Momentum Investing Can Achieve Market-Beating Returns
July 5, 2011
by Matthew Tuttle, CFP
The problem with MPT
MPT is pervasive because there are a number of powerful arguments in favor of it:
In an MPT-based portfolio, you would have made money every year from 1982-99.
You would have lost money in 2002 and 2008, but those years were just anomalies (the proverbial 100-year storm).
If you held on for the 10 years from 2000 to 2010, you wouldn’t have made much, but you wouldn’t have lost either, so MPT is safe for long-term investors.
While on the surface these arguments make sense, a deeper inspection shows some serious flaws:
The 1982-99 bull market was one of the biggest we have ever seen. Any strategy that recommended stock-buying would have made money. MPT, like everything else, benefited from an up market.
The years 2002 and 2008 were not flukes. It is human nature to create bubbles in asset classes that eventually burst. Human beings are ruled by emotions. Greed and ignorance cause bubbles to form, and fear causes them to burst. Until we all become Vulcans, bubbles will continue to form and burst. (Unless you believe that the government can head off bubbles in the future. If you do, I have a bridge in Brooklyn to sell you.)
Not making any money for 10 years is not a successful investment strategy. No matter what age you are, the opportunity cost of missing 10 years of returns is enormous.
Why momentum works
Conventional market wisdom is split between two theories, that t markets are efficient and/or random. If emotionless computers made all buy-and-sell decisions , this would probably be the case. However, human beings trade in markets, and they are ruled by fear, greed and ignorance. This causes trends in the market to be persistent.
Typically, trendsetters identify an undervalued area of the market and move in, driving up prices. Other investors see this and buy as well, further moving prices up. Eventually fear kicks in, money moves out of the asset class into something else and the process starts all over again.