Jan-26 Jan-29 Jan-32 Jan-35 Jan-38 Jan-41 Jan-44 Jan-47 Jan-50 Jan-53 Jan-56 Jan-59 Jan-62 Jan-65 Jan-68 Jan-71 Jan-74 Jan-77 Jan-80 Jan-83 Jan-86 Jan-89 Jan-92 Jan-95 Jan-98 Jan-01 Jan-04 Jan-07 Jan-10 Ruminations on Market Timing with the PE10 In his classic text Securities Analysis (1934), Benjamin Graham (Warren Buffett s mentor) and David Dodd observed, We would suggest that about sixteen times average earnings is as high a price as can be paid in an investment purchase in common stock. To smooth out the effects of the business cycle, they suggested using an average of five to ten years of earnings. Perversely, when profit margins are abnormally high (expansions), those temporarily high earnings are given a premium valuation (i.e. the market s PE is high also) and when profit margins are abnormally low (recessions), those temporarily low earnings are given a lower valuation (the market s PE is low). Given that profit margins are mean reverting there should be a negative correlation between margins and PE ratios, but instead we find a positive one. Yale professor Robert Shiller, in his book Irrational Exuberance (March 2000), popularized his version of this metric to demonstrate how overvalued the stock market was. In Dr. Shiller s version, the current price of the S&P 500 is divided by the average of the past ten years of earnings adjusted for inflation. Since his timing was impeccable, the metric, referred to variously as the Shiller PE, the PE10, the Graham PE, or the Cyclically Adjusted PE (CAPE) has become popular. (I should also note that the second edition of the book, published in 2005, was presciently updated to warn of the housing bubble. Dr. Shiller is now two-for-two and is currently predicting U.S. farmland as the next bubble.) Dr. Shiller continues to update his data, and I have graphed 1926-2011, as well as the average value over the time period (17.43), here: 50 PE10 45 40 35 30 25 20 15 10 5 0 Prior Month PE10 Average PE10
Jan-26 Jan-28 Jan-30 Jan-32 Jan-34 Jan-36 Jan-38 Jan-40 Jan-42 Jan-44 Jan-46 Jan-48 Jan-50 Jan-52 Jan-54 Jan-56 Jan-58 Jan-60 Jan-62 Jan-64 Jan-66 Jan-68 Jan-70 Jan-72 Jan-74 Jan-76 Jan-78 Jan-80 Jan-82 Jan-84 Jan-86 Jan-88 Jan-90 Jan-92 Jan-94 Jan-96 Jan-98 Jan-00 Page 2 As you can see, the market s valuation in 2000 really was unprecedented. After the dot com crash most investors felt the market had become fairly valued (or even cheap) because of the magnitude of the decline when in fact because the decline began at such an extraordinary level it was still expensive. Even with a further roughly 50% decline, the 2008 financial crisis did not bring the market to the typical lows that a crisis produces. In any case, the PE10 has continued to gain popularity due to graphs like the one below (which I created, but similar ones are widespread): 50 45 40 35 30 25 20 15 10 5 PE10 & Subsequent S&P 500 Returns -10% -5% 0% 5% 10% 15% 20% 0 25% Prior Month PE10 Next 10 yr. CAGR The graph shows the remarkable fit between the PE10 on the left scale (I used the prior month s number to avoid any look-ahead bias) and the subsequent decade s annualized returns on the S&P 500 (Compound Annual Growth Rate on the right scale, inverted). There is a correlation of -73% which means 54% of the subsequent decade s stock market returns can be explained by the PE10 at the beginning of the period (i.e. the R-squared is 54%). Clearly we can use the PE10 as an indicator of subsequent market returns and set more appropriate expectations. For example, as I write this the current PE10 is 21, which doesn t portend high market returns to come. What some commentators seem to imply is that the PE10 can also be used to time the market when the PE10 is high obviously you should get out of stocks and when it is low obviously you should get into stocks. The primary purpose of this paper is to explore that strategy to see if it is feasible.
Subsequent 10 Year CAGR Page 3 One of the problems with trading strategies of this sort is that they have low breadth. In other words, even given some skill in identifying when to get in and out of the market, there aren t enough opportunities to exercise it the randomness will overwhelm the skill without a high number of chances to use it. This is why basketball and baseball playoff series are seven games; there has to be a number of matchups to actually determine who the better team is and remove the effect of luck (good or bad). Similarly, a skilled investor would rather have an opportunity to bet $1 a million times than $1 million once. Before plunging into the merits of using the PE10 as a trading strategy, one more popular graph a scatter chart with the best fit trend line added to illustrate the attractiveness of using the PE10 to predict market returns (again, I created this, but similar ones are ubiquitous): 25% 20% 15% 10% 5% 0% -5% -10% y = -0.0058x + 0.2029 R² = 0.5373 0 5 10 15 20 25 30 35 40 45 50 Initial PE10 Using the best fit regression line shown, we can see that for any PE10 below 35 there is a positive expected return over the next decade. Of course, the investor can always invest in bonds, and stocks are risky so the hurdle should be higher than a zero percent rate of return going forward. This brings us to the strategy in question. Suppose for PE10 s above X the investor held only bonds, and for PE10 s below X the investor held stocks. This would seem to be a superior strategy to simply holding stocks or bonds or some constant mix. The lowest PE10 in our 1926-2011 time period is 5.57 (July 1932) and the highest was 44.20 (January 2000). Thus if we invested in bonds when the PE10 was above 5 we would have always been in bonds, above 45 and we would never have gone to bonds. Comparing a buy and hold portfolio of the S&P 500 vs. switching between the S&P 500 and five year treasury notes upon the attainment of some threshold PE10 we find:
Page 4 1.00% Alpha to Stocks 0.00% 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45-1.00% -2.00% -3.00% -4.00% -5.00% On the graph we see that only being in stocks when the PE10 is below 5 means we will never be in stocks and will underperform by 4.38% (the premium of stocks over bonds). If we increase the threshold our performance improves as we hold stocks more and more frequently. If we set the threshold to only move to bonds at PE10 s above 45 we are always in stocks so the difference is zero. It appears that at PE10 s above 15 there is little reward for being in stocks in fact setting the threshold at 30 there has been an alpha of 66 bps. The average time between switching at a threshold of 15 is 3.44 years and at 30 it is 10.75 years so this is not a strategy that requires a great deal of trading. It appears from the analysis so far that this may be a good market timing strategy but there is a problem. Even using perfect hindsight to know to set the threshold at 30 we have an alpha of only 66 bps, we would expect less in practice. But remember the earlier comments about this strategy having low breadth you are trading less than once a decade on average so even across an entire lifetime the number of opportunities for the strategy to work is fairly low and most people would have difficulty following it. If we set the threshold at 15 (i.e. if you aren t getting paid for the riskiness of stocks don t own them) we would have been in bonds from February 1989 until now with the exception of November 1990, and March through May of 2009. Very few people would have felt good missing the entire 1990 s bull market. The return on the switching strategy over that period was 8.38% while the return on the S&P 500 was 8.85%. Here is a graph of an investor long the strategy and short the S&P 500 over that period (i.e. when the line is going up you are winning and when going down losing):
Page 5 If we set the threshold at 30 (i.e. optimizing with perfect hindsight) we would have been in bonds from July 1997 until May 2002 (with the exception of October and November of 2001 and March 2002) and then would have remained in stocks from May 2002 until today. The return on the switching strategy over that period was 5.98% while the return on the S&P 500 was 4.29%. Essentially all of that alpha came from being in bonds rather than stocks in April 2002 (the market was down over 6% that month). Prior to that month the strategy lost and then gained to get back to exactly even. Psychologically, it may be easier to buy and hold than to follow this strategy. Doing nothing is far easier for folks to do than to make a switch. Inevitably emotion interferes with the decision to move to or from the market rather than it being a mechanical process. [Technical notes: First, the best fit line used in the third graph is linear but changing it to a more sophisticated curve increases the r-squared only trivially. Second, while I am discussing this using five-year treasury notes as the fixed income alternative, the conclusions are robust to changing that to treasury bills, long-term government bonds, long-term corporate bonds, or a mix. Third, using a band to reduce the frequency of switching (i.e. switch to bonds above 25 and to stocks below 20, between 20 and 25 don t change from the previous holding) does not meaningfully change the conclusions.]
Page 6 Notes: The analysis in this report has been prepared by David E. Hultstrom, MBA, CFP, CFA. Mr. Hultstrom is the president of Financial Architects, LLC, a financial planning and wealth management firm. Questions or comments are welcome, and he may be reached at David@FinancialArchitectsLLC.com or (770) 517-8160. Reasonable care has been taken to assure the accuracy of the data contained herein and comments are objectively stated and are based on facts gathered in good faith. We disclaim responsibility, financial or otherwise, for the accuracy or completeness of this report. Opinions expressed in these reports may change without prior notice and we are under no obligation to update the information to reflect changes after the publication date. Nothing contained in this material is intended to constitute legal, tax, securities, or investment advice, nor an opinion regarding the appropriateness of any investment, nor a solicitation of any type. The general information contained in this publication should not be acted upon without obtaining specific legal, tax, and investment advice from a licensed professional. Past performance is no guarantee of future results. This is not an offer, solicitation, or recommendation to purchase any security or the services of any organization. The foregoing represents the thoughts and opinions of Financial Architects, LLC, a registered investment advisor. Your mileage may vary. This report was originally written in August, 2012 and was last reviewed/updated in April, 2013. Financial Architects, LLC