The Decline In Day-to-Day Correlation

Yesterday, I posted a round-up of some research I had done on Calendar Effects on the stock market.

I noted that I had been impressed by the correlation in the market’s day-to-day returns. In simple terms, what the market does on one day, it has been very likely to do on the next day.

This relationship, however, is more complicated than I thought. In the years just before World War II, the day-to-day correlation was absent. In fact, it was slightly negative. After the war, the correlation grew and it reached a high point between the Kennedy Assassination and the market plunge of 1973.

Here’s a look at the day-to-day correlation for changes in the S&P 500 for rolling periods of 1,000 trading days which is roughly four years (though it’s a little shorter before 1950 due to Saturday trading).

From mid-1968 to mid-1972, the correlation was an amazing 0.35. After that, things started to change and the correlation gradually sank. By the time of the 1987 crash, the day-to-day correlation was effectively zero.

Correlation hasn’t merely declined but it’s actually started to appear again, just headed in the other direction. From 2003 through 2010, the correlation has been -0.12.

Posted by on April 13th, 2011 at 3:24 pm


The information in this blog post represents my own opinions and does not contain a recommendation for any particular security or investment. I or my affiliates may hold positions or other interests in securities mentioned in the Blog, please see my Disclaimer page for my full disclaimer.