Cameron Passmore March 7, 2020 Advanced Investing Personal Wealth Renaissance Technologies Medallion Fund: An Exception to the Indexing Rule The Renaissance Technologies Medallion fund has reportedly returned an average annual 66% before fees from 1988–2018. That’s right, 30 years of 66% annual returns on average. So, why haven’t I put all my eggs in that basket and retired from Common Sense Investing? For one, it would not have given me the opportunity to opine on the many questions this sort of incredible performance has raised: Is the market really all that efficient? Do index funds still make sense? Should we all try replicating the trading success of the Medallion fund? To get to the root of these questions, we need to start with Jim Simons. To the extent that I can, I’m going to describe how he has apparently “solved the market”, and what his success means for you. Or, if you’re as intrigued as I am by the Medallion fund’s story, you’ll find the information I’m going to share and much more in Greg Zuckerman’s 2019 book, “The Man Who Solved the Market”. I highly recommend it. The Genius of Jim Simons If Jim Simons had never been a famous hedge fund manager, he still would have been a renowned code breaker and mathematician. Simons graduated at age 20 from the Massachusetts Institute of Technology (MIT) with a bachelor’s degree in mathematics. He completed his doctoral studies there as well, and then taught briefly at MIT and Harvard. He went on to work as a code breaker for the Institute for Defense Analyses. In both his academic and institutional careers, he made several meaningful breakthroughs. Simons’ contributions to mathematics cannot be overstated. At age 37, he received the American Mathematical Society’s Oswald Veblen Prize in Geometry. His award-winning work included the Chern-Simons theory, which now has tens of thousands of citations in academic papers. He also chaired the math department at the SUNY Stony Brook University, building it into one of the best math departments in the world. The time that Simons spent working with world-class cryptologists and mathematicians did more than build his hard skills. He also became uniquely qualified to recruit, manage, and work with some of the smartest people on the planet. Cryptologists Crack the Market’s Mysteries Throughout his time as a cryptologist and an academic, Simons maintained an interest in financial markets and trading. He had dabbled a bit, and even written a paper on the topic. But Simons did not leave academia to become a full-time trader until he was 40. At that time, he also recruited Leonard Baum to join him. Baum was one of Simons’ code-breaking colleagues and, like Simons, had also made significant mathematical contributions. Baum had developed the Baum-Welch algorithm for finding the unknown parameters of a hidden Markov model. Why do I mention this? Because it’s highly relevant to trading. Financial markets can be described as chains of hidden Markov models; so can speech recognition patterns. I’ll touch on that again in a moment. The First to Qualify as Quants Initially, Baum and Simons developed some models, but they also relied on human intuition to trade. This led to a mix of early successes and losses. Over time, Simons continued to attract some of the world’s smartest mathematicians to his operation. In 1988, his firm – by then called Renaissance Technologies – launched the Medallion fund. Other than attracting some of the smartest mathematicians in the world, Renaissance was doing something few other firms were doing at the time. They created and cleaned massive sets of historical data to feed their models. Remember, collecting data at that time meant recording data from written records to build out data sets to feed into a computer. In other words, they were among the first quant investors. Between their brilliant mathematicians and their data advantage, Medallion was able to build algorithms to identify patterns. Unlike traditional traders, they did not care why these patterns existed. They simply placed frequent short-term trades, augmented by lots of leverage, to try profiting from what their models deemed to be irregularities. Their models were so successful at generating winning trades that they eliminated any human intervention, to allow for trades that would make no sense at all to a human. They continued to tweak and develop this system, adding more brilliant mathematicians to do so. The Medallion Fund and Markov Models A big breakthrough came in the early 1990s when Renaissance recruited Robert Mercer and Peter Brown from IBM’s speech recognition group. Remember, speech recognition and financial markets are similarly described as chains of hidden Markov models – thus the connection. The Medallion fund had been doing well, but Mercer and Brown began what has become the greatest investing track record in history. Specifically, they developed a strategy to trade stocks that Medallion had been struggling with. Most of their success had come from trading futures. Mercer explained that their trades only won 50.75% of the time, but that’s all it takes when the fund is making millions of trades and employing leverage. When Mercer and Brown conquered stocks in 1995, Medallion took off. Implementing the Impossible As a success story, the Medallion fund is fascinating. The data seem to suggest that Simons and his team have been able to consistently uncover profitable patterns. Doing so consistently should not be possible in a highly efficient market. I can’t tell you exactly how they have done this. Not many people can, and Renaissance’s employees are sworn to legally binding secrecy. In an attempt to crack their code, well-known UCLA financial economist Bradford Cornell wrote a 2019 paper, “Medallion Fund: The Ultimate Counterexample?” He notes that $100 invested in Medallion in 1988 would be worth $398.7 million in 2018, and Medallion never had a negative return over this period. The fund maintained negative loadings to known factors, meaning that its success has not been driven by known risk premiums. Cornell comments: “In forty plus years of reading hundreds of papers on investment anomalies, including some that benefited from data snooping and ex-post selection bias, I have never seen any performance approaching that reported by Medallion.” Cornell built a market-timing model with perfect foresight, investing in stocks when their subsequent returns beat Treasury bills, and buying Treasury bills when they did not. Using monthly returns, this perfect-market timing model turned $100 into $331,288 from 1988–2018. So, even spot-on market-timing couldn’t amass as much as 10% of what a Medallion investor would have earned in the same timeframe. Cornell concludes: “To date, there is no adequate rational market explanation for this performance.” Can You Achieve Medallion’s Success? Now, let’s think about what Medallion’s huge success story means for you as an investor. I offer three takeaways here. First, it’s important to note that actual Medallion investors are as rare as the fund’s amazing performance. Medallion has been closed to outside investors since 2003. So, unless you were a fund employee, you would not have been able to partake in most of the gains. The fund also distributes its profits each year, capping the fund at $10 billion. These actions prevent one of the biggest problems successful active managers typically encounter. When their successful fund grows, it sows the seeds of its own destruction by having more capital than its strategies can handle. Second, while the success of Medallion is without question a challenge to market efficiency, it’s important to remember that market efficiency is a model, not reality. In other words, while real-life markets cannot be perfectly efficient, this does not mean just anyone can beat them. Let’s take a moment to think this through. Any time you make an active trade, you need to ask yourself who is on the other side. What do they know that you don’t? In the specific and rare case of the Medallion fund, they may truly have better information. They may have better data, and better ways to interpret it using sophisticated models developed by some of the world’s smartest people. But again, even if they have figured out a persistent long-term advantage, almost nobody except their employees can access it. There is a finite amount of these trades to go around. Medallion is arbitraging away what might be the last scraps of market inefficiency, and they are keeping the profits for doing so. Third, none of this makes Renaissance invincible. There are other quant investors trying to crack the same code, competing with Medallion to mop up price irregularities. Someday, this might trigger the skill paradox: If two equally skilled investors are competing for alpha, the winner will be determined by luck. Medallion seems to have had a skilled edge so far. But your guess is as good as mine on how long that will persist, especially in light of the fund’s recent uptick in publicity. Bottom line, unless I can get access to Medallion, I’ll stick to my index funds. Even if I could get access, betting on the fund’s persistence would make me nervous. My plan is to keep publishing these Common Sense Investing posts and Rational Reminder podcasts for non-Medallion investors to employ. After all, this is a considerably larger audience of investors. Share: Facebook Twitter LinkedIn Email CIRO Advisor Report
Investing For Your Future Episode 301 – Optimal Government Pension Claiming, and Life Well-Being w/ Prof. Meir Statman Apr 18, 2024 Investing For Your Future
Investing For Your Future Episode 300 – Abby Sussman: Financial Judgment and Decision Making Apr 11, 2024 Investing For Your Future
Investing For Your Future Episode 299: The Most Important Lessons in Investing Apr 4, 2024 Investing For Your Future