The Equity Risk Premium: Unified Modeling and Python Automation
The US equity risk premium (ERP) is discussed in the context of broad-based market indices. The forward-looking ERP is estimated from options markets, which provide a market-based risk-neutral pdf for prices in the future. These pdf’s are transformed to real-world distributions via a well-known exponential tilt transformation. The tilt parameter is interpreted as the Coefficient of Relative Risk Aversion (CRRA) of a representative agent. Historically, I show CRRA in the range (2,4) Is a robust result and narrow this to my preferred range 3 ± 0.8 via a post-WW2 bootstrap. Two methods are given for the risk-neutral distribution estimation and other topics may be briefly touched upon.