Yesterday I heard a Wall Street guru on CNBC say “The S&P500 P/E is now 20, so the market is overvalued by historical standards”. I think this argument is misguided as market valuation, in my opinion, is not an absolute. Rather, equities valuation should be a relative exercise which may or may not make sense based on the opportunity cost of an investment in the S&P500 at any given level.
My hypothesis is that one cannot say for certainty whether a P/E of 15, 20, or 25 indicates an overvaluation in the market without considering what the expected return on alternative investment options are. For example, a P/E of 20, which implies an earnings yield of 5%, may be attractive if all other investment alternatives yield less than 5%. Therefore, there must be a historical convergence between the prices investors are willing to pay for equities and other alternative investments such that the ratio between the equity yield of the S&P500 and the alternative investment of choice is 1. This I call the “Market Valuation Parity”.
Consider, for instance, the historical yield of the 10Y treasuries, which is near 7%. This is exactly what the earnings yield of the S&P500 historical P/E of 15 would imply. Therefore, it may not be a coincidence that the average P/E ratio for equities seems to be around 15. Whether by coincidence or not, the historical average of the ratio between S&P500 earnings yield and the yield offered by 10Y treasuries is 1. Therefore, it seems to me that this ratio could be a better indication of whether equities are an attractive asset class. Although we have chosen the 10Y treasuries, it would make sense to perform the exercise with other asset classes with similar liquidity. But for now, the chart of the ratio between S&P500 earning yield and the 10 year treasuries is shown below:
When the parity is less than 1, it is said that equities is the better option, or has an “attractive valuation”. As the chart shows, March of 2009 was the most attractive time for equities since 1974. Additionally, the parity indicates that although the market is trading at a P/E of 20, it is still attractive based on what other investment options (in this case the 10Y treasury) offer at the moment. The chart also shows the excess that characterized the late 90s, when the parity reached 2.5 just before the collapse of the internet bubble.
A clear flaw is that the chart does not indicate any over-valuation before the 2008 crash, although it may also be indicative of artificially high earnings in the P/E ratio. Furthermore, the relationship that as long as the Federal Reserve keep rates low, the market earnings yield will be above the historical mean of 15. However, that does not automatically mean that the market is overvalued, as it may still offer a more attractive return than other investment alternatives.
The same exercise could be done using the average 10-year earnings yield (courtesy of Matt DeLano), as shown below:
Jeff // June 17, 2010 at 7:19 am |
Another clear flaw — your chart shows bond preference during the great bull stock market from 1980 to 2000.
Matt // June 17, 2010 at 8:38 am |
This reasoning makes MUCH more sense than just saying the average PE ratio over the last 100 years is 16, so anything above is overvalued and anything below is undervalued. Thank you so much for this article.
widemoat // June 17, 2010 at 10:27 am |
Just to play devil’s advocate, Jeff… if you had bought LT US Treasuries from 1980-1982 and reinvested the coupons, you would have done very well. And you would have had the option of taking capital gains along the way.
So bond preference in 1980 doesn’t look foolish. If you could have market timed the stock rise from 1980-2000, sure, that would have been better. But that was an extremely anomalous period for equities, in my view.
ValueHuntr // June 17, 2010 at 10:53 am |
Also notice that by definition, half the companies in the market will have a P/E less than average, so it is possible to invest on those. So a ratio greater than 1 doesn’t mean not to invest, it just means that the market as a whole is overvalued and that you may want to be more conservative than usual.
sam // June 17, 2010 at 4:28 pm |
I’ve been pondering the “treasury bubble” as Warren called it in his feb ’09 2008 letter to shareholders…(which has gotten worse since)
It seems to me that if he is right, and when yields start to climb, and everyone is losing money in “safe” and “risk-free” bonds, and selling them hand over fist, that they won’t likely be buying stocks during that massive unwind unless stocks are in an obvious bull or bubble market
Morgan // June 18, 2010 at 6:05 pm |
Recently, there was another study similar to this done on Greenback.com(http://greenbackd.com/2010/04/09/grahams-pe10-ratio/).
In this they use Grahams idea to compare the price and 10 year average of earnings. They used data from Robert Shillers website. Both your chart and the Greenbackd one conclude the same essentially the same thing; the market is overpriced, but not extremely so. Both are interesting to look at and may provide one more barometer of general market valuation.