an annualized return of 44.3 percent.

For those with an itchy buy/sell finger, using this method with a number of stocks, say five or more, may provide you with the trading action, and profits, you are looking for. Used as an adjunct to long-term, Buffett-style investing, this method may, as I said, add some cream to your profits.
_________________________________

STAEGR is a trademark of John Price.

Do you tend to sell your losers or hang on until they get back to the price you paid for them. If you do the latter, then next month's article "Get-Evenitis and Other Investor Maladies" might help you to rethink the way you invest.