LATEST NEWS

Mark Carhart is the Chief Investment Officer and a founding partner of Kepos Capital. Prior to founding Kepos in 2010, Mark was a Partner and the Co-Chief Investment Officer of the Quantitative Investment Strategies Group at Goldman Sachs Asset Management (GSAM), and before that he was a professor at USC and a Senior Fellow at Wharton.

Mark spoke on the “Rise of the Quants: The New Kings of Wall Street” panel at CAIS 2018. We sat down with him after the panel to discuss his career and to get his thoughts on the evolution of quant investing.

Q: You are known as an avid cyclist. What made you get into cycling and how does it help you professionally?

A: I fell in love with cycling in grad school because it provides such a powerful psychological release for me. I think it’s important for the brain to be distracted from focusing on life’s stresses, whatever they are. Cycling really helps me clear my mind and gets me centered for the day ahead. Others find that yoga, running, or even skydiving provides that same release, but for me it is cycling.

I also enjoy bringing cycling to others and have been tandem cycling with visually-impaired cyclists for many years. In 2013 I co-founded a not-for-profit organization called InTandem Cycling to provide cycling opportunities to the visually impaired and otherwise disabled. With the help of hundreds of sighted “captains,” we operate a fleet of tandem cycles in over 200 rides in New York City every year for our members to experience the love of cycling. So cycling is also my primary philanthropic activity.

Q: You used to teach at the University of Chicago and USC. How does working in academia compare to working on the investment side?

A: I’ve been out of academia for about 20 years now and I can say, without a doubt, that I far prefer my life today. Running investment strategies involves a lot less overhead and administration than teaching and publishing, so I spend more time conducting and consuming research on markets and also interacting with interesting people in our industry. I also think educational institutions tend to be very slow-moving about almost everything. In contrast, we get real-time feedback on the success of our investment strategies, and this feedback is also a primary source of our new research ideas.

Q: What was quant investing like when you started out and how has it evolved over the last few years?

A: When I started at Goldman Sachs in 1997, quant investing was very simple, slow and transparent. The sophisticated, automated, data-hungry computer algorithms that we think of as mainstream today really didn’t exist at that time. Even if we did have the data and the insight, the computing power necessary to build and execute such a system simply didn’t exist yet. Instead, I would say that “quant investing” was more about implementing a systematic decision-making process to remove emotions from what was basically fundamental insight. In effect, these programs protected us from our inherent behavioral biases. If the market consensus was the opposite of our position, we didn’t just abandon our investment strategy. Of course, there are other benefits of quantitative investing like testing your inputs vigorously and weighting variables appropriately, activities that the human brain does not do well.

Today, technology—data, computation power, and computational methods—have advanced to the point where it is possible to build much more sophisticated algorithms that attempt to decipher every market move into noise and information. Our models today are less about mechanizing fundamental data and more about inferring subtle behaviorally driven dislocations.  Of course, you still can’t completely remove behavioral bias from the trading algorithms–they are programmed and operated by humans after all–but our systems manage far more decisions today than they did 20 years ago.

Q: What’s next for quant investing?

A: Markets are still too complex to create a fully automated system that both uncovers and trades dislocations. There are too many variables–not to mentions competition among traders—that influence market behavior. That’s why it’s critical to employ human insight alongside machines to help define what it is the algorithms are trying to capture.

I’m seeing an explosion in interest around risk-factor investing, which basically involves building slow-moving, transparent trading models to deliver documented risk premia in markets. I believe there is also still opportunity in statistical arbitrage strategies, which take advantage of short-term movements that aren’t consistent with the underlying fundamentals. The larger risk that quants–and frankly all investors–worry about is crowding as strategies are successful and become overpopulated with investors. Time will tell if recent interest in quantitative strategies will develop into crowded factors.