Broadly speaking, quantitative investors think of the market as having two major regimes: momentum and mean-reverting. In plain terms, the underlying assumption of momentum strategies, or “trend-following”, is that when prices go up, they will continue to go up, and when prices go down they will continue to go down. Mean-reverting strategies, on the other hand, compare current prices to historical averages and believe that at some point, prices that are too low will converge to what they have been, and prices that are too high will fall back down.
There is economic intuition for both. AQR, a quantitative hedge fund and perhaps the most vocal proponent of momentum investing, has written numerous papers about the strength and persistence of the phenomenon across centuries, even after widespread investor knowledge of its existence. Plausible economic interpretations of the momentum phenomenon range from behavioral models that hypothesize chronic investor under-reactions to new information to risk-based models that frame momentum outperformance as compensation for higher-risk assets or speculators exploiting commercial hedging activity [1].
Mean-reverting strategies, the counterpart to momentum, have similarly intuitive justifications. In some sense, a “mean-reverting” assumption is what underlies traditional long-short value investing. Investors that are long assets believe that those assets are trading at a discount to some “true” value, and that over time the market comes to reflect that truth (and vice-versa in the short case). Again, there are countless views and papers on the “value” phenomenon, ranging from the classic fundamental investors Benjamin Graham and Warren Buffett to the erudite Chicago Booth academics of the efficient markets tradition (Eugene Fama, Kenneth French, Henry Markowitz, etc).
(A small but important aside: that the momentum and value factors have persisted decades after publishing flies directly in the face of the efficient markets hypothesis, and should engender some degree of confidence in the possibility of generating consistent returns in the markets).
Conceptions of Momentum: Reflexivity
George Soros, legendary macro investor infamous for “breaking the Bank of England”, didn’t originally aspire to be the investment juggernaut he is today. Originally, after escaping Nazi persecution, he hoped to become a famous philosopher. However, so the story goes, after he woke up one morning and was unable to understand the esoteric philosophy that he wrote the night before, he decided to devote his efforts to making money instead. It turned out that the philosophy that he had developed was immensely useful in understanding financial markets.
Soros proposed his theory of reflexivity in the Alchemy of Finance. Glossing over a great deal of the philosophical underpinnings of his theory, one can conceive of reflexive phenomena as those in which cause and effect are commutative. So, the cause causes the effect, and the effect can also cause the cause. Examples of this in financial markets include:
- The Latin American debt crisis, in which loans made on the basis of a debtor country’s debt-to-income ratio served to directly influence those same ratios.
- Business investment leads to more jobs, which lead to more demand (as a result of rising incomes), leading to more business investment.
- In bubbles, rising prices directly lead to more demand, leading to rising prices.
- The conglomerate boom of the 1960s, in which acquirers made acquisitions on the basis of their high P/E ratios, while acquisitions increased P/E ratios.
(Another interesting aside is that the metaphysic of reflexivity in finance is not simply limited to macro investments. It is just as descriptive in technology ventures. It’s simply known by a different name in that domain: network effects, in which a platform with more users attracts more users).
Framed another way, one cause of the momentum phenomenon (among many possible causes) is the degree to which the economic drivers of prices are self-reinforcing. Of course, each of the above phenomena did not last into perpetuity, and in fact suffered painful reversals (i.e., mean-reversion) at some point. Investors must be as able to protect themselves in those scenarios as they are able to capitalize on the initial wave itself. However, a discussion of the underlying dynamics of mean-reversion is beyond the scope of this essay.
For now, we explore the process of building a rough signal to help us detect one such reinforcing cycle. You can find our research linked here, though you will need a password to access it.
[1] Asness et al, 2014.“Fact, Fiction and Momentum Investing”
