Factor Performance Across Market-Driven Scenarios

Abstract

We develop a methodology to model factor returns in scenarios which are yet to occur or do not correspond to past market conditions. We infer factor performance conditional on hypothetical market-driven scenarios, which are determined by a parsimonious number of underlying policy shocks. We derive regime-dependent corresponding security and asset class returns, which are functions of factors and policy drivers. The analysis is helpful in constructing robust portfolios designed for more stable performance across various scenarios, and can aid in framing and communicating portfolio risks.

Keywords: market-driven scenarios, factors, stress testing, scenario analysis, risk management, asset allocation, portfolio construction