A chart we've been watching a lot lately is the Citigroup Economic Surprise Index, a proprietary gauge by Citi designed to track how well economic data is doing in comparison with economic expectations.

Anyway, if you look at it over the last few years, it seems to have good predictive value for the S&P, which makes sense, since "beating expectations" is what the market really wants to see.

If you look at the recent peaks and troughs, they match decently with the troughs of 2011 and 2010, and the peaks in early 2011 and early 2012.

So how good is the index, really, at predicting the market?

Well, the other day we whipped up a couple of charts

The first compares the index vs. the performance of the S&P 500 over the next 6 months. The chart below compares the performance of the index vs. the S&P 500 over the next month.

Business Insider

Business Insider

Anyway, there's really no pattern at all -- no relationship between the surprise index and the future performance of the stock market.

When we made the chart, that seemed odd since it appeared that the market and the surprise index were moving together lately.

Well, a note from Nomura's Brent Donelly (of the bank's FX desk) helps clarify why the confusion.

Note that turns in the Economic Surprise Index have been informative for the direction of SPX recently (since 2008) but historically they are not very useful for trading equities. Here is the chart back to 2002; note how the surprise index went up and down, up and down from 2002 to 2007 and stocks went straight up the whole time:

So basically, there is a pattern these days. In the old days there wasn't one.

We're not really sure why that is, and so we can't be sure that the new pattern will hold, in fact if the pre 2008 period was "normal" it stands to reason that CESIUSD will become less important.