Compensating asynchrony effects in the calculation of financial correlations

TL;DR

This paper introduces a method to correct statistical errors caused by asynchrony in financial time series, helping to better understand the Epps effect by quantifying and compensating for it using only trading data.

Contribution

The paper proposes a novel compensation method for asynchronous data errors in correlation calculations, addressing a key factor in the Epps effect in financial markets.

Findings

Statistical errors significantly contribute to the Epps effect.

The method effectively quantifies and compensates for asynchrony-induced correlation biases.

Application to real financial data demonstrates improved correlation estimates.

Abstract

We present a method to compensate statistical errors in the calculation of correlations on asynchronous time series. The method is based on the assumption of an underlying time series. We set up a model and apply it to financial data to examine the decrease of calculated correlations towards smaller return intervals (Epps effect). We show that this statistical effect is a major cause of the Epps effect. Hence, we are able to quantify and to compensate it using only trading prices and trading times.