Frog in the Pan Momentum: International Evidence

This article analyzes various reasons why momentum strategies might work outside US borders. While the US story is firmly rooted in behavioral biases, is the same true on an international scale? That seems logical and likely. In fact, the authors conclude that a “slow diffusion of news best explains momentum in the international context…across all of our tests, we find supportive evidence for the FIP (“frog in the pan”) proxy in both emerging and non-US developed markets.” The results indicate that underreaction to continuous news plays a key role in generating momentum internationally. I suppose that means that when investors underreact to steady news, it’s very like ignoring the frog in the pan – momentum begins to boil before any of the frogs notice.

Empirical determinants of momentum: a perspective using international data

• Amit Goyal, Narasimhan Jegadeesh, and Avanidhar Subrahmanyam
• Review of Finance
• A version of this paper can be found here
• Want to read our summaries of academic finance papers? Check out our Academic Research Insight category.

What are the research questions?

1. What is the frog in the pan phenomenon?
2. Do momentum results extend to international markets?
3. Given the results are similar to those of the US, do behavioral biases like overconfidence or limited attention influence momentum profitability globally?
4. How do various empirical proxies (e.g., Book-to-Market ratio, turnover, information discreteness) perform in explaining momentum internationally? The robustness of different hypotheses using cross-sectional and portfolio-level tests.

What are the Academic Insights?

1. FIP. The “Frog in the Pan” is an amusing analogy that describes how investors react to information depending on how that information is presented. Think of how a frog reacts when placed in water that is gradually heating up (not much of a reaction) vs. how a frog reacts when dropped into boiling water (immediately jumps out). This bias in behavior illustrates how the investor tends to underreact to information, good or bad, that trickles in gradually over time, yet responds more sharply to sudden, significant news. The result is what has been labelled “momentum” in stock prices at least in the US.
2. YES. Take a look at Table 1: Momentum outside the US. The returns on the WML (“winner minus loser”) portfolio, which long the winner decile and short the loser decile. The average monthly returns were 0.89% for all non-US, 0.85% for developed non-US, and 0.74% for emerging markets. The results are consistent with the results reported by the original momentum study, US data only, by Jagadeesh and Titman in 1993. It appears that momentum is a profitable strategy globally. The other stats provided in Table 1 benchmark the momentum performance for comparison elsewhere. For example, the standard deviation of momentum returns was highest in developed non-US countries at 6.11%, then 5.12% for emerging markets. Negative skewness was observed across all international markets reflecting occasional large losses, again similar to the US. The median return was larger than the average and was more volatile and negatively skewed for non-US developed markets. These results were consistent with earlier studies on global momentum: Griffin, Ji, and Martin (2003) and Rouwenhorst (1998). These results set the stage and are really necessary to establish context for the examining behavioral explanations of momentum in international markets.
3. YES. The authors argue that evidence for the FIP explanation of investor behavior was present as momentum was stronger when information flows are continuous rather than in arriving in large, discrete chunks. On the other hand, momentum was larger in low-vol and up-market conditions, suggesting overconfident and more active behavior on the part of investors. They appear to be more active and less rational during those periods.
4. Proxies for behavioral hypotheses included: book-to-market, information discreteness, turnover, and residual analyst coverage. Fama-MacBeth regressions confirmed the significance of coefficients for interaction terms between momentum and explanatory variables or proxies representing behavioral biases. The interaction term for book-to-market was 0.034 for all non-US markets, 0.038 for developed non-US), and 0.043 for emerging markets, all being consistent with the overconfidence hypothesis.

Why does it matter?

The market efficiency debate is central to the field of finance, and the ability of the field to explain investor behavior. The debate, however, continues to this day and momentum has been a sticky thorn in the side of those promoting rational and efficient markets. In particular, momentum flies in the face of the most conservative level of market efficiency: the weak-form notion of market efficiency. Weak-form efficiency holds that a stock’s relative performance is unrelated to its past performance at any time horizon. Strategies like momentum investing should not produce excess profits, nor should they be outside the ability of generally accepted risk models to account for results. So there you have it: Behavioral biases in global markets. Just another nail in the coffin of market efficiency.

The most important chart in the paper

The results are hypothetical results and are NOT an indicator of future results and do NOT represent returns that any investor actually attained.  Indexes are unmanaged and do not reflect management or trading fees, and one cannot invest directly in an index.

Abstract

Although momentum exists in many markets throughout the world, explanations for momentum have
largely been tested using US data. We investigate the extent to which US-based momentum explanations extend to the international context, using regression-based and portfolio approaches. Among the several hypotheses we consider, we find reliable support for the hypothesis that due to limited attention, investors underreact to information arriving in small bits rather than in large chunks, which results in momentum. We also find secondary support for the overconfidence hypothesis for momentum. Finally, we find that momentum is stronger in up-markets and less-volatile markets in the international context just as in the USA. This finding also accords with the investor overconfidence hypothesis, under the proviso that investors are more confident in rising, low-volatility markets.