Is P/E ratio a useful stock valuation measure?
Posted on: July 9, 2008
Warren Buffet said: “Price is what you pay and value is what you get.”
Wall Street uses the price-to-earning ratio, or the P/E ratio in short, to determine whether one gets what one pays for when buying a stock. Is this ratio just a myth? Or is it a useful valuation measure?
To answer this question, I examined the whole stock market data for the past 50 years from 1958 to 2007. For each year, I separated stocks into three portfolios: the top 30% P/E portfolio, the middle 40% P/E portfolio and the bottom 30% P/E portfolio. (Stocks with negative earnings are all in the top 30% P/E portfolio.)
If I had invested $1 in each of the three portfolios at the beginning of 1958, by the end of 2007, the top 30% P/E portfolio would have grown to $91; the middle 40% P/E portfolio would have grown to $322 and the bottom 30% P/E portfolio would have grown to $1698! (The chart below shows the growth of $1 in the three different portfolios in logarithmic scale.)
In fact, in the past 5 decades, there was not a single decade in which the bottom 30% P/E portfolio did not outperformed the top 30% P/E portfolio. The decade spanning 1968 to 1977 was especially eventful: two global recessions, the Arab-Israeli war and the Arab oil embargo. The returns of the three portfolios in that decade are as follows:
Top 30% P/E portfolio: 31%
Middle 40% P/E portfolio: 61%
Bottom 30% P/E portfolio: 137%
It is safe to conclude that the P/E ratio is a very useful valuation measure for long-term stock investment. The lower the P/E ratio, the higher is the expected long-term return. That does not mean that low P/E stocks outperform every year though. In the last 50 years, there are 12 years in which the top 30% P/E portfolio outperformed the bottom 30% P/E portfolio. Take 2007 for example, the top 30% P/E portfolio outperformed the bottom 30% portfolio by more than 13%.
Get my white paper: The Informed Investor: 5 Key Concepts for Financial Success.
Share this:
8 Responses to "Is P/E ratio a useful stock valuation measure?"
March 25, 2009 at 11:39 am
Okonkwo,
The answer is yes. Regardless of what valuation measure you use, P/E, P/B, P/S or P/C, the result is the same: value outperforms growth over the long run but not every year.
Note: P/S is price to sales and P/C is price to cashflow.
Caution: One commonly used valuation measure that does not work is PEG, since the “G” part (growth rate) is subjective. and tend to reflect investors’ illusion more than anything else.
July 20, 2009 at 4:20 pm
Interesting analysis. Thanks. However,I would question whether timing of the markets play any role in the analysis. Your ‘timing’ model basically states that if you bought every January 2nd (or whatever opening day), and sold on the last day of December, that your PE model would be this way. That theory is not how people usually buy stocks. I’m not questioning the notion that P/E is useful, but people obviously panicked last year and early this year and bailed out, after probably buying in during 2005/2006/2007. Would P/E have made a difference there? If a person bought this group of stocks at the highs in those years, and sold in the crash, or after the .com bubble burst, what would the split in P/E ratios be? Would it have affected the rate of returns for times when people actually sell heavily?
July 20, 2009 at 10:16 pm
Alberg,
I can tell you the result is robust to “timing”. In other words, if I pick the rebalance day on any day in a year, the result will be the same.
July 22, 2009 at 2:08 pm
I’ve done a lot of analysis on the difference between fundamental and technical analysis, including market timing (but not looking at day trading timing, for instance). Here’s an example… maybe a sample of one is not valid, but take Microsoft as one that everyone knows. MSFT’s PE ratio has been as high as 44 and as low as 8.65 over the last 5 years. However, it’s stock has been terrible in relationship to shareholder value, when compared with Apple over that period of time, and I don’t think that buying when the PE ratio was low was any better returns than when high. Only in the last three months has the stock been a ‘value’ and it took a major crash for that to happen. Thoughts?
July 25, 2009 at 8:27 am
Alberg,
I’ve never used P/E alone in stock picking. A low P/E stock may not have better returns than a high P/E stock. However, a low P/E portfolio will more likely (but not always) outperforms a high P/E portfolio. Because of that, I have a value biased in my portfolio composition. For example, I have small cap value instead of small cap growth.
March 30, 2014 at 2:58 pm
30% of the index would mean over 1000 stock. It is not feasible hold that many stocks in a portfolio. Can you perform the same strategy for the bottom 50 or 100 stocks so that we have a practical portfolio?
April 3, 2014 at 10:44 pm
There is an index ETF that does that – check out WisdomTree’s website for the symbol.
The Investment Scientist 
March 24, 2009 at 10:34 pm
I must say that I really enjoy reading your site.But this particuar article,is that to say that value stocks outperforms growth stocks