The PEG ratio: Does it work?
PEG has become a popular ratio in financial analysis. Yahoo has even made the PEG ratio available in their "Key Statistics" section of companies, as can been seen on a quote I pulled up for Yahoo.
Most frequent arguments around the PEG are based on theoretical arguments or references to famous people in the media that recommend it like Jim Cramer on his Popular TV show Mad Money. What I feel is lacking in the debate of PEG are hard numbers. So let's take a look at the hard numbers behind the debate. I have taken 20 randomly chosen stocks from the S & P 500 that produced a growing net income for the years 2009-2011. I've decided to go with the prior 2 years of returns because there is no reason to believe that the future returns, or projected returns, should differ greatly. The other reason I decided to use this proxy for future growth instead of actual projected growth is the difficulty of getting past projections for future growth.
Above is the basic formula for PEG. It's an elegant formula where you get P/E per unit of growth. Logic would dictate that getting a cheaper price per unit of growth is better than getting a higher price for the same growth in earnings. Over time the companies that produce greater earnings growth should have the higher price appreciation. "PEG ratio of 1 is sometimes said to represent a fair trade-off between the values of cost and the values of growth, indicating that a stock is reasonably valued given the expected growth. A crude analysis suggests that companies with PEG values between 0 to 1 may provide higher returns" (Source)
Let us get right to the results of the analysis.
The statistical regression of PEG vs. returns do not look good for the PEG proponents. The first major statistic to look at is the correlation, which varies from 1 to -1. A correlation of 1 indicates that two variables move exactly in the same direction, for example, when the S & P 500 gaines 10% a stock that has a correlation of 1 would also gain 10%. And when a theoretical stock has a -1 correlation with the market and the market gained 10%, the stock should be down 10%. A Correlation of 0 means there is no relationship between the two. The correlation between PEG and returns is only 0.158. Furthermore the R-squared, which tells us how much one variable influence the other, is only 2.5%. These two metrics are very low for the 20 randomly chosen stocks and would indicate almost no meaningful relationship.
One thing to keep in mind whenever we talk about the PEG ratio is the quality of the earnings. The quality of the earnings have been shown to predict future performance on average. If the equality of earnings of some of these companies I've analyzed are poor, their earnings maybe inflated; making the P/E ratio smaller and hence the PEG ratio smaller. These smaller PEG ratios could give false signals of good investment opportunities.
Another major shortcoming of using a simplistic tool, like the PEG ratio,is that earnings growth is not the only metric the market cares about. Revenue growth, margin growth, cash flow from operations, Free cash flow to equity, dividends, debt, debt ratings and many other factors are also important to valuing a company.
Overall, from these 20 companies, the ability of the PEG ratio to predict future performance is highly suspect.