"Non-Normality Facts and Fallacies"

The global downturn in the fall of 2008 and the failure of many sophisticated funds raised questions of the effectiveness of standard quantitative techniques and basic theories of modern finance.   Market crashes are often characterized as twenty standard deviation black swan events, which are cited as examples of the failure of the normal model and classical statistics.  This provides a convenient excuse for poor performance and a corresponding demand for a complex solution.   An important question is whether it is possible or even desirable to eliminate downside risk?  More importantly, are complex strategies really a useful antidote?

Similar to generals planning for the last war they fought, a crisis often leads to overly specific solutions that may have little value for the future.  David’s well informed paper presents a valuable cautionary tale that focuses on a better understanding of statistical reality.  Overly complex modeling in the context of estimation error often results in poor out-of-sample performance.  Lessons that ignore firm foundational principles may do more harm than good. 

Visit the Society of Quantitative Analysts for more information and registration.

Dr. David Esch, Director of Research, earned his PhD in statistics from Harvard University. He has specialized knowledge in mathematical statistics, numerical analysis and computation, Bayesian statistics, and econometrics. In addition to his work in finance, he has published statistical resarch in many other areas, including image reconstruction for low-photon-count telescope data, psychological test assessment, and various other areas of health care research. He is the author of "Non Normality Facts and Fallacies," (Journal Of Investment Management 1st quarter 2010). A native of Boston, he attended Harvard College before obtaining a master's degree in mathematics from Boston University.