What is RORO Index? It's HSBC's CERSI!

Wednesday, 29 May 2013 15:17

One of the features of our monthly FX correlation reviews since early 2012 is the colorful two-dimensional heatmap of correlations which can be summarized in what we called CERSI, an abbreviation for Currency Exchange Rate Strength Index.   The definition of the index has been published in the ForexAutomaton on March 24, 2012. We clarified the meaning of this index by complementing its definition with the following interpretation:  

"Bouts of panic in the markets, such as seen in Fall 2008, during the "Flash Crash" of 2010, or after the US credit rating downgrade in 2011, are characterized not only by high volatility, but also by unusually high degree of correlation among the instruments. The concept of portfolio diversification collapses, as the universe of financial instruments degenerates into risky assets and safe haven assets. Mathematically, this degeneration does not follow from a rise in volatility, and represents an independent quantitative aspect of the panic phenomenon. Being independent of volatility by construction, CERCSI is designed to isolate and monitor this particular aspect."

At least since November 2010, HSBC has been publishing its RORO Index, the first mention of which known to me is dated by Novemer 10, 2010, and is made in a Currency Quant Special Issue, published by HSBC.

HSBC monitors not 14, but 34 assets:

FX cross-market correlation heatmaps, as published on this site on February 23, 2012 (left, describing the panic of October 2008) and on April 1, 2012, describing market conditions of March 2012.

A snapshot from HSBC research circulating on the net.

The essence of the interpretation framework is very similar to ours, even though the technical details are quite different, making RORO, in our opinion, more difficult to interpret: HSBC's RORO is the variance of the first principal component in the PCA decomposition of the return time series. What matters more for the problem, in our opinion, is the degree of inequality among the PCA eigenvalues. Although CERCSI has been developed independently, we are happy to acknowledge the HSBC work which we only today became aware of.

  Based on the description provided by HSBC in the November 10, 2010 note, RORO is the variance in the market return "explained by the first PC". It appears that HSBC uses PCA (principal component analysis) to solve the problem which is the opposite of the portfolio optimization: instead of minimizing the overall volatility of the portfolio given the return, they decompose the 34 markets into a set of long-short portfolios, and find the one which has the largest volatility. The recipe to construct it becomes different with any addition of new data. One can understand HSBC as saying that variance in the return of this portfolio is the RORO Index. The risk-on risk-off phenomenon is well conveyed by the colorful heat maps. From the description of RORO Index as "the variance in market returns explained by the first PC" it is not clear, whether this is simply the variance of the first PC or its variance in comparison with the rest of the PCs. The information about strength of the RORO phenomenon is in how the volatility of this most volatile portfolio compares to that of the next most volatile, and that -- to the next most volatile, and so on. That would characterize the strength of the multidimensional correlaiton.

The problem is how to convey this information compactly in a single quantity. CERCI solves this problem by summing the absolute values of cross-market correlations.

Bookmark with:

Deli.cio.us    Digg    reddit    Facebook    StumbleUpon    Newsvine