Temptations abound on the road to Long Point Provincial Park, which sits on the northern shore of Lake Erie. On the way to its beaches, I drive perilously close to the small town of Vittoria, whose inhabitants make devilishly tasty butter tarts.
There are only a few things in this world that are as tempting as a good butter tart. When it comes to investing, the urge to time the stock market is one of them.
The idea of being able to jump into stocks when the going is good, and then to get out before it all goes bad, can be very seductive. But I’ve yet to find a reliable way to do so.
Another nail was hammered into the market-timing coffin by money-manager Tobias Carlisle in a recent informal study. He looked at trying to use Robert Shiller’s cyclically-adjusted price-to-earnings ratio (CAPE) as a timing trigger for value portfolios.
You’ll remember that Prof. Shiller’s CAPE compares the price of the market to its long-term earnings trend. In the U.S., the ratio is calculated by dividing the S&P 500’s current price by the average of its inflation-adjusted earnings over the past 10 years. Stocks are thought to be generally cheap when the market’s CAPE is low, and expensive when it is high.
Mr. Carlisle calculated the performance of investors who moved from a simple value portfolio to Treasuries when the market’s CAPE fell below a set trigger point. They then moved back when the trigger point was exceeded.
He used Kenneth French’s value portfolio to obtain long-term return figures from 1926 through 2013. It follows an equally-weighted basket of U.S. stocks with the lowest 10 per cent of price-to-book-value ratios and is refreshed annually.
Mr. Carlisle considered three different trigger points. The first was the average CAPE for the full period, which came in at 17.6. The second was set one standard deviation higher at 24.8. (The ratio exceeded this level roughly 16 per cent of the time). The third used a very high CAPE trigger at 32.0, which is two standard deviations above the norm and occurred only about 2 per cent of the time.
Using the average CAPE as a timing trigger resulted in average annual returns of 13.4 per cent from 1926 through 2013. Swapping value stocks for Treasuries at a CAPE of 24.8 yielded 18.2 per cent annual returns over the same period.
Value investors who hid out in Treasuries when the market’s CAPE exceeded 32.0 – an extraordinarily high level seen only near the turn of the century – gained 19.4 per cent annually.
All of the timing methods reduced long-term returns. Investors who stuck with Prof. French’s value portfolio, through thick and thin, gained 20.0 per cent annually from 1926 through 2013.
The steady-Eddie value portfolio suffered a maximum drawdown of 85 per cent. (The crash of 1929 was not kind to investors.) But market timing also left investors exposed to some pretty big downturns, despite being in cash for long periods.
Those using the average-CAPE trigger suffered from a maximum drawdown of 69 per cent. The moderately high trigger at 24.8 resulted in a maximum drawdown of 80 per cent. The highest trigger yielded a drawdown of 84 per cent.
In addition, all three timing strategies failed to beat the simple value portfolio on a risk-adjusted basis, based on their Sharpe ratios.
I hasten to add that Mr. Carlisle’s study suffers from a touch of look-ahead bias. After all, investors in the 1930s didn’t know what the market’s CAPE would be in the 1990s. But it probably makes the case against timing even stronger. After all, investors in the early decades would probably have used lower trigger points, which would have generated even worse results.
If you’re a value investor who gets the urge to time the market, it might be better to take a drive to the beach to clear your head.