You are here: Knowledge Center Reading Room

Market Timing

E-mail Print PDF
Market timing refers to an attempt to time investing decisions so as to invest in assets which are expected to go up in the near term and divest from assets which are expected to go down in the near term.  The most simple implementation may be to shift one's assets between cash and stocks generically in order to take advantage of anticipated stock market movements.  As you can see from the below studies, attempts at market timing are only likely to work spuriously, due to occasional good fortune.  We strongly advocate against market timing in all its various forms.  Also, see Dollar Cost Averaging.

"The long, sad history of market timing is clear: Virtually nobody gets it right even half the time.  And the cost of getting it wrong wipes out the occasional gain of getting it right.  So the average investor's experience with market timing is costly.  Remember, every time you decide to get out of the market (or get in), the investors you buy from and sell to are the best of the big professionals.  (Of course, they're not always right, but how confident are you that you will be 'more right' more often than they will be?)  What's more, you will incur trading costs or mutual fund sales charges with each move—and, unless you are managing a tax-sheltered retirement account, you will have to pay taxes every time you take a profit." — Charles Ellis, Winning the Loser's Game

"The mathematical expectation of the speculator is zero. ... The expectation of an operation can be positive or negative only if a price fluctuation occurs — a priori it is zero." — Louis Bachelier, Theory of Speculation, 1900

William F. Sharpe, "Likely Gains from Market Timing," Financial Analysts Journal, March-April 1975, pp. 60-69. "... unless a manager can predict whether the market will be good or bad each year with considerable accuracy, (e.g., be right at least seven times out of ten), he probably should avoid attempts to time the market altogether."  Written by a Nobel Prize winner.  Also, see the QuickMBA summary below.

Robert H. Jeffrey, "The folly of stock market timing: no one can predict the market's ups and downs over a long period, and the risks of trying outweigh the rewards," Harvard Business Review, July-August 1984,  pp. 102-110. "... no one can predict the market's ups and downs over a long period, and the risks of trying outweigh the rewards."  "... only a few wrong [market timing] decisions ... deflate the long-term results produced by market timers to the point where the timers would be just as well off out of the stock market entirely."  Also, see the updatesummary by Rothery below.

Kirt C. Butler, Dale L. Domian, and Richard R. Simonds, "International Portfolio Diversification and the Magnitude of the Market Timer's Penalty," Journal of International Financial Management and Accounting, 6(3) 1995. This excellent study looks at the magnitude of the "market timer's penalty."  Basically, the market timer of this study desires to move assets from one country's stock market to another, presumedly in order to increase risk-adjusted returns.  This study compares a random timer (i.e., a market timer known to have no timing skill) with a buy-and-hold investor.  It finds that the timer's portfolio is 26.2 percent more risky with the same expected return as that of the buy and hold investor.  Equivalently, at the same level of risk, the unskilled market timer gives up 20.8% of the buy-and-hold investor's expected return over the risk-free rate.  The study correctly notes that the potential benefits of market timing are greatest when correlation between assets is lowest.  But this is also precisely the time when the "market timer's penalty" is the greatest.  A market timer without skill should clearly get out of the market timing business.

Jess H. Chua, Richard S. Woodward, and Eric C. To, "Potential Gains from Stock Market Timing in Canada," Financial Analysts Journal, SeptemberOctober 1987, pp. 50-56. This interesting paper finds that it is more important to correctly predict bull markets than bear markets.  And buy-and-hold investors effectively have a 100% accuracy in correctly forecasting bull markets.  "If the investor has only a 50% chance of correctly forecasting bull markets, then he should not practice market timing at all.  His average return will be less than a buy-and-hold strategy even if he can forecast bear markets perfectly [which is an extremely heroic assumption]."

William G. Droms, "Market Timing as an Investment Policy," Financial Analysts Journal, JanuaryFebruary 1989, pp. 73-77. "Gains from market timing over the long run require forecast accuracies that are likely to be beyond the reach of most managers.  More frequent forecasting increases the potential return available and reduces the level of accuracy required to outperform the market, but the transaction costs incurred in more frequent switching reduce the advantage."  Note that the study ignored the largest transaction cost for taxable accounts: short-term capital gains taxes.  If taken into consideration, this would have dramatically lessened the attractiveness of market timing as a strategy even more.  Note also that prediction becomes much harder as the time period of the predictions becomes shorter.

John R. Graham and Campbell R. Harvey, "Market timing ability and volatility implied in investment newsletters' market timing recommendations," Journal of Financial Economics, December 1996, pp. 397-421 (2.2mb). This excellent paper examines whether investing newsletters exhibit market timing ability.  "We find no evidence that letters systematically increase equity weights before market rises or decrease weights before market declines."

Roy D. Henriksson, "Market Timing and Mutual Fund Performance: An Empirical Investigation," Journal of Business, January 1984, pp. 73-96 (4.55mb).  This paper investigated whether mutual funds exhibited market timing ability.  "The empirical results ... do not support the hypothesis that mutual fund managers are able to follow an investment strategy that successfully times the return on the market portfolio."

Burton G. Malkiel, "Models of Stock Market Predictability," Journal of Financial Research, Winter 2004, pp. 445-459 (159kb). This study finds that, while there appears to exist some reversion to the mean behavior, "... there is no evidence of any systematic inefficiency that would enable investors to earn excess returns."  In other words, market timing models aren't likely to be fruitful.

Mario Levis and Manolis Liodakis, "The Profitability of Style Rotation Strategies in the United Kingdom," Journal of Portfolio Management, Fall 1999, pp. 73-86 (131kb). This paper looks at the feasibility of market timing between growth and value stocks, and between large and small stocks, in the United Kingdom (i.e., if you wanted to tactically switch between growth and value, or between large and small, how accurate would your forecasting need to be in order to beat a buy and hold strategy?).  "Our simulation results suggest that forecasting the size spread with a 65%-70% accuracy rate may be sufficient to outperform a long-term small-cap strategy. Beating a long-term value strategy, however, is markedly more difficult; it requires more than 80% forecasting accuracy."

Stephen L. Nesbitt, "Buy High, Sell Low: Timing Errors in Mutual Fund Allocations," Journal of Portfolio Management, Fall 1995, pp. 57-60. "Our finding is that market-timing activity by mutual fund investors costs over 1% per year in return performance versus what mutual funds actually report their performance to be."

Austin Pryor, "Timing isn't as Significant as You Might Expect," Sound Mind Investing, July 2003. An excellent article pointing out that even perfect timing doesn't do much better than essentially random timing (actually, it compares a perfect timing of deposits routine with a dollar cost averaging routine).

S P Umamaheswar Rao, "Market Timing and Mutual Fund Performance," American Business Review, June 2000, pp. 75-79. "The empirical results do not support the hypothesis that mutual fund managers are able to follow an investment strategy that successfully times the return on the market portfolio."  Note that if highly paid, highly educated, highly experienced mutual fund managers can't successfully time the market with the assistance of large support staffs of brilliant analysts, it seems imprudent to think that any particular layperson can expect do so.

Norman Rothery, "Timing Disaster," Stingy Investor. An update to (and summary of) the Jeffrey article above.

Robert Sheard, "Market-Timing Futility," Motley Fool, July 1 1998. This article supports disciplined periodic investment.  It suggests that, since the long-term return difference between making periodic investments with perfect timing and with perfectly imperfect timing is small, the important thing is to be making the periodic investments, not to try to time the markets.  "...for a genuine long-term investorsaver ... it makes precious little difference when you invest."

John D. Stowe, "A Market Timing Myth," Journal of Investing, Winter 2000, pp. 55-59. This paper points out that the often cited reason for avoiding market timing isn't a valid reason.  Many articles suggest that, because the majority of the stock market's gains are confined to a small number of days or weeks or months, imperfect market timing is likely to result in missing those runups, with catastrophic results.  This article argues that the same rationale might suggest that the market timer is just as likely to miss the few worst days, weeks or months, which would tend to increase returns.  Both results would likely be due to luck.  Market timing still seems imprudent, but this isn't the reason why.

Jack L. Treynor and Kay K. Mazuy, "Can Mutual Funds Outguess the Market?," Harvard Business Review, July-August 1966, pp. 131-136. "Are mutual fund managers successfully anticipating major turns in the stock market? ... [the study] shows no statistical evidence that the investment managers of the 57 funds have successfully outguessed the market."

Back to Reading Room