Neural Functionally Generated Portfolios

TL;DR

This paper presents a neural network approach to learn the generating function of functionally generated portfolios from market data, enhancing flexibility and performance over classical methods.

Contribution

It introduces a neural network framework for learning FGP generating functions, allowing data-driven adaptation while maintaining key financial properties.

Findings

Neural FGPs outperform classical benchmarks in investment returns.

The approach learns market dynamics without estimating drifts or covariances.

Neural FGPs preserve self-financing and pathwise properties.

Abstract

We introduce a novel neural-network-based approach to learning the generating function $G(\cdot)$ of a functionally generated portfolio (FGP) from synthetic or real market data. In the neural network setting, the generating function is represented as $G_{\theta}(\cdot)$, where $\theta$ is an iterable neural network parameter vector, and $G_{\theta}(\cdot)$ is trained to maximise investment return relative to the market portfolio. We compare the performance of the Neural FGP approach against classical FGP benchmarks. FGPs provide a robust alternative to classical portfolio optimisation by bypassing the need to estimate drifts or covariances. The neural FGP framework extends this by introducing flexibility in the design of the generating function, enabling it to learn from market dynamics while preserving self-financing and pathwise decomposition properties.