Arrogant and Delusional is No Way to Go Through Life
I was struck by the arrogance and self-assuredness of this proclamation by Romney, which he made at the infamous Boca Raton fundraiser on 17 May:
If it looks like Im going to win, the markets will be happy. If it looks like the presidents going to win, the markets should not be terribly happy.
I know Krugman has just today referenced this quote (here: http://krugman.blogs.nytimes.com/), showing an apparent negative correlation between Romneys Intrade favorability and the S&P 500 index. Thats wonderful, but I wondered if the Intrade players were reacting to the stock market, rather than to Romneys perceived chances of winning; Im guessing that a booming market is likely to favor any incumbent, whoever it may be.
Being a bit of a stats-dork, and also terribly inept at stock market predictions, I decided to take a closer look. I downloaded the S&P data from 18 May 2012 (the day after Romneys statement) through yesterday (21 September 2012). I then downloaded the Gallup weekly running data for the Obama-Romney matchup. I then ran a time series (cross correlation) analysis.
In essence, and without going into some of the finer particulars, I find a 2-day-lag positive correlation between the poll numbers and the S&P index (r = 0.235, p = .008, n = 126). In English, Obama poll favorability correlates positively with the S&P index and, importantly, it is the index that tends to follow the polls. Obama goes up; the stock market follows. Obama goes down; the stock market follows.
However statistically significant, the strength of the association is not strong. As a rule of thumb, I consider a correlation coefficient of 0.250 to be of moderate strength. Of course, the relationship could also just be coincidence.
That being said, it cannot reasonably be denied that
not only was Romney wrong in his arrogant prediction, but the opposite of what he said would happen
seems to be happening. Way to go, Mitt. Youre even worse at stock market prediction than I am, and Im not good.