CM 003: Prediction Markets Under Attack; An Introduction to Subsidized PMs
The CFTC goes after Polymarket, while we introduce the Crowd to prediction market incentives and subsidies.
❤️ Please click the little heart at the top of the post—it’s like a tip, but free!
✍️ Share your thoughts with us about the issue in the comment section below.
📧 And a warm welcome our 8 latest Crowd members who joined since our last podcast. Share the newsletter with your friends to help us grow by sending them to crowdmoney.io to #JoinTheCrowd!
💰 Welcome, yet again, to Crowd Money, a monthly newsletter covering real-money prediction markets (PMs). Two weeks after our newsletter, we release the Crowd Money Cast, a podcast where we discuss our latest prediction market thoughts and interview key players in the space.
We had planned to dive deep into subsidized prediction markets, the different types of automated market makers, and our visions for a fee-less prediction market structure in this issue. Then came a wild week for the world of prediction markets, upending those plans and our written interview. We still plan to deliver that non-interview content to the Crowd—we think it is quite informative and useful—but that will be for another day (or most likely, week).
So in today’s Crowd Money 003, we will instead cover:
Prediction markets under attack as the CFTC allegedly begins an investigation into the leading DeFi crypto-PM Polymarket
An introduction to subsidized prediction markets, incentives, and the ways in which PMs can bring liquidity and interest to their platforms
The CFTC Investigates Polymarket
PredictIt. Augur. Kalshi. Polymarket. Scott Alexander used to tell us stories about the old days, a time of peace when low interest kept balance between the various prediction markets. But all that changed when…
On October 23rd, Bloomberg reported that the CFTC (the U.S. Commodity Futures Trading Commission) opened a probe into whether Polymarket has been “letting customers improperly trade swaps or binary options and if it should be registered with the agency.”
For context, trading swaps and binary options constitute the behavior of a regulated exchange. Polymarket notes on their website that “Polymarket is for informational and educational purposes only. We take no custody of anyone's money or cryptocurrency, extract no profits, nor do we host any markets ourselves. Polymarket displays existing markets live on the Ethereum blockchain (or sidechains) and is a graphical user interface for both visualizing data and market trends from on-chain activity, and interacting with said smart contracts directly via your Web 3 enabled wallet.”
TLDR: they claim that they are merely a display interface for transactions on the Ethereum blockchain, and do not actually operate the market mechanics behind the transactions.
In response to the alleged probe, Polymarket reportedly retained the former head of the CFTC’s enforcement division, James McDonald, a partner at Sullivan & Cromwell LLP. So far neither the CFTC nor Polymarket have released anything about the probe nor confirmed it.
Ever get that feeling of Déjà Vu?
Of course, there was never truly a time of peace in the prediction market space. Only a mere armistice. Back in 2012, the CFTC investigated and charged InTrade—an Irish-based prediction market with both real-money and play-money markets—with violating the agency’s “off-exchange options trading ban and filing false forms with the CFTC.”
The CFTC brought forth a civil case against the company in Washington D.C., seeking an injunction preventing American citizens from trading on the platform. Shortly after the lawsuit was filed, InTrade closed all US-based accounts causing a sharp decline in liquidity. Just six months after the suit was filed, InTrade suspended all trading and by the end of 2013 liquidated all remaining bets.
Will Polymarket be shut down by the CFTC?
Is the Polymarket case like InTrade’s? As OldBullTV pointed out on Twitter, InTrade was accused of misappropriating member funds in addition to off-exchange options trading. So far Polymarket is only being investigated for the latter.
Nevertheless, some users on Twitter have noted that an investigation into the crypto-platform will reveal other wrong-doings. Others have noted privately to us that Polymarket has been playing a dangerous regulatory game and would not be surprised if the CFTC attempted to shut down the platform.
On the other hand, the “and if it should be registered with the agency” part of the probe gives us hope that Polymarket will live on.
How did Polymarket get reported to the CFTC?
The news only broke a few days ago. As of writing, we do not know—but we want to know. While we have heard rumblings both publicly and from connected sources privately about the origins of the probe, they are uncorroborated. One likely scenario, we believe and feel confident sharing, is that the CFTC’s actions are part of larger US-government action against the crypto space which recently included the SEC seeking to rein in stablecoins.
Regardless of the cause, we want to figure out the truth. If you are connected in the space or know of individuals who may have useful information please get in contact us at email@example.com or more securely on Signal (+1 720-464-7408).
The clear loser is Polymarket—particularly in the short-run. As the Bloomberg article noted, Polymarket was in the stages of raising another round of venture capital funding at a near $1B valuation. So far Polymarket has only publicly raised $4M from their Seed round, although we have heard rumblings of another founding round that took place earlier this year. We are not sure whether the $1B raise is the same round, but seems less-likely than not to be different given the timeline.
In the medium- to long-term, the company faces other issues. While other competitors grow or enter the space, Polymarket has to operate under the pressure of regulatory scrutiny. The probe might also make traders hesitant to continue trading on the platform or cause new ones to find another platform instead, both of which would decrease the liquidity on the platform. As of now, Polymarket still has more trading volume on its platform relative to Kalshi, but will that continue to hold?
The other losers are the myriad of other crypto-based prediction markets in the space, such as Futuur, Hedgehog Market, Zeitgeist, and others. Some like Augur which do not currently operate in the United States will be less affected, but these platforms still want to open official operations in the largest user market. This makes that more complicated and difficult to navigate.
The only clear winner in the short-term is Kalshi, the first and currently only CFTC-regulated and approved real-money prediction market. How big will their benefit be? We’re not so sure, but it’s clear being approved by the agency has its benefits.
In the medium- to long-term, probes such as these could make it easier for other companies to get CFTC-approval otherwise the agency would be de-facto creating a monopoly in the space. Plus, Kalshi already blazed the path for regulatory approval. Moreover, in the long-term we think this development could be good for the PM space in general. Having regulated markets builds trust, and trust will make it easier for new people to enter.
What is clear, however, is...
An Introduction to Subsidized Prediction Markets
In the first issue of Crowd Money we explored what predictions markets are, how they operate, and why individuals should care about them. We discussed the role of market makers in the prediction market ecosystem and why liquidity, in any form, is key to prediction market success.
In the second issue we looked at a specific use case to contextualize these ideas: the policy-making space. We spoke with experts from the field on the feasibility of leveraging prediction market intel to make policy decisions and delved deeper into how liquidity and trading volume are critical to generating accurate information. We also discussed how market manipulation is costly to maintain in real money markets in comparison to their prediction alternatives.
In this, the third issue of Crowd Money, we will be exploring just how prediction markets can achieve liquidity, trading volume, and question interest through arguably one the largest categories of market stimulants that exist: subsidies. But first we need to talk about why prediction markets outside of individual reasons.
The Larger Logic Behind Prediction Markets
Prediction markets are themselves a type of incentive.
From Tetlock and Gardner’s Superforecasting, to the heap of academic papers on the subject, we know that people are capable of making accurate, quantifiable predictions. From public health officials, to financial professionals, generating accurate information on the outcome of events can be used to make important decisions.
Prediction polls such as Metaculus and Good Judgement Open are places where people make predictions about future event outcomes for gamified rewards and status symbols. Often the idea behind prediction polls or any place that aggregates predictions about the future is delivering the “wisdom of the crowds.” The idea is that if you get enough people to forecast a question, the aggregate of their wisdom will deliver true foresight.
That thought is largely…incorrect. Recent research from Satoppa et al. (2021) reveals that this basic aggregation largely benefits predictive accuracy through its reduction of bias: The systematic over- or under-estimation of future event likelihood. This is okay, especially for questions where the signals are largely public and easily accessible. But this improvement is likely insufficient for decision-makers and others to seriously consider the information.
As the paper continued to show, the best way to improve predictive accuracy—by reducing bias and noise (non-systematic, unpredictable error) and increasing information—is by aggregating the forecasts from the best forecasters in the group (aka “Superforecasters”). It’s not the “wisdom of the crowds” that we are after, but the “wisdom of the forecasting elite.”
This is further enforced by findings such as Human Forest, which despite using a small-N, was able to generate great predictive accuracy, both from their individual forecasters (incentivized through prizes) as well as through their different methods of aggregation.
But getting people, particularly those with elite skills or access to useful private information that is currently not known by the poll or market, to participate is hard. The data used in the Satoppa is from tournaments run by IARPA which included large cash prizes. As Robin Hanson noted nearly a decade ago, places like Metaculus or Good Judgement Open tend to fail in mooching information off of people.
Prediction markets are the incentive to get those people off the sidelines. By allowing people to generate alpha off of their forecasting ability or access to private information, we incentivize them to add their knowledge to the elite pool and generate more accurate foresight.
And there’s evidence for that. That same Satoppa paper showed that prediction markets were the best at increasing the amount of information of the market, which is likely why Atanasov et al. (2017) found prediction markets to be more accurate than their prediction poll counterparts 6-months to a question’s end date.
Why Do We Need Subsidies?
Great. So if the prediction market is the incentive, why the need for subsidies? Two reasons:
To directly incentive people with the right information or skills to participate in the market, who might otherwise not due to being risk-adverse or too busy
To indirectly incentive those same people by increasing liquidity and interest in a question, causing dumb money to enter and increase their potential gain from participating
By increasing liquidity, volume, question interest, and the information in the market, subsidies help transform prediction markets from the theoretical world of usefulness and the practical world of small-time gambling into serious markets that deliver useful information that also generate significant liquidity and volume.
So, what are market subsidies? Market subsidies are financial incentives given to individuals or groups often with the goal of encouraging consumption or purchasing. These can come in the form of tax breaks, cash outlays, price off-sets. Subsidies also lower the requisite capital and confidence needed to trade in prediction markets. This should result in greater market participation which will drive increased liquidity and trading volume in turn.
More Than Just Money
Before diving into our market subsidy thoughts, it is important to note that market subsidies are not substitutes for some of the fundamental aspects of a successful prediction market. In other words, you can’t just spray some Axe cologne and call it a day, you need your deodorant (Middle Schoolers need not apply).
No subsidy will fully compensate for a prediction market platform with a poor user interface, slow payout times, or questions that cover irrelevant or uninteresting topics. Long resolution periods, a high likelihood of an ambiguous resolution, and low marketing efforts can be groups with these as well. These basics, noted by others such as Zvi on LessWrong, are still required in any prediction market before considering the positive effects of market subsidies.
Types of Prediction Market Incentives
Market subsidies are one type of incentive for traders. There are others which we grouped into two types: non-monetary and monetary.
Leaderboards or other reputation-aimed programming (like Good Judgement’s title of “Superforecaster”)
Clearer questions with defined resolution criteria
Relevant questions about current events
Sweepstakes or tournament prizes (like Hedgehog Market’s no-loss competitions...although their version is arguably a prediction tournament rather than market)
“Free money” by a third party looking to generate interest in a market to gather data
“Free money” by the prediction markets themselves for new users
Capital from automated market makers
Lower or non-existent fees
Speaking of monetary incentives...if you want to start making plays, Polymarket is a great place to start! Use referral code ‘CrowdMoney’ here to get up to $100 reimbursed on your first day of trading! https://polymarket.co/crowdmoney
Which is Better?
Generally speaking, while incentives are powerful as we’ve seen on platforms like Metaculus and Good Judgment Open, we believe that in the long-term subsidies will be a more important and impactful strategy.
Within subsidies there are two routes a prediction market platform can take. They can either provide money, i.e. outlays, through the aforementioned methods and the like. They can also take the opposite approach, and instead of giving money away, they can cut costs for the user.
This is where we see the prediction market space headed! The largest cost to cut for a trader is fees, and this is what we will briefly examine in the rest of today’s issue .
Please No Fees
We will be focusing on cutting fees for a few reasons. First, it’s impact on the ecosystem cannot be understated. Many of the market arbitrage opportunities in prediction markets exist because of fluctuating fee structures between platforms. Going fee-less would shake up the entire space, and force other platforms to find ways to support new revenue streams.
Second, going fee-less requires fewer initial outlays of cash, although as we will see, this framework is still costly. It does require, however, a substitute for the fee revenue that currently supports most platforms.
Third, this is a change that happened recently in the e-broker industry and greatly increased trade volume as well as catered towards retail investor activity. When retail trading firms offered no-fee stock trading, daily average trades for retail investors increased upwards of 100%. If we want to get more traders on these platforms, they will need a fee-less system to compete with the fee-less accessibility to equities trading that currently exists.
Finally, going fee-less would, theoretically, drive both of the most important indicators of prediction market success: liquidity and volume. You will see more money trading on more markets if there is lower cost to trading. This will result in more accurate information being generated from markets. It will also allow for more “suckers at the table” which exist in any market, and provide incentive for talented traders to make plays and generate alpha from the uneducated players in the market as well.
The natural next question is, how can any prediction market platform support a fee-less pay structure? We at Crowd Money have devised two potential models to achieve this, one of which includes a tournament, margin trading, and a subscription service. If you want to hear about this framework and the other in detail, let us know and we will outline it either next month, in our news supplement, or a separate post in the weeks ahead.
Please No Fees?
Is there a case that there should be fees? Potentially. We are not talking about fees for depositing funds or withdrawing them, or anything of that nature. Those fees should be sent to the shadow realm, never to return again. But what about trading fees?
Well, that comes back to automated market makers. So keep your eyes peeled...or simply #JoinTheCrowd to hear that case in the weeks ahead 😉
We have a new newsletter partner, PredictIt! Featured by websites like Bloomberg, Yahoo Finance, Axios, and others, PredictIt is one the premier political betting platforms. Use referral code ‘CrowdMoney20’ to get up to $20 reimbursed on your first day of trading! http://predictit.org/promo/crowdmoney20
If you want to submit plays and you want to have more of a conversation about it, come find us in the Prediction Markets Discord server! We like to discuss plays there and get feedback on our content.
We are also in partnership with the Young Professionals in Foreign Policy organization, a global group focused on developing the next crop of foreign policy leaders. Check them out and sign up if you want to network with like-minded individuals around the world!
Just like last month, if you have ideas, want to contribute, know a potential guest, or just have feedback please reach out to us!
If you came to Crowd Money but have not heard of Global Guessing, go follow us on Twitter! We recently released two interviews related to forecasting. The first is with David Manheim where we chat about Covid-19 preparedness and predictions. The second is with David McCollough, an elite forecaster and Managing Director from the Good Judgement Project where we discuss the importance of question quality and more. Go check them out!
And finally, stay tuned for a new piece of content which will be released in the week between Crowd Money issues and the Crowd Money Cast. Instead of including news items in these monthly issues, we will be sending out a separate, shorter, post with interesting and important articles from the month. If there are stories you think we should cover from the news, send them in!
And a reminder, if there are topics you’re interested in, you have feedback, or just want to say hi, you can follow us at the below: