•  
• Tweet
•  
•  
• 

Engineer of Contrarian Signals

"To persistently generate alpha, you need to give highly talented quantitative people state-of-the-art tools to design and test algorithms."

Opalesque Futures Intelligence: How did you get into futures trading?
Brad Paskewitz: For seven years I worked as an engineer. Then one day I received a call from a recruiter who said that a client, a bank, was looking for someone with my exact background. It did not make sense to me but I said to go ahead and submit my resume. That's how I was hired by Banque Indosuez in 1987 to develop and trade quantitative strategies in the futures market. As a commercial bank they could trade only financial futures because of regulation at the time. We traded a combination of stock indexes, interest rates and currencies.

OFI: What were the skills you brought from engineering?
BP: My core skill set was artificial intelligence, statistics and computer programming. So I was able to program and test trading models. After three years at Banque Indosuez I moved to Lehman Brothers, where I led a project with the goal to apply advanced machine learning methods to forecast futures market movements. This used artificial intelligence and was cutting edge at that time.

OFI: What is your approach?
BP: It is a contrarian strategy. The basic idea is that the market is ready to reverse course after going in one direction for a time. After the market rises, our models will move to sell; after the markets goes down, we'll buy. Essentially our models predict tops or bottoms.

OFI: So you're not a trend follower?
BP: The logic of our strategy is the exact opposite of trend following. Our returns are uncorrelated with the Barclay CTA Index and negatively correlated with many of the largest CTAs.

OFI: What's the time frame for your contrarian approach?
BP: Our models have been good at forecasting on a short-term basis. The trades average six trading days but can be as short as one day or as long as three weeks.

OFI: Is there a discretionary component?
BP: No, it is purely systematic. We take a strictly quantitative approach to all aspects of trading, from selecting the strategy to risk control. We've never had to override signals.

OFI: You say you use more than one model. Are they all contrarian?
BP: We use three contrarian models, which look for different patterns in market data. These are weighted equally in the portfolio. They're like three traders looking at market behavior from different angles to determine tops and bottoms.

OFI: Why do you trade only S&P 500 futures?
BP: Our first client wanted an S&P 500 trading program. S&P 500 futures have the advantage of being very liquid-$100 billion to $200 billion is traded per day in this market. Given that liquidity, our program has an estimated capacity of $1 billion.

OFI: Does the strategy work with other instruments?
BP: It works in some markets, not all. But it works well with most stock indexes around the world as well as individual equities and other select markets. We have been developing a separate program, expected to launch in month or two, that will trade other markets and models. These are all liquid futures markets.

OFI: Trend followers lost money in 2009. How did contrarianism work out?
BP: 2009 was a relatively tough year for us as well, in the sense that we did significantly worse than in previous years. We forecast short-term tops and bottoms, but the market had very long streaks going up for extended periods some days. Based on history, our models signaled that the market was likely to turn down. There is no precedent of a 60% gain in one year unless you go back to 1930s. It was an explosive, one-directional move.

OFI: But unlike many CTAs, you made a profit last year. How did that happen?
BP: There were times when market kept going straight up. But in between those times, there were up and down movements-zig zag price action. That's what we made money on. So it was not as bad as it might have been. Our ideal market is one where volatility zig zags. When the market persistently goes up, the model short sells and then market moves against us.

OFI: What do you expect for the business looking forward?
BP: To persistently generate alpha, you need to give highly talented quantitative people state- of-the-art tools to design and test algorithms. We've built an R&D platform that allows us to more easily create programs and run back tests. That's how we developed the new diversified program. We have an R&D team in China of nine full-time people, in addition to people in New Jersey. We're also automating operations.

OFI: How did you hire researchers in China?
BP: Our chief financial officer, Diana King, is Chinese and her brother is an entrepreneur in China. They know how to do things there.