Market Timing: Does it Make Sense?

If you ever want to start a heated conversation among investors, ask them if they think market timing makes sense. One side will claim that academic studies show it doesn't. The other side will say that practical experience shows that it does. Unfortunately, the heat generated by this argument too often hides a central truth: both sides are right.

In our view, the crux of the matter is the failure by too many academics and financial writers to distinguish between what we call “four types of market timing”, based on (a) whether it is done systematically or episodically, and (b) whether its objective is earning additional returns or limiting downside risk.

Most of the academics who caution about the alleged evils of market timing are focused on episodic attempts to time markets in the hope of earning higher returns by adding to ones holding of asset classes expected to do relatively well, while reducing allocations to asset classes whose returns are expected to lag. As we have repeatedly noted, this is a hard game to play consistently well for the simple reason that the required returns forecasting accuracy is beyond the skills of most investors (for the classic analysis of this, see "Likely Gains from Market Timing" by William F. Sharpe).

On the other hand, we also believe that prudent risk management can logically lead to episodic market timing when an asset class appears to be substantially overvalued. Our examples of such circumstances in the past include the UK pound in 1992 and US equities in 2000. Under these circumstances, reducing one's allocation to an overvalued asset class to a level below one's long-term target may make sense, because it can limit large losses. Is this not the same as episodic forecasting to earn high returns? We do not believe it is. Historically, relative asset class risk has been much more stable than relative asset class returns. Hence, when trying to limit risk, rather than earn higher returns, an investor confronts a somewhat easier forecasting problem. The problem, of course, is that hindsight is always much clearer than foresight. In both the cases we cited, right up until the big downside move occurred, there were still people arguing that prices would rise higher. As we repeatedly note in our writing, while overvaluations are corrected over a shorter time than they take to develop, and those turning points are virtually impossible to call in advance. They are always shrouded in uncertainty, and require investors to make judgment calls that may turn out to be premature.

In contrast to episodic market timing, which results from the analysis of a specific set of circumstances, systematic market timing entails actions which are taken automatically. A good example of this is a disciplined rebalancing strategy. If markets aren't always perfectly efficient (and we don't believe they are), index funds will naturally be overweight in overvalued securities within an asset class, and underweight undervalued securities. Systematically rebalancing out of asset classes in which one is overweight relative to one's target, and into those in which one is underweight helps to offset this “momentum risk” that is inherent in index funds in a market that isn't always perfectly efficient.

An investor also has to decide whether to rebalance back to his or her target asset class weights, or to attempt to enhance portfolio returns over time by rebalancing overweight asset classes to slightly below their target weights, and underweight asset classes to slightly above their target weights. The available evidence suggests that, over long periods, mean reversion in returns is a fact of life in most asset classes. Hence rebalancing to slightly above or below target weights should be able to exploit this tendency, and enhance portfolio returns a bit over time. Backtesting results show that this has been true historically. However, as is always the case, this is no guarantee that it will continue be true in the future.

We help subscribers make marketing timing decisions in a number of different ways. In March and September, we publish our Economic Update, which includes our most likely and most dangerous scenarios for the future, along with their implications for asset class returns and key indicators to watch. Each month, we publish our valuation updates for the world's major bond and equity markets. These are based on an analysis of the likely future supply of, and demand for returns in these markets. Finally, along with our model portfolios, we include a detailed discussion rebalancing strategy.

If you found this article to be useful and informative, more is available every month along with all of our archives back to 1997. We hope you subscribe today. A year subscription of The Index Investor is only U.S. $195.

Download Free Section PDF
Subscribe to The Index Investor