skip to Main Content

Why you might want to dilute a winning strategy with a lower-returning one

A motorsailer may not be the fastest boat, but it will always get there

In my newsletter I track two exceedingly simple models, one for value and one for momentum, as applied to 20 different stock markets, a REIT index, and a host of commodities. These aren’t portfolios, but just rubrics for each market on its own. See the letter for all the details, but the momentum rule will be familiar to anyone who has read about tactical asset allocation: it’s simply in or out based on the monthly close vs the 12-month moving average. The value rule scales in and out as thresholds in the CAPE ratio are crossed. For example, here are the results for Italian and Japanese stocks:


In some markets, the value approach has worked better, but overall, the 12-month momentum rule is the clear winner, both in terms of returns and drawdowns. Here are the averages of these models over the 20 countries tracked. For this, I just divide our theoretical funds evenly, and add each new market as it becomes available (when my data series starts). The bottom chart shows maximum drawdowns (percent decline from highest peak to lowest subsequent trough). 

Summary stats

Give up some long-term return for better peace of mind

Take a close look at the periods from 2000-2003 and 2007-2012. During those stretches, which spanned deep bear markets, the momentum strategy was essentially flat. Sure, it preserved capital while stocks plummeted, but the value strategy did an adequate job in that regard, and it recovered to new highs faster. In 2000-2003 especially, it didn’t even register a decline, but kept chugging higher due to its heavy bond allocation. 

The 50/50 combination of value and momentum (each deployed independently) has a substantially lower maximum drawdown than either of its components. This is because those components suffered their worst hits at different times (momentum in 1987, and value in 2008). Value only drew down about 10% in 1987, when momentum had the rug pulled out from under it, and momentum did a nice job at stepping aside prior to the 2008 financial crisis. 

Value and momentum derive their returns from fundamentally different sources. One looks at underlying earnings, the other at price vectors. Especially when you are working only with stocks and a cash vehicle (here Treasuries) it’s prudent to diversify factors. The combination will be easier to stick with than either strategy in isolation, since value and momentum tend to perform differently at critical times. Value sells before the top and buys before the bottom, while momentum holds on beyond the top and re-enters well after the bottom. You are bound to develop doubts at such times when your account differs radically from a conventional equity portfolio. 

With a tactical portfolio or a conventional one, frustration is the price of long-term outperformance. But by focusing on risk and not return, we can coach ourselves into following a different benchmark, and the longer we stick with such a plan, the better our chances for outperforming all the benchmarks.


This Post Has 2 Comments
  1. Great site and the newsletter is really helpful.i was wondering what specific ETFs for the 5class global momentum model provide the closest tracking.

    Is the difference in returns of the actual ETF performance vs the indexes minimal?

Comments are closed.

Back To Top