Few people bothered to see Trouble with the Curve, a recent baseball movie starring Clint Eastwood and Amy Adams, and most critics didn’t like it. I did see the movie, and without giving away the plot, it is fair to say that the film is a cry against quantitative analysis in sports. Eastwood plays an aging baseball scout with failing eyesight who has to rely on his daughter (Amy Adams) to evaluate the home office’s number one prospect. In the end, all of the number crunching in the world can’t come up with a better analysis than Eastwood, who can hear the sound of the bat on the ball and subsequently knows better than to sign the prospect. It was impossible to watch this movie without thinking of last year’s hit film Moneyball.
Moneyball is the story of Billy Beane, the GM of the Oakland Athletics, who in 2002 turned the baseball world upside down by using quantitative player analysis — called sabermetrics — to take a team with the lowest salaries in baseball to the playoffs. Lest anyone think that Beane was a flash in the pan, he just won the Sporting News 2012 Baseball Executive of the Year award for another amazing season. The film makes fun of traditional methods of scouting baseball talent, and shows how computer analysis can lead to a more ‘optimized’ approach to tactics.
Obviously, Trouble with the Curve and Moneyball tell two very different tales about how to evaluate player talent in sports. Where Moneyball celebrates quantitative decision making, Trouble with the Curve warns that relying too heavily on quantitative methods can lead to mistakes, and touts qualitative decision making instead.
All of this brings me to my latest speaking engagement at the IMCA (Investment Management Consultants Association) Special Winter Conference in Scottsdale, Arizona. The theme of the conference was managing money in the Post-Modern Portfolio Theory world, with session titles like, “The Evolution of Alternative Investing: New Thinking, New Approaches, and New Opportunities,” “Creating Stronger, More Resilient Portfolios to Counterbalance Risk and Enhance Modern Portfolio Theory Construction,” and “Managing Volatility with Post-Modern Portfolio Theory.” Conference attendees left armed with all of the statistical evidence they needed to show how the correlations from investing in alternative strategies lowered portfolio volatility and increased the chances of increasing risk adjusted returns.
Unfortunately, few in the audience had any idea what caused a variety of hedge fund strategies to have low correlations in the past, and therefore could only assume that they would do so again in the future. This disregard for knowing why asset classes performed the way they did in the past in favor of simply plugging data into models is the legacy left to us by modern finance. The stochastic approach to decision making seems to be accurate, scientific, systematic, and objective. After all, when plugging historical data into an optimizing model you don’t need to know why the returns occurred the way they did. You only need to know that the output of the model is statistically significant. That’s why we at Pinnacle believe that quantitative approaches to decision making are invaluable tools for determining market valuation, portfolio volatility, and the future direction of a variety of markets. Sauro Locatelli, our quantitative specialist, delights in tweaking a variety of data to give us objective and systematic guidance in our portfolio construction.
Nevertheless, I still see myself much like Clint Eastwood in Trouble with the Curve: While I appreciate the biases and heuristics that cause us to make mistakes as human decision makers, I don’t believe we should be relying solely on quantitative methods to make portfolio decisions. Our insistence on adding our collective judgment, experience, and right-brained intuition to our portfolio construction process is a lousy business decision, at least in terms of our productivity. The financial markets are confusing and our quant models tell us what to do with an assurance that is convincing to all concerned. It is hard to argue when the answers from the model are “statistically significant.”
The problem is that while we have unbelievable amounts of data to work with, none of it will help us predict whether Congress will go over the fiscal cliff… or the impact of the Federal Reserve buying $40 billion per month in Agency mortgage-backed securities… or the European Central Bank’s Ongoing Monetary Transaction Program… or the impact of five-years of zero interest rate policy at the Fed… or any of the other unprecedented financial conundrums we are currently faced with. We just don’t have the historical data to populate our model, and so we are left to fend for ourselves. The only way science can make us feel better about our forecast is if we choose to forget that there are problems with the data and with the assumptions we make in our model.
Given what I’ve seen at the conferences I attend and the audiences I speak to, just about everyone is willing to do exactly that. But if you were to sit in on a meeting of the Pinnacle Investment Team, you would see that for a significant part of the time, we have our figurative eyes closed as we listen to the sound of the bat on the ball and the pop of a fastball hitting the catcher’s mitt. Our collective gut instinct matters, even if everyone in the investment industry shudders to say it out loud.
Copyright: 104paul / 123RF Stock Photo