Machine learning based forecasting of significant daily returns in foreign exchange markets

TL;DR

This study explores the use of modern machine learning and outlier detection techniques to forecast significant fluctuations in currency exchange rates, demonstrating superior performance over traditional methods across multiple currency pairs.

Contribution

Introduces the novel application of outlier detection methods, especially PKDE, for forecasting significant currency exchange rate fluctuations, outperforming traditional approaches.

Findings

Outlier detection methods outperform traditional machine learning techniques.

PKDE outperforms other outlier detection methods in this context.

Results are consistent across different currency pairs and time horizons.

Abstract

Asset value forecasting has always attracted an enormous amount of interest among researchers in quantitative analysis. The advent of modern machine learning models has introduced new tools to tackle this classical problem. In this paper, we apply machine learning algorithms to hitherto unexplored question of forecasting instances of significant fluctuations in currency exchange rates. We perform analysis of nine modern machine learning algorithms using data on four major currency pairs over a 10 year period. A key contribution is the novel use of outlier detection methods for this purpose. Numerical experiments show that outlier detection methods substantially outperform traditional machine learning and finance techniques. In addition, we show that a recently proposed new outlier detection method PKDE produces best overall results. Our findings hold across different currency pairs, significance levels, and time horizons indicating the robustness of the proposed method.