Asynchronous Data and Serial Correlation in Financial Time Series

In our increasingly integrated global economy, news in one time zone will often have a significant impact on markets in other time zones. Tracking cause and effect as it moves around the globe can be difficult. If markets are closed in one time zone, it may appear that two closely related events occur on different days. This, in turn, can significantly impact a number of risk measures. 

In this paper, we begin by looking at actual market data to see how asynchronous data can impact traditional measures of risk. Next we examine how to quantify the degree to which different markets are linked using serial correlation. We then explore how risk measures can be corrected for asynchronous data. A separate section is devoted specifically to incremental risk measures. Finally we discuss some practical issues to consider when reporting risk statistics.