Intelligent Portfolio Rebalancing
Introducing a statistical portfolio rebalancing test based on resampled portfolio optimization.
Portfolio rebalancing is often handled in an ad hoc way, despite its crucial importance for any investment portfolio.
Chapter 6 of the Portfolio Construction and Risk Management book1 introduces a new method for assessing the similarity between the current portfolio and a desired portfolio that we are considering rebalancing towards.
Portfolio Construction and Risk Management Book
The PDF version of the Portfolio Construction and Risk Management book is freely available online at the bot…
Hence, instead of just rebalancing at fixed frequencies, we can statistically assess how much our current portfolio deviates from the desired one and potentially rebalance at times that are favorable in terms of transaction costs, for example, as new money comes into the portfolio or execution is particularly cheap.
The new rebalancing method builds on the Resampled Portfolio Stacking2 approach, which is also introduced in Chapter 6 of the Portfolio Construction and Risk Management book. Compared to Exposure Stacking, which was introduced by Kristensen and Vorobets (2024)3, Resampled Portfolio Stacking allows us to use other targets than the portfolio exposures. For example, we can use the marginal risk contributions, the marginal return contributions, or a ratio between the elements of the two vectors.
Portfolio Optimization and Parameter Uncertainty
Portfolio optimization remains one of the most researched areas in quantitative investment management. Yet it is also one of the most criticized aspects due to its inherent sensitivity to parameter estimates or more generally market model uncertainty.
The benefit of the new method is that it allows us to incorporate differences in risk contributions, including complex diversification interactions for any investment risk measure.
The fundamental idea is to perform B resampled portfolio optimizations having the same characteristics as our target portfolio4. This could, for example, be the same overall risk and tracking error targets in addition to the natural constraint of having the same investment restrictions.
The new rebalancing method is thoroughly described in the Portfolio Construction and Risk Management book. The book also includes accompanying Python code5 with practical examples of how you can use the new rebalancing approach and much more related to Entropy Pooling and fully general CVaR optimization6.
Entropy Pooling and CVaR Portfolio Optimization in Python
Entropy Pooling (EP) is a very powerful method for implementing subjective views and performing stress-tests for fully general investment distributions. It can be seen as a generalization of the Black-Litterman model that does not rely on the oversimplifying assumptions of normally distributed …
Portfolio Construction and Risk Management Book: https://antonvorobets.substack.com/p/pcrm-book
Resampled Portfolio Stacking: https://antonvorobets.substack.com/p/resampled-portfolio-stacking
Portfolio Optimization and Parameter Uncertainty SSRN article: https://ssrn.com/abstract=4709317
Resampled Portfolio Stacking: https://antonvorobets.substack.com/p/resampled-portfolio-stacking
Portfolio Construction and Risk Management Book’s Python code: https://github.com/fortitudo-tech/pcrm-book
Entropy Pooling and CVaR Portfolio Optimization in Python Substack post: https://antonvorobets.substack.com/p/entropy-pooling-and-cvar-portfolio-optimization-in-python-ffed736a8347