Learning low-frequency temporal patterns for quantitative trading

TL;DR

This paper presents a modular online machine learning framework that learns low-frequency financial signals from time series data, combining unsupervised and supervised learning techniques to improve trading predictions.

Contribution

It introduces a novel combination of autoencoders, restricted Boltzmann machines, and online neural networks for low-frequency financial data analysis, with rigorous overfitting assessment.

Findings

Effective learning of low-frequency signals from daily, weekly, and quarterly data

Robust backtest results with reduced overfitting risk

Insights into market phenomenology and data analysis value

Abstract

We consider the viability of a modularised mechanistic online machine learning framework to learn signals in low-frequency financial time series data. The framework is proved on daily sampled closing time-series data from JSE equity markets. The input patterns are vectors of pre-processed sequences of daily, weekly and monthly or quarterly sampled feature changes. The data processing is split into a batch processed step where features are learnt using a stacked autoencoder via unsupervised learning, and then both batch and online supervised learning are carried out using these learnt features, with the output being a point prediction of measured time-series feature fluctuations. Weight initializations are implemented with restricted Boltzmann machine pre-training, and variance based initializations. Historical simulations are then run using an online feedforward neural network initialised with the weights from the batch training and validation step. The validity of results are considered under a rigorous assessment of backtest overfitting using both combinatorially symmetrical cross validation and probabilistic and deflated Sharpe ratios. Results are used to develop a view on the phenomenology of financial markets and the value of complex historical data-analysis for trading under the unstable adaptive dynamics that characterise financial markets.