Go big or go home: how a Kalshi supertrader built a seven-figure edge

Earlier this year, as New York’s Democratic primary took shape, most people believed Andrew Cuomo was a lock. But full-time prediction market trader Iabvek wasn’t buying the conventional wisdom. Instead, he went long Zohran Mamdani, believing the polls were about to miss a youth-driven surge. His theory proved correct, and he was handsomely rewarded.

I recently spoke to Iabvek about the upcoming races in New York and New Jersey, how collaboration creates edge, and the risk philosophy behind his seven-figure run.

(Click here for our current NYC mayoral markets.)

Trade at your own risk. Not financial advice.

Kalshi: What got you into prediction markets, originally?
Iabvek: The first time I heard about prediction markets was reading about Intrade in 2012. I told my parents they should put all their money on Ron Paul to lose. They thought that sounded insane. I was also 12, so they probably shouldn’t have listened to me.

Then in 2015, I saw people talking about the Bernie versus Clinton primary markets. They were saying, “I made all this money, it’s so obvious.” I’d always been interested in elections, so I badgered my mom to sign up for an account because I was too young. She deposited a tiny amount of money, and we checked it out.

Kalshi: How did your first trade go?
Iabvek: My first trade was nonsense on the Wyoming primary. I lost about $20. Then with $100 going into 2016, I made two trades that both won: Democrats to win the Nevada Senate and the West Virginia governor’s race. Even though West Virginia is a Republican state nationally, it’s Democratic locally. That made sense to me. But I still didn’t think this would become something I did seriously.

Kalshi: When did you start taking it more seriously?
Iabvek: After a summer internship in 2018, I had my own money for the first time and traded heavier during the midterms. I followed things casually, but 2020 was when it clicked. I started taking it seriously, made money, and went from there.

Kalshi: What were your biggest trades from that period?
Iabvek: In 2020, one of my favorite markets was the Iowa Democratic caucuses. I saw that the satellite caucuses were full of ultra-left groups. “Latino Worker Local 212,” “Somali Leftists of Des Moines,” etc. Other traders treated the satellite caucuses as a wash, but it was obvious to me they’d overwhelmingly favor Bernie. I went big on that, and they swung it. The market mispriced a weird new thing.

Kalshi: What about more recently? What did you see that others didn’t?
Iabvek: A lot comes from experience. There are weird election-specific factors people don’t account for. Working with a team also helped. We spent a lot of time nailing down details across the country.

There were bizarre situations like the Adam Gray versus Duarte primary [CA-13]. There were ballots in UC Merced that wouldn’t be counted until the end and were 90% Democratic. Duarte led the whole time, so the market assumed he’d win. I was convinced those ballots existed, even though they hadn’t shown up. When they finally counted them, they put Gray over the top.

Kalshi: Your total profit on Kalshi is near $1 million, correct?
Iabvek: Yeah.

Kalshi: What has been your biggest win, and which are you most proud of?
Iabvek: My biggest was likely the [2024] popular vote margin market. We had errors in our model. It was insane to put all my money on it, but the market had even more errors. It worked out.

The market I’m most proud of was probably the House seats market [2024]. Maybe also my biggest.

Kalshi: As someone who is up big, what would you say to someone just starting out?
Iabvek: Estimate what the “sharp conventional wisdom” should be and use that as a baseline. Then find something it isn’t pricing in, and move from there.

Often, trading on conventional wisdom is enough because markets get skewed by people who think polls are fake. But I don’t think it’s enough to win big, and you can certainly get your face ripped off, because sometimes the crazy theories turn out to be right.

So I’d say it’s good to start with the conventional wisdom, and then try to find some factor on top of that that will really give you an edge.

Kalshi: You mentioned sometimes you get your face ripped off. Any examples?
Iabvek: Romania [2025]. After round one, the right-wing candidate had a 20-point lead, so I flipped right.

Then the moderate unexpectedly won in round two. We researched examples from around the world, trying to find cases where a candidate who led by ~20 points in round one ended up losing in round two. We really struggled to find examples like that, which gave us confidence that the right-winger would win. Many people cited Le Pen in France 2017 as a counterexample, but in our view, even though Macron consolidated support, consolidation on that order still wouldn’t have been enough for Dan to win.

At any rate, I certainly lost a huge amount.

Kalshi: You mentioned working with a team. How did that start?
Iabvek: Mainly through Discord. We sometimes pool money when it makes sense. We share models. Collaboration is enormously better than doing it alone. Other people catch your blind spots.

I work with a lot of smart people. But even if someone wasn’t that smart, they could still be really useful because they can just flag ideas. And maybe ten of their ideas are ridiculous, but then the 11th idea is very good. And you can stick to your guns with the bad ideas, and change your mind for the good ones.

That’s why I also read the comment section on Kalshi. Most of the comments are complete nonsense. But it’s the same idea. If you find that one good comment, it can change your mind.

A comment once helped us flag a big error in our margin model about New York vote totals. It turned out there were a lot more votes left to count. I thought the commenter who claimed this was crazy, but when I looked it up, he was right. It turns out New York City is just bad at counting ballots.

Kalshi: I’ve seen you post on X about risk management and Kelly criterion. What’s your advice on that?
Iabvek: I personally feel strongly that if you’re not doing this full time, you should go really, really big (within money you can afford to lose) and be completely willing to go bust. Part of that is just Kelly criterion stuff. If you have a future source of income [a job], your actual bankroll is much bigger.

Prediction markets are really popping right now. If you’re thinking about going full time, that transition is a lot easier to accomplish if you hit a big win. It can be a huge cushion. If you’re going big while you still have a job, I think it’s easier to make that transition.

It’s totally subjective, but to me, I think going small can delude you into thinking you have an edge. If you lose big, you learn, and you can’t make any excuses for yourself, and you probably should not do this.

You have to be responsible. You don’t want to literally lose your savings. But within the parameters of money that it wouldn’t materially affect your life to lose, I think it’s good to be on the higher end.

Kalshi: So within reason, go big or go home*.
Iabvek: But obviously if you’re doing it full-time, you should be more conservative, because you don’t want to blow up.

You should also be cognizant that the marginal value of the additional dollar tails off more sharply than the logarithmic function that Kelly says, because there’s less really good trades you can make with a million dollars than you can with $1,000, because there's less liquidity. But Kelly assumes your opportunities will stay the same as your bankroll increases, and that’s just not accurate.

Kalshi: So at a certain point, the markets are going to move if you’re putting that kind of money in, so Kelly doesn’t always scale.
Iabvek: Right. It’s a good baseline, but not gospel.

^{*Please note that while Iabvek believes in you, the interviewer, Old Terry, thinks you probably suck at this, and should not “go big.”}

Kalshi: Let’s pivot to this year’s elections. How did you do in the primaries, and what do you see next, to the extent that you can talk about it without jeopardizing your trades?
Iabvek: First off, I'm actually not very concerned about jeopardizing my trades. And maybe that's stupid. Maybe people are going to yell at me for leaking stuff, but my philosophy, basically, is most people who listen to this just won't hear these random details. And if they weren't already the kind of person who would be able to figure this stuff out, then they just don't really know how to piece it together.

(Click here for our current NYC mayoral markets.)

And then the flip side is: if you hear all this stuff and you do know how to piece it together, please reach out to me on Twitter, right? Let's talk. I'm always happy to talk to people and collaborate with people who are smart and interested in this stuff. And so if you hear all these random details I mention and you're like, oh my gosh, that all makes so much sense now, I suddenly know what to do, that's great. I'm glad I made you money. Please, DM me to return the favor.

But back to the election. I was long Mamdani, and it was a very big win.

Leftists always say, “We're gonna activate the youth who are super left-wing but they don't want to vote Democrat,” and it never works. I always thought it was a stupid theory, but it actually worked for Mamdani.

Early-vote data showed a huge youth-turnout surge. It was obvious in the voter file. He changed the electorate. But polls underestimated him because they won’t adjust screens for new young voters.

I also didn’t think Sliwa would drop out. I was somewhat skeptical that Adams would drop, but not completely surprised. But I never thought Sliwa would go.

Kalshi: Right, he’s really got nothing else to do. So why would he drop out?
Iabvek: Early on, I was curious about Sliwa finishing second. Cuomo was consistently losing Democratic support, so I thought Republicans might give up on consolidating around Cuomo, who they hate, and go for who they actually want: Sliwa. But obviously it didn’t happen.

In safe blue states, GOP primaries nominate maniacs all the time. It’d be easy for Sliwa to get second if they had that mentality.

Kalshi: Thoughts on the New Jersey race?
Iabvek: Democrats are extremely likely to win. I was buying in the mid-80s. It’s possible the price drops if people get hyped, but if I think fair value is 96, it makes sense to buy high and keep buying as the price goes down. And maybe this will look completely idiotic when the GOP wins. But it’s hard to see either candidate straying far from generic Democrat vs. generic Republican. The turnout environment for Democrats in off-years is incredibly good. I think generic D in New Jersey should win by double digits in 2025. So it’s hard for me to see Sherrill losing.

One helpful data point is the 2022 House results. Republicans won the national popular vote, but Democrats still won New Jersey by 10 points with lower minority turnout. People confuse “trending right” with 2024 minority swings toward Trump. And I don’t see anything Trump has done for lower-income Latino voters. He didn’t lower prices. His immigration policy won’t help.

Also, the Arizona 7th special showed a 40-point swing back in Latino areas like San Luis, which had swung strongly to Trump last year. So it’s hard to see Trump maintaining that. New Jersey’s rightward trend should be over.

Kalshi: Anything else you’d like to touch on, any last-minute advice, or any markets we didn’t cover?
Iabvek: Not really, I think that’s everything. This was a lot of fun.

Kalshi: Good luck trading on Tuesday.

Follow Terry Oldreal on X: @realOldTerry
Follow Kalshi on X: @Kalshi

The opinions and perspectives presented in this article belong solely to the author, who is using a pseudonym and cannot trade on Kalshi, as well as the interviewees. This is not financial advice. Trading on Kalshi involves risk and may not be appropriate for all. Members risk losing their cost to enter any transaction, including fees. You should carefully consider whether trading on Kalshi is appropriate for you in light of your investment experience and financial resources. Any trading decisions you make are solely your responsibility and at your own risk. Information is provided for convenience only on an "AS IS" basis. Past performance is not necessarily indicative of future results. Kalshi is subject to U.S. regulatory oversight by the CFTC.