O'Reilly logo

Finding Alphas: A Quantitative Approach to Building Trading Strategies by Igor Tulchinsky

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

16 Techniques for Improving the Robustness of Alphas

By Michael Kozlov1

The main goal of alpha research is to predict and outperform the market. However, most investors desire not only returns, but also low risk exposure, predictability of trading results, etc. All these conditions lead us to definitions of a robust alpha.

A robust alpha should have the following properties:

  1. Invariance under modification of traded universe: An alpha should reflect a statistical property that is independent of the choice of the traded universe. In other words, we wouldn’t like our alpha to depend on some specific instrument or group of instruments. This constraint is frequently imposed by the market as the result of regulatory changes, customer preferences, liquidity decrease, short bans, etc.
  2. Robustness to extreme market conditions: An alpha should not have a sharp decline in any of the benchmarks. The most common benchmarks used for alpha performance measurement are information ratio, maximum drawdown, return, etc.

SIMPLE METHODS FOR ROBUSTNESS IMPROVEMENT

We will describe below some useful methods for improving the robustness of an alpha.

  1. Ordering methods
    1. Ranking: Ranking is an operation that replaces a vector’s elements with their ranks (from 0 to vector_size-1), under some sorting. Usually all vector’s elements are also divided by vector_size-1, to be in the interval [0,1]. If two values are the same, they are supposed to have the same rank that equals to the average of their corresponding ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required