In his recent letter to investors, fund manager John Hussman stated, “This is an obscenely overvalued market.” There are reasons to worry that he might be right. The current market capitalization of U.S. equities stands at $40 trillion, more than twice the level of GDP. That is the highest ratio of market cap to GDP ever recorded. The previous high-water mark was 1.9 at the height of the dot.com bubble. The Shiller CAPE index (reproduced from Prof. Shiller’s website below), and most other measures of price to earnings and cash flow, are also close to all-time highs. On the other hand, as economists like to say, times are also good. Unemployment is at record lows, inflation is subdued, the economy is growing steadily, and monetary policy is accommodative. As a result, some analysts argue that current market levels are justified. Prices are high but not “too high.”
Not surprisingly, our clients are asking what they should do. Are prices so high that they should move predominantly to fixed income? This turns out to be an extraordinarily difficult question to answer and rather than drone on, we feel the best way to illustrate the problem is with a simple interactive software program using Excel. We use the software in conjunction with the Shiller index, but it can be applied with any market indicator. The data to which the software is applied are given by the Shiller CAPE index, the monthly return on 30-day U.S. Treasury bills, and the monthly return on the overall US stock market as measured by the CRSP index of most all listed US stocks. The spreadsheet has two data periods. The full sample covers the period from January 1926 through December 2018. The shorter sample begins in January of 1950 to avoid both the Second World War and the great depression. It can also be thought as using more “modern” data.At Cornell Capital, we are cognizant of the valuation issue. It is our view that the market is richly valued, but we do not believe the proper response is exiting entirely, or even more dramatically going net short. As the spreadsheet software demonstrates, timing the market is hazardous. However, the current ten-year bull market has been far from uniform. By taking account of variation in fundamental valuation across securities and by using options as hedging tools, we believe the risk-return tradeoff can be managed effectively even in the current elevated market environment.
All the software requires in the way of inputs are the high level at which the investor exists the market and the low point at which he re-enters. The defaults in the sheet are 25 and 18. Given these inputs, the spreadsheet calculates the ending value, including reinvestment of interest and dividends, of three strategies: buying and holding Treasury Bills, buying and holding the stock market, and the trading rule of going in and out of the stock market based on the chosen low and the high points. When the investor is out of the stock market, it is assumed that he is holding Treasury bills.
Comparing first the difference between the Treasury bill and the stock market strategies, the distinction is striking. One dollar invested in Treasury bills in January 1926 grows to $20.98 by December 2018. The same dollar invested in the stock market grows to $5,089.23. For the shorter period, a dollar invested in January 1950 grows to $16.50 in Treasury bills and to $1,219.38 in the market by 2018. Clearly, it pays to be in the market most of the time, but perhaps you can do even better by getting out when the market is “overvalued.” The software makes it easy to check that.
At the default values of 25 and 18, the trading rule fails to deliver superior results. For the full period, the trading rule produces a final value of $3,547.75 compared to the buy and hold value of $5,089.23 for the stock market. For the shorter period, the trading rule does worse. The final value of $599.29 is less than half the buy and hold value of $1,219.38. However, these results are highly sensitive to the choice of high and low points, particularly for the full period. The reason that that full period results are so sensitive is that clever choices of the high point get the investor out of the market before the crash and back in at the low points of the depression. For instance, at inputs of 28 and 15 the trading rule produces a final value of $10,137.85, more than twice that of buy and hold. However, even with that choice the trading rule fails to outperform in the later period. You have to search carefully to find inputs that lead to outperformance in the shorter period.
Of course, if you search over enough alternatives high and low inputs, or if you use a variety of additional indicators other than the Shiller CAPE index, you will eventually find a trading strategy that dramatically beats buy and hold. But such an approach is data mining in the extreme. For the test to be fair, you must pick both the index and the high and low points in advance as you would have to do in the actual investment world.
The spreadsheet demonstrates how hard it is to beat buy and hold. But the fact that it is difficult does not mean it is unwise. As the Grossman-Stiglitz theorem holds, markets must be sufficiently inefficient that smart capital can make superior returns based on insightful fundamental analysis. And part of such fundamental is an assessment of the general level of the market.
 We discuss data mining in detail in our book, Conceptual Foundations of Investing.
 The Grossman-Stiglitz theorem is discussed both in previous posts on our website and in greater detail in Conceptual Foundations of Investing.