Integration of the Quadratic Optimizer into the Python Backtester (Code Dump of the updated Russian Doll Engine)
In the previous work, we introduce ledoit-wolf constant correlation shrinkage methods:
We also introduced the quadratic optimization problem for the n-alpha (extension of the n-asset) problem with linear transaction costs.
This post integrates the different techniques into our advanced, proprietary Python backtesting module (Russian Doll engine). Our updated Russian Doll engine code should now allow us to seamlessly go from arbitrary number of single strategy portfolios to multi-strategy portfolios with different optimization options. More to be added.
The alphas and benchmark datasets used to run the code in the example are given in this post:
