Probability versus Ambiguity: Perceiving and Predicting Markets

Looking at markets as a social question can help participants better navigate times of uncertainty.

By Denise K. Shull, President, Trader Psyches

 

Looking at markets as a social question can help participants better navigate times of uncertainty.

What question does a market technician want to answer? What about a global macro hedge fund or fundamental analyst? Regardless of their substantial differences, does not each of these individuals’ methods attempt to accurately predict future asset prices via the lens of mathematics and probability?

Yet, can numbers really tell the whole story? Is not the more basic question, the one the numbers try to reflect, the price that another human being – inevitably also relying on numbers – will be willing to pay in the future? What if we asked that question directly?

 

Neuroeconomics and Price

The new science of neuroeconomics, the study of risk-decisions from the point of view of brain processes, suggests a reward exists for doing so. Recent experimental evidence even points to the ability to predict another person’s behavior as more helpful in reading markets than pure probabilistic thinking. Think of the best market players you know – whether they run a billion dollar hedge fund or trade their own capital. If you listen over their shoulders you hear them speak to the screens as if the monitors can hear. Implicitly their perceptions are not about the bars, lines and formulas, but about the people behind the flashing electrons.

Imagine that hypothetically ABC Fund strategically includes human behavior analytics and predictions in devising trading ideas and risk strategies. In the summer of 2008, no algorithm for massaging historical data could ever have foretold that if a major investment bank like Lehman Brothers went bankrupt, AIG would also be severely wounded. ABC Fund on the other hand, looking at the drama of Bear Stearns’ demise, and the pervasive chatter regarding who else might fall, would have noted the cries of moral hazard and the potential limits of then-Treasury Secretary Henry Paulson’s and Federal Reserve Chairman Ben Bernanke’s appetite for bank rescues. “A bank fails” would have topped the human scenarios list and ABC would have either been flat financials as a risk management strategy or short as a tactical bet. Conversely, numbers alone very well may have justified being long Lehman or AIG given the value prices at which they were trading.

 

Known Versus Unknown Probabilities

So if our brains can do it, what prevents us from writing a winning trading algorithm
to do it? It appears to be a result of our perceptional abilities in ambiguous situations – or what the decision scientists define as “uncertain outcomes with uncertain probabilities.” In 1921, Frank H. Knight wrote in Risk, Uncertainty and Profit , that markets are always ambiguous and what we call risk or uncertain outcomes with known probabilities is but an illusion. Predicting future prices is not like the algebra we did in ninth grade. The market is not hiding a set of absolute numbers with immutable relationships where the answer emerges from solving for the missing factor.

Yet behavioral finance also confirms that humans consistently prefer known, quantifiable risks to ambiguity. This explains our affection for the quantitative, but omits that we are simultaneously quite adept at making judgments in uncertain circumstances. In other words, our innate ability to interpret ambiguity gives us information that numbers alone will never find. Brain imaging – pictures of brains making decisions – indicates that we perform different mental gymnastics when we realize we lack all the factors. We seem to use the context to fill in the blanks when we consciously or unconsciously sense that the equation cannot be solved for X.

Should we abandon all efforts to leverage projected probabilities? Of course we should not. Meteorologists do not quit because they cannot say for sure where a hurricane will make landfall. Projected probabilities offer a valuable tool, but we must admit that they are now clearly insufficient in market storms.

 

Rethink Your Approach

The violent volatility that began in August 2007 demands that we refine our approach. The next level lies in the convergence of models and algorithms with “what if” thinking about the human game. The systematic merger of such qualitative insight with traditional quantitative approaches will open up the market to more accurate predictions about “fat tails” and “black swans.”

Andrew Lo, director of the M.I.T. Laboratory for Financial Engineering, has proposed the adaptive markets hypothesis. It is based on the idea that an ever-evolving ecosystem is a better model for markets than the efficient markets hypothesis. This idea makes complete sense given that biological humans would naturally produce biologically behaving markets.

The future of both outsized returns and best-of-breed risk management should go one step further. Traders should systematically incorporate the idea of a social markets hypothesis into their approaches. Strategically magnifying the focus on the foundational question will expand the universe of strategies and risk management options by an order of magnitude that no mere chip-set can reach.

 

Denise K. Shull has traded since 1994 and founded Trader Psyches as a consultancy focusing on the psychology of trading in 2003. She invites feedback via denise@traderpsyches.com.

Back to Spring 2009 Issue Home