Saturday, November 12, 2016

Prediction Post-Mortem

As is normal, I've been getting some ribbing about my election forecast that an unexpected Trump victory would cause the Dow to drop 500 points before rebounding in a week.  This prediction was informed by the Brexit election in late June, where the pundits were wrong, and the FTSE tanked, only to be normal a week later.

What happened here is that the U.S. pundits were also wrong, the Dow futures dropped 800 points before the Dow finally closed up 250 points - a thousand point swing in one day.

So, my prediction was correct directionally - there was a violent drop.  I predicted 500 points but it was 800 points, which was due to the fact that the markets were still open in Asia and Europe, where Trump is far less popular.  My prediction was correct that the stock market would snap-back quickly, and even I was shocked just how quickly it actually snapped-back..  (I give some credit to Clinton's gracious concession speech, which was probably the best speech of her career.)

All-in-all, I feel good about my prediction,but I am very curious as to why polling has become so much less reliable than even a few years ago.  I recognize many young people don't have land-lines, making them difficult to contact and therefore to include in the survey.  Even when contacted, an increasing number of people refuse to participate.  Are people less cooperative?  Are people just less available, working in a 24/7 global economy?  Maybe, people suspect the caller is not really a pollster and don't trust them.  I like to think that people are finally discovering how little of their privacy they still have and choose, at the very least, to keep their political opinions to themselves . . . I know that I never talk to pollsters.