The Medallion Fund, the Definitive Counterexample to the Efficient Market Hypothesis

The performance of the Medallion Fund, a New York–based quant hedge fund run by Renaissance Technologies, is the ultimate counterexample that directly refutes the grand premise of Eugene Fama’s Efficient Market Hypothesis. If you had invested 100 dollars in the Medallion Fund in 1988, it would have grown to 398.72 million dollars by 2018. On an annualized basis, that is a staggering 63.3%. No academic paper has ever recorded numbers on this scale. And it is not just the numbers that are astonishing. Despite two massive market collapses—the dot-com bubble and the global financial crisis—the Medallion Fund did not post a negative return in a single calendar year. Its returns are also unrelated to traditional risk premia. Its beta (correlation with market direction) was negative, and all factor loadings were negative as well. In other words, these gains were not compensation for bearing more risk. To this day, no market theory can reasonably account for this performance.

In “The Man Who Solved the Market,” author Gregory Zuckerman traces the story of mathematician James Simons, founder of Renaissance Technologies, and the design principles behind the Medallion Fund. Appendix 1 in particular contains Medallion’s year-by-year results—numbers every investor ought to see. Even the word “overwhelming” feels inadequate.

In this article, the analysis is based on Medallion’s gross returns. Net returns reflect the industry’s highest fee structure that Medallion charges; even after fees the performance is extraordinary, but the fund’s pure investment skill is better captured by the gross numbers. Ironically, outside investors cannot invest in the Medallion Fund. All early external accounts were redeemed, and Renaissance Technologies created separate, institution-only funds (for example, RIEF and RIDGE). Medallion was reserved for Simons and the firm’s employees.

Looking at the return history, there is simply no precedent. I have read countless investment theories and papers, but nothing comparable to the Medallion Fund’s performance. From 1988 to 2018—31 years—there was not a single year with a negative return. During the dot-com bubble, the fund returned 56.6%, and even in the 2008 financial crisis it made 74.6%. From its second year of operation onward, the lowest annual return was 31.5%.

The most intuitive way to grasp this performance is to look at the power of compounding. If you had invested 100 dollars in early 1988 in the CRSP index (the market-cap-weighted average of the market), it would have grown to 1,910 dollars by the end of 2018—an annualized return of 9.98%. That is a respectable figure for the market as a whole. But if you had put the same 100 dollars into Medallion? 398.72 million dollars. Not 4 million, not 40 million. In just 31 years, the seemingly impossible—100 dollars turning into 400 million—actually happened.

Of course, there is some exaggeration embedded in these numbers. At some point the fund stopped taking outside capital, and even employees were forced to withdraw money along the way, so in reality investors could not fully enjoy the theoretical compounding effect. During that period, assets under management grew from 20 million dollars to 10 billion dollars, yet returns barely declined. That is evidence of just how robust Medallion’s strategy was.

According to the author, Medallion’s strategy is to open and close thousands of ultra–short-term positions simultaneously. Even if only 50.75% of all trades are winners, repeating that process millions of times produces astronomical profits. The central challenge with such a strategy is transaction costs, but Medallion managed to minimize those as well. The reported gross returns are after subtracting transaction costs.

Could Medallion’s returns be explained as compensation for bearing “risk premia”? No. On a calendar-year basis, it never once recorded a negative return. Its annual standard deviation was a fairly high 31.7%, but the average annual return was 66.1%. The Sharpe ratio was well above 2.0. Its beta was -1.0. In other words, Medallion also functioned as a hedge against market risk. Even when analyzed through the Fama–French factor model, both the SMB and HML coefficients were negative. The Medallion Fund shattered the basic finance rule that more risk equals higher returns.

Today Medallion is closed to outside capital, but Renaissance Technologies runs external funds such as RIEF and RIDGE for outside investors. However, as Zuckerman notes, these funds do not use the same strategy as Medallion. Their returns are ordinary. This underscores that Medallion’s strategy was a finely tuned system with strict capacity limits.

This article does not fully demystify Medallion’s extraordinary performance. Some might argue that Renaissance was simply the best market maker—that tiny gains on millions of trades accumulated into huge profits. But Medallion’s returns are so overwhelming that it is hard to attribute them solely to that explanation.

One thing is clear: in the context of the prevailing paradigm of the Efficient Market Hypothesis, Medallion is a shock on the scale of the Michelson–Morley experiment. It is a counterexample that cannot be ignored in the history of markets, and one that is well worth serious thought.