On Wednesday March 26, 2014 at 11:15am Marcos Lopez de Prado and David H. Bailey, two of the bloggers on this site, jointly presented a talk How to spot backtest overfitting at the Battle of the Quants meeting in New York City.
The Battle of the Quants conferences, organized by Bartt C. Kellerman of Global Capital Acquisition, are held regularly in New York City, Shanghai and London. They gather together academicians, asset managers and other professionals in the area of quantitative finance and investment.
The viewgraphs of the talk by Lopez de Prado and Bailey are available here. Their talk discusses backtest overfitting — why it is such a widespread problem in the field, how to spot it and some tools to help prevent it. The talk also discusses, more generally, the misuse of mathematical-statistical methods in finance (overfitting, questionable charts and graphs, misuse of probability theory and stochastic methods, misleading technical jargon, etc.), and why it is so important for professionals in mathematical-quantitative finance to be more open about such problems in the field, thus protecting its good name.
The technical material presented in the conference is based on these two papers:
- David H. Bailey, Jonathan M. Borwein, Marcos Lopez de Prado and Qiji Jim Zhu, Pseudo-mathematics and financial charlatanism: The effects of backtest over fitting on out-of-sample performance”, Notices of the American Mathematical Society, to appear, May 2014.
- David H. Bailey, Jonathan M. Borwein, Marcos Lopez de Prado and Qiji Jim Zhu, The probability of backtest overfitting, manuscript, 12 Feb 2014.