Introduction to Style Analysis
Thursday, January 24, 2019, 03:00pm - 04:00pm
Contact Chris Bartlett, Andrew Grauberg

Hedge Fund Style Analysis

Though the term 'style analysis' has been known among traditional portfolio managers for ages, its underlying math is totally different when it comes to hedge funds. The traditional style analysis implies a decomposition of asset contributing styles (e.g. a strategy or a category) into a predefined set of investment styles, e.g. value, growth etc. It is hardly doable for hedge funds, because of their non-transparency, a huge diversification of trading strategies and investment vehicles. In other words, it could difficult, or even impossible, to explain hedge fund returns in terms of a pre-defined and fixed set of factors (or categories), because these factors may be not correlated with the fund being analyzed.

This tutorial discusses the main problems of hedge fund style analysis and guides you through the steps of using Risk Shell to build reliable factor models that can be further used for stress testing, manager exposure analysis or style drift analysis.

Hedge Fund Style Analysis and Factor Analysis

  • Learn the basics of regression factor models and style analysis.
  • Understand why the common direct regression is hardly applicable to hedge fund style analysis.
  • Understand why relying on the 'labeled' hedge fund manager categories may be misleading.
  • Implicit and Explicit macro factor models.
  • Introduction to stepwise regression, LASSO, Elastic-net, AIC and the Kalman filter.

Static and Dynamic Style Analysis for Hedge Funds - Step-by-step Tutorial

  • Risk Shell Static and Dynamic style analysis. Application controls.
  • Working with factor subsets.
  • Risk Shell Dynamic Style Analysis charting.
  • The Kalman filter: pros and cons.

Potential Audience

Institutional portfolio managers, hedge FoF and multi-asset portfolio managers, CIOs, advanced family offices.

Location online
EST time zone. For existing customers only
Registrations are now closed


We use cookies to improve our website. By continuing to use this website, you are giving consent to cookies being used. More details…