financial markets newspaper
November 1, 2016
John Authers

We had our October Surprise, and it should not have been surprising. We should also be prepared for a November Surprise, and treat rising volatility over the week before the US election as a racing certainty.

What is beyond doubt is that the markets are scared of a victory for Donald Trump. Volatility remains low, but the Vix index has spiked twice since June’s Brexit referendum — in September 12, after Hillary Clinton was taken ill and admitted to having pneumonia, and again this week after a tracking poll showed Mr Trump in the lead.

To deal with this, we must break the issue into component parts. First, who is going to win, with what probability? Second is that probability priced by markets? And third, what effects would each outcome have on markets?

(Note that the outcome is not binary. A Clinton victory may be accompanied by control of the House, while a Trump victory could be accompanied by Republican loss of the Senate. There are also possibilities that Mr Trump does not accept a defeat, or of a constitutional crisis following a tie in the electoral college — extreme and unlikely events that would spell disaster for global markets.)

So, who is going to win? Prediction markets capture prevailing opinion and put a number on it. They may not be right, but they express the prevailing wisdom. Growing publicity for their remarkable success in predicting elections (over time they have been more reliable than Gallup polls), means they have assumed a role in forming that prevailing wisdom as well.

There are different markets, but all show remarkable assuredness that Mrs Clinton would prevail. Predictwise, which aggregates prediction markets, has never shown the Democrats’ chances as lower than 72 per cent and now puts them at 84 per cent. But this conceals a wide variation. Predictit shows Mr Trump’s chances doubling in the last week to 37 per cent. The Iowa Electronic Market, the longest continuously operative prediction market, which is less liquid than some others, now shows Mrs Clinton’s chance at only 57 per cent.

The latest polls show Mrs Clinton’s advantage in the RealClearPolitics average dropping to only 1.7 percentage points, from 7.1 percentage points two weeks ago, and behind in the critical states of Florida, Ohio, Nevada and North Carolina. That makes the prediction market odds on a Trump victory sound very generous

Therefore, it looks as though the probability of a Trump victory remains underpriced on markets, and could easily swing wildly in the next few days. University of Iowa research shows that we should expect this in any case. It is a property of betting on a binary outcome; as the event approaches, there is less time for a bet to pay off. That means volatility will rise. If anyone betting on Clinton has doubts now, they only have a few days to sell, and doing so will depress the price further. Big swings in prediction markets in the last few days of a campaign are the norm.

What will the impact be? Markets have not moved much until now in part because it is unclear what a Trump victory would mean. His policies are unclear. This intensifies the chance that risk premia will rise. But US Treasuries are havens. People buy them even on news that is ostensibly bad for the credit of the US. Treasuries’ prices rose in 2011 after Standard & Poor’s downgraded their credit rating. So victory for a man who says he might deliberately default on Treasury debt might lead people to buy more US government paper.

A further issue is that Mr Trump might try a big fiscal stimulus, with infrastructure spending and tax cuts. This would mimic Ronald Reagan, and imply a strong dollar and higher bond yields — which might be good for the US economy, but terrible for international asset prices.

So the next week is perilous. A little gold, or some Vix futures, might be a good idea, treated as an insurance premium. Volatility could easily rise far higher from here. And within reason, it is good to hold cash. It gives optionality. Cash can be rapidly redeployed, and that is appealing.

Finally, Trump futures offer value. Anyone who bought one for 9 cents on the dollar on the Iowa market two weeks ago has made a 365 per cent profit, and should consider selling. But some exchanges still put his chances below 20 per cent. Too bad that these markets have limited capacity; such odds are very tempting.