Stable Linear-Time Optimization in Arbitrage Pricing Theory Models
The author’s take on the paper, Dr. Gordon Ritter:
“The computational method in the paper gives a research tool that will be appreciated by researchers working on equity market-neutral strategies. It’s less useful for strategies that are not factor neutral.
Of course, a real portfolio optimizer would be more complicated and have all kinds of constraints, more realistic t-costs, multiperiod, etc., but the method I describe there can be implemented in a few lines of R or python, and gives a way of quickly running a “cheap” (in the sense of computationally cheap) backtest with no t-cost or with quadratic t-costs, that can be a precursor to a more detailed simulation.”
Dr. Ritter presents an explicit formula for mean-variance optimization in the context of APT models (also called multi-factor models),and related generalizations with trading costs.
Dr. Ritter’s explicit formula has two desirable features:
1. the solutions are well-defined and numerically stable in the presence of approximate or exact colinearity in the design matrix, and 2. the computational complexity is (manifestly) linear with respect to the number of assets.
Stable Linear-Time Optimization in Arbitrage Pricing Theory Models by Dr. Gordon Ritter
Download The Full Paper
About Dr. Ritter:
Gordon Ritter completed his PhD in mathematical physics at Harvard University in 2007.
His publications while at Harvard were in quantum field theory, differential geometry, quantum computation and abstract algebra, including a well-known simplicity theorem for Kac-Moody groups and a mathematically rigorous treatment of Euclidean QFT on Riemannian manifolds. Prior to Harvard he earned his Bachelor’s degree with honors in Mathematics from the University of Chicago.
Dr. Ritter currently teaches at Cornell, Columbia, NYU, University of Chicago and the Baruch MFE program.
His academic research is on portfolio optimization and statistical machine learning; his finance publications include Risk, the Journal of Portfolio Management, Journal of Financial Data Science, European Journal of Operations Research, and others.
Named Buy-Side Quant of the Year in 2019 (researchers awarded the same honor in other years included Marcos Lopez de Prado, Alex Lipton, and Jean-Phillipe Bouchaud). In parallel with his teaching responsibilities, Dr. Ritter works full time in the industry. In 2019 he founded Ritter Alpha LP, a registered investment adviser running systematic absolute-return trading strategies across multiple asset classes and geographies, based on cutting-edge technology and rigorous applications of statistics and the scientific method to investment problems.
Prior to founding Ritter Alpha, he built a successful trading system from scratch at GSA Capital, a firm which won the Equity Market Neutral & Quantitative Strategies category at the Eurohedge awards four times. Prior to GSA, Dr. Ritter was a Vice President of Highbridge Capital Management (HCM) in addition a core member of the HCM statistical arbitrage group, a small team sharing full discretion over systematic trading models that generated billions in profit for investors and directed trillions of dollars of trades across global equities, futures and options. Several senior members of the HCM statistical arbitrage group later joined Gordon at Ritter Alpha. Lastly, in his spare time Gordon enjoys scuba diving in Hawaii, and is passionate about preservation of Native Hawaiian culture and marine ecosystems.
