Discrete signature and its application to finance

TL;DR

This paper introduces flat discrete signatures and discrete signatures, enabling direct computation from discrete data, with applications in financial time series analysis, notably improving stock market estimations with fewer data points.

Contribution

The paper proposes a novel discretized signature method that directly processes discrete data and captures quadratic variation, enhancing financial data analysis.

Findings

Flat discrete signatures can represent quadratic variation.

Discrete signatures emphasize recent data over older data.

Effective stock market estimation with fewer data points.

Abstract

Signatures, one of the key concepts of rough path theory, have recently gained prominence as a means to find appropriate feature sets in machine learning systems. In this paper, in order to compute signatures directly from discrete data without going through the transformation to continuous data, we introduced a discretized version of signatures, called "flat discrete signatures". We showed that the flat discrete signatures can represent the quadratic variation that has a high relevance in financial applications. We also introduced the concept of "discrete signatures" that is a generalization of "flat discrete signatures". This concept is defined to reflect the fact that data closer to the current time is more important than older data, and is expected to be applied to time series analysis. As an application of discrete signatures, we took up a stock market related problem and succeeded in performing a good estimation with fewer data points than before.