It’s that reflective time of year again. And as we get into that retrospective pondering, two 2016 events stand out like blazing beacons: the Brexit referendum and the US election.
In the space of a few weeks the world gasped as a pair of entirely unexpected results — the proverbial ‘black swan’ nobody saw coming — vaulted onto our front pages.
When you chat with Peronet Despeignes, a former Financial Times economics journalist now working with the decentralized, blockchain-based prediction markets startup Augur, you can’t help these two events bounding into your thoughts. Especially when he uses that very piece of avian terminology favoured by philosophers.
While Augur’s novel decentralized model is still in beta, we can only look at what other prediction markets, for example France-based Hypermind, managed on the Trump/Brexit front earlier this year. While they did not deliver Donald or ‘Leave’ as anywhere near favourites, they did do better than traditional polling methods. Despeignes is right behind Hypermind’s two blog posts on that subject, one each for Trump and Brexit.
The exciting question now is whether, with a global and decentralized model such as Augur, the prediction markets will be a lot further ahead of the other forecasts next time we’re on the eve of something so big. With some of the geographical restrictions of current markets eliminated by Augur’s concept, Despeignes believes they can.
“If you think about it, people who are too emotionally close to an event like Brexit or the US election might not be able to look at the details as objectively as a foreigner can. Our hope with Augur is that people in, say, China, can weigh in on these predictions. It’s a level global playing field that allows a wider variety of more people to participate on an equal basis, and that’s likely to make the markets more accurate.
“The wider the pool of people involved and the more variety there is in the type of people participating, the more accurate prediction-market results tend to be.”
Fast-forward to a time when prediction market odds — fueled by a much larger and diversified trading population — are accepted as exactly the kind of outlier you want ahead of a big day like an election. Imagine that on a hypothetical future Brexit vote, these market have ‘Leave’ at a 60% probability, and this gets more widely circulated in the media than it might have been in 2016.
Might the ‘clear and present’ nature of the prediction (by people who have put their money where their mouths are) make voters (who don’t have the same financial incentives to research the matter) sit up and reconsider? If financial markets were already reflecting sharp responses to a strongly anticipated outcome, then a tanking Sterling might well give those voters a last-minute nudge.
“That scenario could create some strange feedback loops,” chuckles Despeignes. “In democratic action people are simply voting, but there’s no immediate financial incentive like there is with prediction markets.”
His message is that the wisdom of a financially incentivized crowd is the best forecasting tool we have. And with blockchain technology making that tool ever more powerful through greater security, transparency, accessibility and user numbers, its power is set to grow.
“The basic rationale is that none of us has a monopoly on the truth, insight, awareness or access to facts about everything,” says Peronet. “We all have bits and pieces and fragments. Markets like this are one of the most effective ways of pooling those fragments of knowledge together.
“If you search Google Scholar for prediction markets, you find one of the closest thing to a consensus in economics. Which is that they’re pretty darn good on average.”
Consensus in economics? If nothing else, that rare phenomenon tells us it must be worth keeping an ever closer eye on prediction markets. And April’s Presidential elections in France might be a very good place to start…
Peronet Despeignes spoke at Fintech.Pioneers in Berlin on 16–17 in February 2017.
