Scientific research is tricky. I am always skeptical, as a default stance, of any research, and scientific research is as prone to error and erroneous interpretation, as any human endeavour. Having said that, our conclusions about how the world and how our society work should have some sort of logical underpinnings. That is to say that any observation of what works (e.g. momentum investing) should be supported by an explanation that makes sense.
To draw a comparison with fundamental investing (buying companies that are more efficient, better run and hence more profitable), this is an investment approach that appears to work. Warren Buffet has been very successful at it, as have other. Why should it work? Because profit creates value, good management creates profit and if an investor can identify good management, that should translate into increased stock market gains when the rest of the investing public realizes the company’s value (either by paying more for the stock, or refusing to sell it cheaply).
I am satisfied with backtests and real-time tests of my momentum investing model that it works. Why should it work? An article in The Economist about Popularity, Luck and Herding refers to a study that holds the answer. Abhijit Banerjee at MIT suggested that people make decisions based on two factors: their own information and the actions of others. Because it’s impossible to collect complete information and difficult to collect adequate information before making a decision, we can use a mental shortcut by accounting for the decisions of others, assuming that they have different or better information than we do. This leads to herding.
Herding was tested in a 2008 experiment that offered pop songs for download, but reversed the order of the popularity rankings. Sure enough, the least popular song (but highest ranked) was downloaded most often. Interestingly, after people listened to the songs, the rankings righted themselves over time.
I see two implications for momentum investing. First, it relies on popularity. It doesn’t matter which company is best run or most profitable, it only matters which company other investors are buying. We try to get in front of the herd to profit from later buyers. Second, these waves of popularity will right themselves as new information becomes available and is digested and acted on. If a company is popular (stock price rising as investors buy), but releases disappointing earnings, investors will begin to sell. The rankings will begin to change and our signals will tell us to sell and buy another company that is becoming more popular. Because popularity is always evolving, momentum is a relatively active strategy that aims to keep money moving in accordance with investor sentiment.
I believe this to be a solid foundation for an investment strategy. Rather than try to deny sentiment and human irrationality, as fundamental investing does, it responds to irrational behaviour, producing profit by staying in front. As would be expected, there is no guarantee and no ability to avoid fluctuation. No investment program is risk-free, but our momentum model reacts as quickly as possible to available information.