Skip to main content

How Prediction Markets Could Guide Bitcoin's Future

Op-ed - How Prediction Markets Could Guide Bitcoin's Future

While opinions on Bitcoin's future differ, most agree that the current scalability debate has become a mess. Trolling, misinformation, populism, vote manipulation, vocal minorities, censorship and other distractions have made it hard to find a signal above the noise.

And importantly, if protocol development is going to be driven by what seems like popular opinion on message boards, this could potentially ruin the whole Bitcoin project.

The essence of this problem, Yale researcher and Truthcoin Chief Scientist Paul Sztorc argues, is that “talk” does not scale.

“The real function of a debate is for people to examine each others' reasoning, and find some area of agreement, or maybe highlight areas where they don't agree,” Sztorc told
Bitcoin Magazine.

“But debates do not scale,” he said. “It's literally O(n^2) scaling: With five people, you need at least 10 conversations in order to know that everyone is on the same page … with 50 people, you need at least 1,225 connections. If someone changes their mind or learns something new, that’s all reset. So it’s no surprise – to me – that the conversation is becoming socially dysfunctional at around this time.”

Luckily, Sztorc believes there is a solution for this problem: prediction markets.

Prediction Markets

The concept of prediction markets is not new, Bitcoin legend Hal Finney advocated them years ago.

Basically, prediction markets are markets for so-called “event derivatives,” which represent a possible future event. “Hillary Clinton will be elected president of the United States in 2016” could be such an event, for example. The relevant derivative might then be redeemable for one dollar if Hillary Clinton is indeed elected, but will be worthless if she is not elected. Up until the election, this derivative will be tradeable on the prediction market. As such, it will command a market price.

It then holds that this price would reflect the likelihood of Clinton becoming the next president, according to the market. If the Clinton token is worth 40 cents, the market gives her a 40 percent chance of winning the election. After all, if market participants expected a higher chance of her winning, they would buy these derivatives “at a bargain,” and the exchange rate would go up. And if market participants would expect a lower chance of her winning, they would sell (or short) these “expensive” derivatives, and the price would go down.

“The great thing about markets,” Sztorc explained, “is that they are incredibly decentralized – anyone can partake. This is good because, by definition, you don’t know who has what information, and people won’t want to give valuable info away for free. At the same time, markets present information that is unanimously and constantly acceptable. If you have a room full of 5,000 people, those who agree with the current rate would just do nothing. The price is ‘good enough,’ as far as they are concerned. Yet
person with a strong enough conviction can edit the price by buying or selling the event derivative, and update the exchange rate. They can improve the forecast and make money at the same time.”

The Great Noise Filter

Furthermore, prediction markets could be a great filter for uninformed opinion, or “noise.” Unlike message boards, chat rooms and mailing lists, partaking in a prediction market would come at a cost. Sztorc, therefore, expects that people who didn't invest the time and effort to really, deeply understand an issue won't get involved at all, as they would be unwilling to invest money, too.

Talk is cheap,” Sztorc argued. “But people who really don't know anything, or know how complicated a situation is, will probably not want to put their money down. And even if they do, even if a fool pushes the price around, there is now money to be made by well-informed people who can pull the market back toward realistic expectations.”

“The same is true for manipulation,” he said. “Everyone wants more money, while only a few people will want to knock a market in a certain direction. As such, markets inherently resist dishonesty. Unlike anything else, there is a force pulling the market rate towards reality. They just plain do a good job of getting the right answer, over time.”

Joint probability

Prediction markets can show us the expected future. But, much better, they can also be used to compare multiple expected futures.

Through funky constructions, it's possible to make predictions about specific scenarios. For instance: “If Hillary Clinton is the next president of the United States, then the stock market will rise.” Now, if Clinton is not elected president, anyone who bet on that scenario will get their money back, regardless of what the stock market does. It's only if she is elected, that the stock market comes into play. If Clinton is elected and the stock market then indeed rises, whomever predicted that correctly wins money. But if Clinton is elected and the stock market drops, they will lose money.

This also means that market participants can insure themselves against future events. Someone who holds stocks can insure himself against – say – a Donald Trump presidency, by betting that the stock market will drop if Trump wins. The bettor might still lose money on the stock market, but win money on the prediction market at the same time, canceling out his losses.

And, of course, these market rates will also be public for anyone to see. So now, prediction markets are not just predicting possible events, but consequences of possible events. While

presidential candidates will probably claim that the stock market will rise if they are elected, prediction markets reveal who the market believes is right. Rather than empty promises, voters can base their vote on market information.

Bitcoin Governance

So how is any of this relevant for Bitcoin's governance process?

Let’s take the block-size dispute as an example, and assume that everyone wants the exchange rate to go up. (Note that this is not necessarily true; some people value Bitcoin for other properties than its price. However, even then, the expected price could inform specific aspects of the debate, such as the security offered by the value of the mining reward.)​

First, a prediction market regarding bitcoin's price as resulting from the block-size limit could be utilized as a “hard fork insurance.” Lets say, for instance, that someone named Greg believes that bitcoin's exchange rate will drop if the block-size limit is increased. Greg doesn't want the bitcoin exchange rate to drop, because he holds a bunch of bitcoin. Therefore, Greg bets that if the block size is increased, the exchange rate will drop. If the block-size limit is then indeed increased, and the exchange rate indeed drops, Greg will win money. He would have effectively insured himself against that scenario.

That would be an advantage for Greg personally, as he wouldn't risk losing money. But it would also help the rest of us. After all, by making the trade, Greg has effectively stated his opinion, which is reflected in the exchange rate. Moreover, Greg could tell other people to insure themselves, or simply make his own insurance public. If people trust Greg's expertise enough, they will follow his lead, which adjusts the exchange rate even further.

But Greg is not the only one betting. Someone named Mike, who believes the exchange rate will rise if the block-size limit is increased, might be betting as well. And people who trust Mike's expertise might follow his lead instead. Now the prediction market reflects a potentially terrific aggregate of expertise, and their expectations of what's best for Bitcoin, offering a great guide for Bitcoin's development.

And there's an additional bonus. Lets imagine that Greg and Mike are talented programmers, working to improve Bitcoin or the ecosystem. Thanks to the prediction market, neither Greg nor Mike need to repeatedly enter into debates on message boards convincing everyone and their grandmother that increasing the block-size limit is either bad or good. After all, they have already insured themselves, and stated their opinion through their trade. Instead of wasting their time on Reddit, they can both get back to work.

No silver bullet

Of course, a prediction market won't solve Bitcoin's governance problem in and of itself. First, since Sztorc's Truthcoin is designed as a Bitcoin sidechain, there is one prediction that it can’t support. If Bitcoin breaks completely, or its exchange rate goes to zero, no payout will be possible, even if predicted correctly. Additionally, Sztorc's proposal relies on “
to insert the truth after an event has happened, which required a bit of trust in these oracles. (Sztorc believes we should be able to find enough reliable people to do this task, to tide us over until his larger “trustless orcale” project is completed.)

But, most importantly, prediction markets cannot govern Bitcoin in and of themselves. They can only advise – and maybe that’s for the best.

Prediction markets don't need to replace anything. You can still have your mailing list, you can still have your forum discussions, and prediction markets won't automatically change Bitcoin's code,” Sztorc emphasized. “But prediction markets do offer something called common knowledge. Everyone will know what the market expects to happen – and everyone will know that everyone knows. And how weird would it be if the prediction market says we'll see a huge exchange rate collapse if we either do or don’t fork to Bitcoin XT, and yet some developers still keep fighting or pushing it?”

For more information on Truthcoin, visit