Comparing Multi-Asset Allocation Models: Mean-Variance, Kelly-CVaR, Black-Litterman, and Risk Parity
Summary
This report surveys several approaches to allocating across asset classes: macro and cycle-based fundamentals, mean-variance optimization, Kelly-CVaR, Black-Litterman, and risk parity. It describes a macro model that separates directional forecasts from expected-return estimates, a mean-variance variant using Kaufman’s adaptive moving average to estimate returns, and a Kelly-CVaR approach that seeks growth subject to a limit on average portfolio loss. Black-Litterman combines equilibrium return assumptions with investor views using Bayesian updating. Risk parity allocates according to risk contributions, with variants for leverage, risk budgets, and factor exposures.
The reported backtests give performance figures for several approaches, including a downside-risk parity variant that substitutes downside deviation for standard deviation. These results are specific to the report’s tests and do not establish general performance. The report also highlights mean-variance instability when inputs change slightly and the subjectivity and complexity of Black-Litterman views and uncertainty settings.
Key ideas
- The report compares fundamental, mean-variance, Kelly-CVaR, Black-Litterman, and risk-parity allocation methods.
- A macro model separates forecasts of return direction from forecasts of return magnitude.
- Kelly-CVaR combines growth-oriented sizing with a constraint on average portfolio losses.
- Black-Litterman updates equilibrium return assumptions with investor views through Bayesian inference.
- Risk parity balances portfolio risk contributions, while downside-risk parity uses downside variability as its risk measure.
- The reported backtests are model- and test-specific, and several methods depend on sensitive or subjective inputs.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.