SANFORD J. GROSSMAN, QUANTITATIVE FINANCIAL STRATEGIES, INC.-JANUARY 1999|
Before he became highly successful in the investment management world, Sandy Grossman taught Finance at Stanford, Chicago, Princeton and The University of Pennsylvania where, in 1989, he became the Steinberg Trustee Professor of Finance
Before he became highly successful in the investment management world, Sandy Grossman taught Finance at Stanford, Chicago, Princeton and The University of Pennsylvania where, in 1989, he became the Steinberg Trustee Professor of Finance. He has been the recipient of numerous awards including, in 1987, the prestigious John Bates Clark Medal of the American Economic Association. His academic work has covered issues such as information in securities markets, market liquidity, transaction costs, program trading and optimal dynamic hedging. He was named President of the American Finance Association in 1994 and was a Public Director of the Chicago Board of Trade from 1992 to 1995. In 1988, he founded Quantitative Financial Strategies, Inc., Conshohocken, PA, a technical financial research firm set up to develop financial investment models using his pioneering work in quantitative finance, reflected in particular in his volume: The Informational Role of Prices (MIT Press, 1989). QFS has compiled an enviable track record compounding its clients' capital at an annual rate of 30% since 1993 in the currency trading program. Another program including currency, global stock index and global government bond futures, was established in 1998 and both programs returned approximately 44% in 1998. QFS and its affiliates currently manage $1.4 billion in client assets and $850 million in currency overlay programs. Sandy Grossman spoke with HFN publisher, Antoine Bernheim, in late January 1999.
Profile of Sanford J. Grossman
Born: July 21, 1953
Education: BA, MA and PhD in Economics from the University of Chicago
Family: Married to Naava; two children
Last vacation: Whistler, BC
Last book read: Wines of the Rhone Valley by Robert Parker
Hobbies: Skiing, wine and high quality food
Favorite quote: " To be a successful hedge fund you need to have at least as much fear as greed."
Q. What are the key factors behind your successful transition from academia to business?
A. I always had a sense, in my academic work, of combining intuition with precision and that is what I have tried to do in this business. There are many quantitatively oriented people who have gone into this business but I have always thought about economics in an intuitive way. I am not a mathematician or an engineer by training and I use mathematics only as a tool. I am an economist and what I bring to the table is a really unique and good understanding of economics as a discipline of people interacting.
Q. Could you describe the main components of the Informational Portfolio Strategies┐ trading system you developed?
A. Our basic philosophy is based on capital flows. When one country has a lot of investment opportunities but not much savings while another country has a surplus of savings but not many investment opportunities, capital should flow from the country with a surplus of savings to the country where savings are relatively scarce. That capital flow will be induced via the offering of high real returns in the country which capital needs to flow into. The return is going to come from interest and currency appreciation. So the first component of our trading system is an economic/econometric model that measures the impact of these capital flows. The model determines the expected returns for various paired positions by valuing the relative benefits of capital being invested from one country to another. The second component is a risk model of a portfolio of pairs which looks at market information to forecast the magnitude of ups and downs, the risk and volatility of a portfolio of pairs. The third component is a drawdown control process to maintain overall portfolio losses within a tolerable level.
Q. Could you describe how you were able to expand your analysis of currency yield differentials to a wider range of instruments such as stock indices and bonds?
A. The concept of pairing is very important in our philosophy and we brought this to the highest level in our macro hedge fund which we started in early 1998 and produced a 40% plus return. It took about two years of research to bring our ideas in trading currencies as pairs to trading stocks and bonds as pairs and crosses. Just as you might trade the Australian currency against the Canadian currency, you might trade Australian stocks against Canadian stocks. It is not an arbitrage trade, it is a risk trade but it is one that can be sensitive to capital flows in the same way that a currency can be. The correlation structure of stock markets is different than that of currencies. For instance, the UK equity market is highly correlated to the US equity market while Sterling and Dollar traded against the rest of the world are not anywhere as highly correlated between each other as the two equity markets are. As a result, the pairs we put on may be quite different in equities compared to currencies.
Q. Could you describe the ongoing research effort in your trading systems?
A. Our research is done in-house and focuses mainly on currency markets which is the research of a lifetime. I don't think we have solved every problem in the currency markets or that we understand every price fluctuation. I find the currency markets really exciting and interesting. I continue every day to think about them and understand exchange rates and capital flows better. We are just about to begin trading the Australian dollar and we spent quite a bit of research effort trying to understand how various aspects of the Australian economy work.
Q. Is there a typical duration of your trades and how has your trading been affected by the increase in volatility of capital flows?
A. Our approach is macroeconomic and duration in our currency trades tends to be as long as there are business cycle discrepancies. It is measured in quarters rather than days or weeks or months. Trend reversals are very often associated with an increase in volatility and we try to measure and forecast volatility. Volatility tends to make us reduce positions because we are always looking at the risk-adjusted return of a trade. So unless something changes the expected return enormously, if risk doubles we tend, roughly speaking, to halve our positions. Accordingly, in the last few months we have had smaller positions than we had previously.
Q. When and how would you use discretion over your trading systems?
A. We would do that primarily when there is some enormously unusual political event that we feel may not be completely captured in the kind of indicators we are looking at. It is a relatively rare phenomenon. Otherwise, we would only use discretion intra-day, for instance by delaying the initiation of a trade during the day. Our approach works best when economics are driving the markets. When politics are driving the markets, it works the least best. My models cannot forecast a presidential assassination or a revolution. Events like that can be disruptive although they happen rarely in the G-7 currencies we trade.
Q. What are your risk management tools and what is your approach to cutting losses?
A. In determining the size of the positions we put on, we go through the scenarios that might cause losses and map out the circumstances under which we might take the trade off. It is analogous to a game of chess where in making a move you try to evaluate a series of moves down the line. Fortunately, our problems are much simpler than chess problems: they are actually soluble. Risk control has to start with putting on a small enough position initially so as to be able to manage the risk of that position if something goes wrong. Secondly, we want to have a process in mind, before we put the position on, as to how and under what circumstances we are going to get out of it and we actually carry out that process. Through dynamic programming also called backward induction, the computer analyzes the decision tree of all the paths that might happen in the future and sets the positions in each of those circumstances to maximize the risk-adjusted return subject to meeting risk control targets.
Q. Could you describe how your organization functions?
A. We have a relatively small, tight-knit organization of 16 people. There are 5 people involved in trade execution who are implementing the trades of the models, which run in real time. We have 4 people in our research and systems department who are devoted to building our front, middle and back office systems which are fully integrated. We have designed all of our own risk management, accounting, portfolio analytics ourselves and probably have over 250,000 lines of C++ code. We have a separate back office/financial control area which controls and checks the operation of the trading area. I am myself primarily involved in managing, making sure that we have the infrastructure and quality control to maintain the standards we had when we were smaller. I only watch the markets as a supervisory function rather than a discretionary trading function, to make sure that nothing is happening that would make us want to change what we are doing.
Q. Where would you like to take your firm in the future?
A. I would like to continue to strengthen and diversify our infrastructure, so that we can accommodate more kinds of trading strategies and handle other products and trades and be comfortable with having twice as many assets under management. My objective is to maintain quality in performance results and client services and I don't want to grow very much unless I can be sure that we can continue to produce the same results. u