Comment Text:
I will start with a general comment and then provide specific responses to the CFTC questions.
General Comment
Our democracy needs prediction markets to produce accurate and unbiased predictions about elections, politics, climate change, health, and more.
At age 6 I won my first prediction market by estimating the number of candies in a jar. This was the first time I had “bet” money. But I don’t think it was a game of pure chance, rather this was the start of a lifelong interest in estimation, statistics and probability.
We call ourselves political gamblers, but the people who are consistently making money on prediction markets are more accurately described as prediction market traders. I don’t gamble. I’ve never gambled at a casino, bought a lottery ticket, or even participated in a charity raffle. Instead, I buy (or sell) contracts in events that I think are more (or less) likely to occur than the market price.
In 1993, I traded my first prediction market: the University of British Columbia Election Stock Market for the Canadian Federal Election. That year I lost money ($15!) as I was a novice and didn’t expect the New Democratic Party to have their worst ever result. My next encounter with prediction markets was the US Presidential election of 1996 which I traded on the University of Iowa Election Stock Market and worked on an (unfinished) arbitrage bot.
Over the next 20 years, I made several good calls including buying a house within 2% of the bottom of the Philadelphia housing market, and in 2011 being one of the few people in the world to make money shorting bitcoin. I made several good calls on the bitcoin price bottoms of 2015, 2018, and 2020 (85% retracement from the top!).
In 2016, I joined PredictIt and started to take my prediction skills to the next level. After some initial success from observing trends, I built a multivariate regression model for the Democratic presidential primary. As that featured a relatively stable two person race (Clinton vs Sanders), after some initial mistakes (oops - I need a turnout model!), I did a new iteration each week and by the end I had built a model that was better than 538. I turned $500 into $5000. On Nov 8, 2016, even while PredictIt was barely functional, very early in the night with FL results coming in I saw that Trump was going to win the election and made several thousand dollars.
Every year since 2016 I have consistently made money on prediction markets. Since I was a kid, I’ve been fascinated by history, international elections and politics. At age 12, I was reading The Economist, world almanacs, and soon became an avid shortwave radio listener. I love international markets!
I’ve predicted who would win the past two Conservative Party leadership races in Canada (both times the second place candidate won), that Servant of the People would win a majority in the last Ukrainian parliamentary election, that the socialist Economic Freedom Fighters would back the right-wing Democratic Alliance over their former ANC comrades to form municipal government in Johannesburg, that the polling would narrow in Peru presidential runoff election of Fujimori vs Castillo, and more recently that Petro would be elected the first leftist president of Colombia. I love doing a deep dive into the politics, economics, and history of a country!
Last year, I had a 20% edge over three months of betting on US primaries on PredictIt. I’ve got a model that can predict who will win even if there aren’t any polls! For instance, in the special election in NY 19 I correctly chose the Democrat even though he was behind by 6%.
The United States has legalized gambling on sports, casinos, lotteries and more. None of this provides the public with any useful information that furthers our democracy. By contrast, the role of political prediction markets is to provide accurate forecasts about elections and other events. This is a PUBLIC GOOD!
Furthermore, prediction markets encourage people to educate themselves about the political process and to increase their civic engagement. I’ve spent thousands of hours researching things that I’m betting on and discussing my findings with others. I’ve also been active in social movements for thirty years including starting the Progressive Student Alliance (Notre Dame) and being members of groups including: the Philly Socialists, Democratic Socialists of America, Student Environmental Action Coalition, United Students Against Sweatshops, and I even campaigned for Nader for president in 1996 and 2000.
There is a mix of people using prediction markets ranging from people who are closer to gambling, those who think they know what they are doing, to those who are consistently making a profit (aka the sharps). The so-called gamblers who are losing money are essential to the system as they pay for the work done by the sharps. The so-called gamblers are funding the provision of public goods - accurate predictions.
Currently our democracy is under attack by people who want to gerrymander districts, discourage voting, threaten election officials, and overturn election results. These people would rather have us listen to partisan political hacks and entertainers than to polls, data models, and prediction markets. Taken to the extreme, a minority of these people participated in the Jan 6 insurrection.
I prefer science to violent insurrection.
I think we need to use the imperfect scientific tools that we have to defend our democracy. We should encourage people to build data models with financial incentives. We should encourage people to trade on political prediction markets as a system of checks and balances for inaccurate polling, as well as to react in real-time to changing fundamentals.
We need expert predictions from people who are paid to eliminate their bias. With journalists becoming more partisan, we need an alternative source of analysis. As a socialist, I’ve bet and won money on Trump and other conservative candidates. I’m able to reduce my own bias enough that I can consistently profit.
There are already billions of dollars in politics. There is no evidence that political gambling has corrupted the political process. There are countless ways that an individual or corporation can make millions or even billions of dollars from government policy whether through changes to the tax code, government contracts, or other methods. By contrast, the amount of money involved in predicting elections is minor.
One cannot logically argue that legalizing regulated prediction election markets would undermine our democracy, unless you have evidence. So far I haven’t seen anyone that argues this position providing evidence. Instead they are spreading fear, uncertainty, and doubt about the impact of election prediction markets. I think this is disinformation that comes from the fear of the unknown and a morality based opposition to gambling in general.
By contrast my evidence is my life story (and that of many fellow prediction market traders who I know) including the research, analysis, and data models that I have created to trade these markets. I’ve made money because I was better than average and helped move the needle of the markets towards being more accurate. Also you can look at the United Kingdom where election betting has been legal for many years. I see no evidence that their elections have been corrupted by betting.
How about using prediction markets without money? Metaculus is an example of a prediction market that relies upon people competing for reputation. However there is no money involved. So while it provides useful insights, the lack of financial incentive causes a lag in predictions. This was evident in the December 2021 market for whether the Omnicron variant would become dominant. We need to have money in prediction markets to eliminate this lag.
I’m good at this. But there are people who are far better than I. If the CFTC cracks down on these markets they will move overseas, often using cryptocurrency, with an increased risk of scams, hacks, and money laundering. This will cause HARM to the general public (and require a lot of work and resources for the CFTC and/or SEC to catch all the bad guys). But if the CFTC and Congress move to permit and regulate prediction markets, then we’ve only seen a fraction of what is possible as we open a vast new area for data driven economic and political inquiry that will strengthen our democracy!
Responses to the CFTC Questions
1. No. These contracts do not relate to “gaming”. Election prediction markets involve research, analysis, data models, polls, and are very different from gambling.
2. The use of a 5000 order multiple is entirely unnecessary as election prediction markets do not involve gaming. It might be useful in establishing election prediction markets by assuaging some of the fears of market opponents, but, in my opinion, will needlessly restrict the development of accurate prediction markets that promote election integrity.
6. I’m not sure if these are different from Nadex contracts, but I would argue that that ruling was wrong.
7. I’m not sure if this is significantly different from Kalshi’s prior proposal, but I would argue that proposal should have been accepted.
8. The contracts can serve a hedging function. My understanding of hedging is that it may only require that some users, often a minority, use it for hedging. That threshold is easily met.
10. Yes. For instance, if I was a full-time prediction market trader or the operator of a prediction market I would expect that the change in Congressional control would impact the regulations surrounding prediction markets as (very generally speaking) Republicans are more likely to favor them than Democrats.
11. The Commission should not consider the position size and should allow smaller multiples.
12. Binary contracts are useful.
14. The contracts are promoting a public good - the accurate prediction of election results. This is essential to defending our Democracy which is under attack.
15. Trading in these contracts is likely to increase the integrity of elections by promoting polls, data models, unbiased research and analysis, and increase political engagement by regular folks. In the medium to long term, they will improve the public’s perception of the integrity of elections (it might take a couple years for people to get over fears that lack any basis in fact).
16. Is there any evidence that this has happened in the UK?
21. The order size limit may slightly increase the risk of manipulation. In general, you want to increase the number of actors in a market to reduce the probability that anyone is able to manipulate it
22. No. That is unnecessary as this market is very unlikely to be manipulated in my opinion as an experienced prediction market trader.
23. Very unlikely.
24. I think these markets not only meet the weak criteria of not being contrary to the public interest, but they meet the much stronger criteria of promoting the public interest.
