GB792 Unit 2 Empirical Study Critique Example

What Else Separates Winners From Losers: A Unit 2 Critique of Jegadeesh and Titman (1993)

[Student Name]

Purdue University Global

GB792: Advanced Topics in Finance

Unit 2 Empirical Study Critique

[Instructor Name]

[Date]

This critique concerns a real published study; its reading of the design and its rivals is the candidate's.

The Study

The article under review is Jegadeesh and Titman (1993), given in full in the reference list.

Part One: The Claim and the Design

The claim can be stated in a sentence: stocks that performed well over the past three to twelve months continue to outperform stocks that performed poorly, over the following three to twelve months. The authors interpreted the pattern as evidence that investors underreact to information, so that prices move toward value gradually.

The design was a set of portfolio strategies. Using New York and American Stock Exchange stocks from 1965 to 1989, each month the authors ranked stocks on their returns over a formation period of three, six, nine or twelve months, sorted them into ten deciles, bought the top decile and sold the bottom, and held the positions for a holding period of three to twelve months. Because a new portfolio was formed each month and held for several, the strategies overlapped. Each strategy was run twice, once starting immediately and once skipping a week between formation and holding, to separate the result from short-term reversals caused by bid-ask bounce. The six-month formation, six-month holding strategy earned close to 1 percent a month, about 12 percent a year.

Part Two: The Central Comparison

Everything in the paper rests on one comparison: the future returns of the winner decile against those of the loser decile. The groups were formed only on past returns, but stocks that had very high or very low past returns may differ from each other in other ways too. If any of those other differences predicts future returns, it could produce the spread without any underreaction. A critique must therefore ask what else differs between the two deciles and whether the paper rules each difference out.

Part Three: Five Rival Explanations

First, systematic risk. If winners carried more market risk than losers, the spread could be compensation for risk. The paper tested this directly by estimating the betas of the portfolios and found that the losers, not the winners, had the higher betas. On this evidence, risk as measured by the market model works against the result rather than explaining it.

Second, persistent differences in expected returns. Conrad and Kaul (1998) argued that momentum profits could arise if stocks differ permanently in their expected returns. Ranking on past returns would then sort high expected return stocks into the winner decile, and they would keep earning high returns, not because of underreaction but because of what they are. This explanation makes a prediction the paper's own evidence can test: if winners are simply high expected return stocks, the spread should persist indefinitely. Jegadeesh and Titman reported that part of the first-year gain reversed over the following two years, and their 2001 follow-up strengthened that finding. Reversal is hard to reconcile with permanent differences in expected returns.

Third, industry concentration. Winners and losers may come from different industries. Moskowitz and Grinblatt (1999) found that much of the momentum effect could be attributed to industry momentum, in which whole industries continue their past performance. That finding does not overturn the result, but it changes its interpretation, since underreaction to industry news is a different story from underreaction to firm-specific information.

Fourth, microstructure. Extreme past returns are more common among small, illiquid stocks, where bid-ask bounce can produce apparent reversals and continuations. The one-week skip addresses the reversal problem, and the authors reported that results held when small firms were excluded, which limits this explanation.

Fifth, trading costs. Lesmond et al. (2004) argued that the stocks generating momentum profits, especially losers, are small and expensive to trade, and that once realistic costs are deducted the profits largely disappear. This rival does not deny the pattern in prices; it denies that the pattern could be exploited. That distinction matters for interpretation: a pattern too costly to trade on is consistent with a market in which no one can profit, even if prices are not perfectly efficient.

Each rival is useful only if it makes a prediction the paper's own tables could confirm or refute, and the reversal evidence is what defeats the expected-return explanation.

Part Four: Economic Magnitude

A return of about 1 percent a month before costs is large by the standards of asset pricing. The question is how much survives. The strategy trades heavily, since portfolios are re-formed monthly, and short positions in small loser stocks are costly to maintain. The paper did not deduct trading costs in its main results. Estimates in later work vary with assumptions about costs and portfolio construction, and the profits are larger when the strategy is confined to larger, more liquid stocks, where costs are lower. The economic significance of the result therefore depends on implementation, and a fair reading separates the pricing pattern from the trading opportunity.

A Note on the Sample Period

The sample ended in 1989, and a critique should ask whether anything about 1965 to 1989 could have produced the result on its own. The period includes the high-inflation 1970s and the 1987 crash, and a pattern found in one twenty-five-year window could, in principle, be a feature of that window. The paper offered some protection by reporting results across subperiods, which showed the pattern in each. The stronger protection came later and from outside the paper: the authors' own test on the 1990s and studies of other national markets, which found the pattern in data the original authors never saw. Those later tests belong to another unit, yet they bear on how much weight the 1993 evidence can carry by itself.

What the Paper Establishes

The 1993 results support a strong claim about prices: over 1965 to 1989, past winners outperformed past losers over horizons of three to twelve months, and the difference is not explained by market risk or by small-stock microstructure. The reversal evidence makes the persistent expected-return explanation unlikely.

The interpretation reaches further than the evidence. Underreaction is consistent with the pattern, but the paper did not test it directly; it inferred it from the absence of alternative explanations. Industry momentum, later documented, shows that the information investors underreact to may be broader than the firm-level news the paper had in mind. And the trading cost evidence suggests that the pattern, though real, may not offer a profitable strategy to most investors. The paper established a fact about returns that later research has found remarkably durable. It proposed, but did not establish, the reason for it.

References

Conrad, J., & Kaul, G. (1998). An anatomy of trading strategies. Review of Financial Studies, 11(3), 489-519. https://doi.org/10.1093/rfs/11.3.489

Jegadeesh, N., & Titman, S. (1993). Returns to buying winners and selling losers: Implications for stock market efficiency. The Journal of Finance, 48(1), 65-91. https://doi.org/10.1111/j.1540-6261.1993.tb04702.x

Lesmond, D. A., Schill, M. J., & Zhou, C. (2004). The illusory nature of momentum profits. Journal of Financial Economics, 71(2), 349-380. https://doi.org/10.1016/S0304-405X(03)00206-X

Moskowitz, T. J., & Grinblatt, M. (1999). Do industries explain momentum? The Journal of Finance, 54(4), 1249-1290. https://doi.org/10.1111/0022-1082.00146