Augur had a big Election night, but the ‘morning after’ may determine the future of the upstart prediction market

The midterm election market will — hopefully — be the first Augur market over $1 million to successfully settle and serves as a large-scale proof of concept that the decentralized prediction market can be useful. PredictIt, essentially the centralized version of Augur, had just slightly more open interest on an identically-worded market. The total volume on the Augur market has surpassed $2M (because shares are purchased in ETH, the total volume fluctuates by day depending on the price of ETH) and the total volume on PredictIt’s market was a slightly higher $2.4M+.

Mechanically, there were some slight rule differences and execution results. On PredictIt, bets were capped at $850 per person, there was a 10% fee on the market, and the servers went down for a short period while votes were being counted and traders were live-trading the results. Augur, by contrast, ran smoothly throughout the entirety of the election.

PredictIt, the law, and what comes next


Results: Auguring the 2018 U.S. Midterm Elections

Read more

For American investors, PredictIt is widely regarded as the only public outlet to bet on political outcomes. The company, however, needs to limit the amount invested on a single contract in order to comply with their regulations from the Commodity Futures Trading Commission (CFTC).

Typically, prediction markets are not allowed in the U.S. because the CFTC considers them commodity trading outlets. But PredictIt has CFTC approval to operate for academic purposes only. Augur, by contrast, is unique as a decentralized application, effectively making it censorship resistant. The result of this regulatory bypass is that bet sizes have no limits for example.

Augur also had much more efficient arbitrage on its market. PredictIt’s high 10% fee made it impossible to play this arbitrage any time it was less than 10%. In a fair market, this limitation wouldn’t exist. 

Challenges remain for the upstart

The successful distribution of funds would prove that Augur can handle large-scale, mass-global betting in real time better than peer-to-peer centralized alternatives that currently exist. But will that occur? The next 6 months — yes, that long —  will be crucial in deciding this. The midterm election market actually expires on December 10th, which seems to be an arbitrary date that was chosen by the creator of the market, and the outcome of the event needs to be settled by what Augur calls reporters.

After December 10th, the creator of the market has three days to report an unofficial outcome on the market, i.e. state their opinion on which side of the bet was the winner. Then, reporters — who hold Augur’s REP token — have one week to contest this outcome. Any REP token holder can stake their tokens to dispute the unofficial outcome of the market, and if they are successful, they can earn up to a 50% return for reporting. Each dispute where a REP holder stakes their coins against the unofficial outcome is considered a “round” and each round of contesting an outcome lasts for one week. If the unofficial outcome is not contested, then the market settles, and the smart contract pays out the Augur shareholders.

The mystery trade better on a Republican win in the House

This brings about something interesting that happened on election night — a single unknown party purchased approximately $40K worth of losing shares right after the election results came in. The bet the trader was willing to make? He would offer up to 0.05 ETH to bet on a Republican House win that would pay out at 1 ETH — a possible return of 20X or more depending on the precise prices paid to acquire the apparent “losing shares.”

If somehow the bet pays off, the trader will make upwards of hundreds of thousands of dollars. But most would agree this bet pays off at zero as Democratic House control is now guaranteed. So what was he thinking? Perhaps that the wording of the market itself would ultimately be contested? The size and timing of the orders — moments after the election was “called” by real-world sources — implies a calculated plan by an experienced trader. But irrespective of his intentions, his trading action has spurred unrest among traders.

Language is critical 

Block by Block

Block by Block: Prediction Markets

Learn more

The midterm market was worded as “Which party will control the House after 2018 U.S. Midterm Election?” and the expiration of the market was one month after the election, specifically December 10th. There is a technicality with this because — even though the Democrats won the House election last week — the Republicans still controlled the House the day after the U.S. midterm Election (Wednesday).

Technically then, the outcome of the market was known at the very time the market was created. This would normally be a recipe for what’s called an “invalid market.” If reporters ultimately deem a market to be invalid, then shareholders are all paid out proportionally. For example, since there are three outcomes in this market — a Democratic, Republican, and Tie share — each would each pay out ⅓ of an ETH if the market is deemed invalid. If the reporters somehow conclude that this market should settle as Republican due to this wording, then this new entrant would be in for a huge pay day. But that shouldn’t happen, given Augur’s dispute process.

Still, there is an important decision for reporters to make here. Up until the day of the event, traders were presumably betting on the market as if it was about who would win the election. Odds on that Augur market mimicked other betting services, suggestion the presumption was indeed valid. Interestingly enough, PredictIt, the centralized version of Augur, had their market titled exactly the same as the Augur market and it settled without any dispute as if Democrats were the winners. (Editor’s note: There is no rational argument otherwise.)

A storm is brewing

It seems as though the most likely outcome is that the market will be disputed using Augur’s decentralized reporting system. This will be the first time that it’s been done with so much money at stake. There was another market, referred to as the “Bastille market” that went 15 reporting rounds and has sort of become “case law” around whether subjective markets should be allowed on Augur. The Bastille market asked if the “weather would be nice” on a certain day in a geographic region. The dispute ultimately ended with a ruling there cannot be subjective markets and it’s the longest dispute we’ve seen thus far. Bastille, however, was a low volume market.

In this case, there are 3 potential outcomes worth looking at for this midterm market, and in order of likelihood, they are:

  1. Democrats: Reporters recognize that traders were better on which party would win majority control of the U.S. House of Representatives.
  2. Invalid: Reporters decide that the outcome was known at the time of market creation and therefore all bets are effectively off.
  3. Republican: Reporters read the English language with a highly technical lens, and resolve the market counter to those who appeared to purchase winning shares. (Reminder that even as of this writing, the House is under Republican control and remains so until next year when the new Congress is convened.)

It seems likely the Augur platform essentially becomes useless if this market settles as anything other than Democrat — even though some purists believe the best course of action is to settle the market as invalid to encourage market creators to create less ambiguous markets in the future. With this kind of money on the line, a ruling other than Democrats signals to traders that they can’t trust Augur markets to settle in the expected manner — a manner consistent with centralized alternatives. There is already a significant amount of friction to use Augur as it is — the UI is complicated, there’s a lot of trading jargon, all bets require traders to first obtain ether, and a trader needs to have a basic knowledge of how to use the Ethereum blockchain as well as Metamask or another wallet.

What happens next augurs the future of Augur

A negative outcome for traders would deter usage of the app. In that event, the utility of the REP token would suffer and the token price would likely plummet. Another interesting technical point is that this market can be dragged out for around 10 weeks for a low-four-figure sum if one party wishes to claim “Republicans” is the official outcome.

In each week long round of contesting an unofficial result, the required bond size grows and it takes just a small — but growing — fraction of the total market to place the result into dispute. Eventually, the growth of the required bond would make it challenging to play this game but it could persist for some time. One counterweight against this tactic would be if the market creator decides to stake more REP than the size of the reporting bond on his currently unofficial outcome. In this case, the process would be much quicker and more expensive for a nefarious actor.

Still it’s hard to imagine a world where Republican becomes the official outcome. If you’re going to play on words — and take the stance that markets need to be settled in a highly technical manner — how do we know that the market creator was talking about the U.S. House of Representatives, and not the House of Lords, or some other house? Once one starts suggesting “it could be this instead of that” it becomes more than equally reasonable to note the market creator did indeed intend this to be about the U.S. House election and the winning party thereof. Other interpretations become increasingly absurd.

The way reporting works is very intricate in Augur and beyond the scope of this analysis. Suffice it to say that disputes are mathematically capped at 20 rounds.  If an official outcome is decided in disputes, then the staked REP on the invalid side of the dispute gets disperses to the reporters who reported from the correct side, so there is a chance to increase your REP holdings by reporting correctly. People who stake their REP reporting on an outcome different from the final outcome lose their REP, and the people who stake it on the right/final outcome get a 50% ROI on their stake. The length of this pending dispute could make for more winners and losers, but is just one challenge ahead for Augur.

Not just one market

Other than this market, there are going to be a ton of disputes in January. More than 300 markets expire at the end of the year, and many of them are a lot more ambiguous than this one, so there could be a rush of demand for the REP token to play the disputes even after these markets expire. Augur will lose 95% of its open interest due to these markets expiring when we enter 2019, and it’s highly likely traders will be very cautious making trades on Augur going forward. The friction, uncertainty, and time to settlement make for a negative experience for a trader.

Here are links to the shareholder addresses backing the three bet outcomes:

Democratic control:

Republican control:


Andrew Macdonald is a financial-market analyst at Canadian-based fintech hedge fund. You can find him on Twitter @macdonaldcrypto.

The post Augur had a big Election night, but the ‘morning after’ may determine the future of the upstart prediction market appeared first on The Block.

Augur had a big Election night, but the ‘morning after’ may determine the future of the upstart prediction market written by Contributor Network @ November 15, 2018 Contributor Network

Comments are closed.