These codes have been used in the article Signature Methods in Stochastic Portfolio Theory by Christa Cuchiero and Janka Möller.

For citations:

Cuchiero, C.; Möller, J. Signature Methods in Stochastic Portfolio Theory.

@article{CM:23,
    title={{Signature Methods in Stochastic Portfolio Theory}}, 
    author={Cuchiero, C. and Möller, J.},
    journal={Preprint arXiv:2310.02322},
    year={2023}}

Code to compute the

• JL-signatuture using the memory efficient algorithm proposed in our article
• Randomized Signature

A collection of classes used for market-simulation an the computation of the (theoretical) growth-optimal portfolio.

A collection of basic functions.

Functions used to optimize signature portfolios using real market data.

Functions used to traine the signature portfolios in the simulated markets and comparing their performance to the the theoretical growth-optimal portfolio.