Featured Image

The Most Significant Change in FX Rate Correlations Since the Beginning of the Century - Observations Pre/Post Brexit

Risk, Performance, and Reporting

By Ivan Mitov  |  August 3, 2016

After the EU referendum vote last month, the British Pound suffered significant depreciation vs. the US dollar. The GBP/USD rate slumped by almost 8% in just a single day (June 24, 2016), which is an all-time record for this currency pair. On the other hand, the Japanese Yen appreciated vs. the US Dollar and the USD/JPY rate decreased by 3.4%. This is a significant joint extreme for these two currency pairs. 

So what does that mean for the correlation between the two rates? And what was the Brexit effect on the major FX Rates correlations? Today's post leverages Cognityrisk solution to look across market risk analytics in the above-referenced currency pairs, and highlights what appears to be the most significant correlation change since the beginning of the century.
 
The following chart provides the Pre- and Post- Brexit daily correlations* for seven major currency pairs, namely AUD/USD, EUR/USD, GBP/USD, NZD/USD, USD/CAD, USD/CHF and USD/JPY.
 
Picture1.png

Figure 1: Pre- and Post- Brexit correlations
* The rolling daily correlation is estimated using a historical time window of 1 year daily observations and an EWMA decay factor, which provides good out-of-sample estimates.

It is clear that there are significant changes in correlations for three currency pairs: the three main European currencies vs the Japanese yen. The other pairs are not as affected. Figure two provides comparisons within Cognity of the correlation matrixes for the two dates: June 23, and June 24, 2016.

Picture2.png
Figure 2: Cognity correlations comparison report – Pre- and Post- Brexit correlations

With Cognity we had the option to highlight large differences. In this case only differences larger than 15% are highlighted. We see the same three pairs – EURUSD vs USDJPY, GBPUSD vs USDJPY and USDCHF vs USDJPY. The most significant change is in the GBPUSD vs USDJPY pair.

Actually this is the most significant correlation change since the beginning of the century. The following figure provides the daily correlation changes in the GBPUSD vs USDJPY currency pair.

Picture3.png
Figure 3: Daily Change in the historical correlation between GBPUSD and USDJPY rates.

The same holds also for the following EURUSD vs USDJPY pair.
Picture4-1.png
Figure 4: Daily Change in the historical correlation between EURUSD and USDJPY rates.

Post-Lehman Similarities

Clearly the major impact and change in the FX markets correlations is driven by the British Pound's large depreciation vs. the US Dollar.  And on the opposite site – the Japanese Yen appreciation vs. the US dollar. It should be noted that similar correlation changes were observed after the Lehman collapse in 2008.

Picture5-1.png
Figure 5: GBPUSD, USDJPY rates and their rolling daily correlation.

Looking at the data above, the long-term daily correlation for the period January 2001 – July 2016 between those two FX rates is slightly negative (approx. -11%) with periods of significant negative and significant positive correlation. The interesting observation here is that the Brexit event entirely flipped the correlation from a relatively large period of negative correlation to a period of high positive correlation comparable to the levels of post Lehman collapse in December 2008.

This article originally ran on BISAM's Insight blog

Ivan Mitov

VP, Head of Risk Research

Mr. Ivan Mitov is VP, Head of Risk Research at FactSet. In this role, he leads FactSet’s risk and quantitative analytics research and software development and builds and manages a team of quantitative researchers, developers, and analysts. Mr. Mitov earned a bachelor’s degree in applied mathematics, a master’s degree in mathematical modeling in economics, and Ph.D. in probability theory and mathematical statistics, all from Sofia University St. Kliment Ohridski.

Comments

The information contained in this article is not investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.