Election 2016: looking back at October and November 2016
The only reasonably likely outcomes for November 2016 was RH_DS (Republican House, Democratic Senate) or RH_RS — Republicans were always going to hold the lower House, Senate control was the only open question.
Throughout 2016, to check my priors, I thought that RH_DS was going to happen, though when I visited DC in mid October 2016 some friends on the right were really confident about the Senate in a way that was quite bothersome (and at least I think I know when they are trolling me). In 2015, I thought it was a toss up while the market thought RH_RS was the most likely outcome.
Taking just August-November 2016, one can see the volatility in October 2016 both in the graph and the data. In early October, Republican senate was in the high 30s (37–38%) while Democrats holding the Senate was at in the low 30s (32–33%). The markets did not initially overreact to the locker room talk “October surprise” on October 6, 2016.
However, while there are many problems with thinking of prediction markets as an oracle (besides the question of whether there is sufficient volume), the markets are can be quite responsive. By the end of October, for example, RS were down to the high 20s (26%) while DS was still in the high 30s (39%).
The day before election day, the markets were smarter than the consensus which believed this would be a wave election with the first woman president having a unified Senate and House in her parties hands. RH_RS was at 52% while RH_DS had collapse to 24%. If one had looked at IEM rather than CNN, one would have known that Hillary was going to face the same conundrum that Obama had struggled with since 2010. Even if she had won, she would have faced a Congress that would try to obstruct (resist?) her while she was hounded about the emails and a possible FBI investigation. It’s not like a divided DC led by HRC would have been able to make the fundamental changes needed on climate change so maybe we shouldn’t blame Trump for end of civilization. Would Hillary have negotiated poorly (or even worse, sold out to Wall Street) as a centrist like her husband ended welfare as we know it?
Reality has a way of making “data driven” predictions seem foolish if one is too “result dependent” in their evaluation. Traders are reflecting what we know, and more importantly, as Nate Silver and other have pointed out, the binary nature of the outcome makes it seems like the “polls”, “experts”, and handicappers on a prediction market are fools. That being said, at least when it comes to Congressional control, those who were putting their money where their mouth was seemed to know what was happening better than those who were just talking (“Democrats enter the final days of the campaign season with a clear shot to take the Senate majority, with seven GOP seats at serious risk of flipping and only one of theirs that could slip away.But a new wildcard has entered the mix: Wisconsin — a seat long thought to be a Democratic lock that has now become a lot more uncertain.” http://www.cnn.com/2016/11/03/politics/senate-races-flip-2016/index.html)