Observations on Correlation

by on March 8, 2013  •  In Miscellaneous

I recently found an old letter sitting at the bottom of an unread pile of papers. A friend had passed it along with a note commenting “pretty cool…quant/technical work that makes intuitive sense.” Feeling guilty for having forgotten about this for so long, I read through the old letter, and thought the following observation on correlation worthwhile sharing: 

“Interestingly we have found that high correlations happen in the most extreme way when macro events dominate (recession, depression or other)… In fact, extreme levels of correlation are reached towards the end of a bear market, most of all. Also…this phenomenon does not occur at the end of bull markets. Things are quite different at that point, as bull markets have strong tendency to become more and more narrow, thus resulting in low – and not high! – correlation between stocks.” 

Something to file away into the mental model archives…

Tags: ,