The authors offer a reappraisal of Keynes’s investment performance and assess his contribution to professional asset management. In the mid-1920s, a time when larger investors regarded stocks as an alternative asset, Keynes believed equities constituted a new, separate asset class. At that time, legal restrictions prevented institutional investors from investing in equities, which were largely the domain of the individual investor. The authors regard Keynes’s shift into equities for the King’s College, Cambridge, endowment fund early in the 1920s as being as radical as well-known U.S. institutional investors’ moving into illiquid assets in the late 20th century.

Using modern research techniques, the authors analyze original records for explanations of Keynes’s success and discover two particular insights. First, Keynes’s record, clarified by the authors, demonstrates the importance of intellectual flexibility to adapt investment approaches to changing times and to learn from mistakes. Second, the organizational setup that Keynes enjoyed at his college endowment fund was important to his success because it allowed him the freedom to make the switch into equities and to experiment widely with different types of risk. According to the authors, Keynes also varied his sectorial emphasis quite frequently. At times other than the period of 1940–1945, the majority of his holdings were in small- and mid-cap stocks outside of the mainstream. He also rotated between high- and low-dividend-yield stocks.

The authors analyze Keynes’s trading in quoted U.K. stocks owned by the Kings College, Cambridge, endowment, over which he held full discretion between 1923 and 1946. A total of 954 transactions by Keynes—567 buys and 387 sells—are examined. Transactions outside the United Kingdom, day trades, misbookings, and transactions with unclear capital raisings or rights issues are excluded. For the benchmark, the authors use the capitalization-weighted 100-share U.K. equity index series estimated by Dimson, Marsh, and Staunton (Global Investment Returns Sourcebook 2011). They also iterate monthly intrayear values using fluctuations in other indices and estimate dividend income using data reported by companies in the index.

Using an event study analysis, the authors assess Keynes’s trading behavior with standard methodologies commonly used to analyze the performance of individual investor trades. Abnormal returns are calculated in a window centered on the month of the transaction. Return estimates cover 12 months before and after the transaction month. Buy-and-hold abnormal returns for each security are calculated as geometric differences between observed cumulative returns over a specified interval and cumulative beta-adjusted returns on the market. The authors find that both bought and sold stocks increased strongly in the 12 months prior to a transaction by Keynes. Post-purchase share prices outperformed the market by 3.3% over the 12 months following the purchase. Sells also outperformed the market by 4% over the 12 months following disposal.

A Quandt–Andrews breakpoint test and other econometric techniques indicate a break in trading behavior in the early 1930s. Correspondence excerpted in the study corroborates a change from a top-down to a bottom-up approach. The authors further examine the data by dividing the analysis into two periods: before and after this change.

In the 1920s bull market, Keynes tended to buy stocks that had outperformed; those stocks subsequently underperformed by 4% in the 12 months after purchase. In the earlier period, purchases and sales were usually made after favorable performance, which is consistent with recent studies of overconfident U.S. individual investor behavior.

In the 1930s and 1940s, Keynes’s buys produced a post-transaction outperformance of 6.2% in the 12 months after purchase. Between his earlier and later periods, Keynes improved the timing of his disposals of short-term holdings. But the timing of his disposals of long-term holdings worsened. Post-transaction performance reinforces the conclusion that Keynes exhibited greater skill in buying individual stocks after 1932.

The authors also test for susceptibility to the disposition effect (the desire to avoid regret by selling winners too early and holding on to losers). They analyze the proportion of realized gains and that of realized losses using a standard test and find that Keynes displayed a disposition effect across all periods.

Finally, the authors use a binomial distribution approach to test whether Keynes’s trading behavior is indicative of contrarian or momentum tendencies. They find that Keynes displayed weak contrarian behavior prior to 1932 and strong contrarian behavior after 1932, which provides further evidence of a change in investing style.

Keynes is still invoked today to support many diverse viewpoints. This study clarifies his investment successes and changes in his investment style. It prompts as many questions as it answers but does provide evidence of his undoubted insight into his own era. One question the study raises is how much of his performance was attributable to what we would consider insider trading today because of his close involvement and personal relationships with the governing class. Given his portfolio bias toward smaller companies, the abnormal return against a large-company benchmark may be evidence of the small-cap effect rather than Keynes’s skill. Another question is how Keynes fared with his multiasset capabilities outside of the narrow field of domestic U.K. equities. Unfortunately, we may never know.