Context
Luana Lopes Lara is co-founder, president and COO of Kalshi, a federally regulated prediction market. The company went from a $5 billion valuation to $22 billion in about a year. It is also fighting lawsuits from several states, including a $36 billion suit from New York, where it is headquartered. Host Bob Safian presses her on whether Kalshi is really different from DraftKings and FanDuel, even though about 75% of its volume is sports. For PMs, the episode is a case study in building a product inside a regulatory gray zone, designing incentives, and keeping a startup mindset at hypergrowth.
The Big Idea
Kalshi's bet is that a product built on an exchange model, where users trade against each other and the company earns only a fee, is structurally different from a sportsbook, and that this difference is what makes both its regulation and its information value defensible.
Because Kalshi does not profit when users lose, it says it has no reason to cap winners or shade prices. Lopes Lara argues the same incentive design makes the prices a more honest forecast than polls or media.
Key Insights
Overnight success took eight years
- What: the company started in 2018 but could not launch anything for four years, because it insisted on being legal and regulated from day one.
- Why it matters: the team spent that time working with the federal government on how to bring prediction markets to the US safely.
- Example: when Kalshi won its lawsuit against the CFTC (the US futures regulator), it could list many more markets. Lopes Lara says "we were very prepared when our time came."
Fast growth preserved the startup mindset
Lopes Lara says growing so fast was a good thing. Slow, steady growth makes a company start thinking "I'm a bigger company, I need to hire more people," and it takes a long time to notice the mistakes that follow. Kalshi kept an early-stage mentality and stayed small in headcount. She says that two years ago, before the election, Kalshi made well under $10 million a year. Now it transacts more than that in a day. Her stated priority is speed.
Exchange incentives differ from sportsbook incentives
- Exchange: users trade against each other. Kalshi does not set prices or odds, and takes a transaction fee.
- Sportsbook: revenue equals customer losses, so the house wants users to lose and caps winners.
- Result: Kalshi says it wants winners, because winners bring price competition. She also argues the visible order book makes prices more transparent.
Why national liquidity matters
Lopes Lara uses a stock exchange analogy. If you could only buy stocks on the New York Stock Exchange while physically in New York, prices would be worse and the market would not be liquid. She uses this to argue that state-by-state rules would hurt every participant, which is the core of her answer to the state lawsuits. She also notes that each lawsuit rests on a different legal theory, and that Kalshi won in the Third Circuit even though an appeals court sided against it on Ohio and Tennessee.
Hedging is a growing business
- Kalshi is hiring for a small-business hedging arm and has partnered with the US Hispanic Chamber of Commerce.
- Her favorite example: a California goat farmer hedged against legislation expected to raise labor costs, by taking the side that it would pass.
- When she asks small businesses why they do not use oil futures, the answer is usually "what is that, and how do I do that?" Futures were built for big institutions.
- She says the market is used to hedge gas, diesel, weather, sports and politics, and some individuals now hedge the weather on their wedding day.
Why prediction markets forecast well
Her argument has two layers:
- Incentive to research: if you win money for being right, you check the weather site, look outside and call your mom, rather than guess.
- Price competition: everyone who did research competes for the best price, and anyone who disagrees trades and moves the price. Markets run 24/7, so new news moves the price right away.
Kalshi published its own calibration study. A market that says 70% should come true about seven times in ten, and she says results are very good, in the high 90s close to resolution. This is Kalshi's own study, so treat it as the company's claim.
Most visitors come for information
Safian cited a report that 70% of visitors only look at data. Lopes Lara said it is probably higher, about 78%. She compares Kalshi to a substitute for Twitter or the New York Times, because it incentivizes truth while social media incentivizes clickbait. A sportsbook's price means little to a viewer because it carries the house margin.
Surveillance is core intellectual property
- Kalshi bans insiders proactively: a politician is blocked from trading their own election, and an athlete from their own sport.
- Every user gives name, address, date of birth, ID and employment verification, which she says makes insiders easier to find than on offshore platforms.
- Detection mixes AI and machine learning on trading anomalies (for example, a new account trading in an unusual way), user tips, and a compliance team that investigates, including social media checks and interviews.
- Cases cited: a White House teleprompter operator, and George Santos, who allegedly tried to trade on his own attendance at the State of the Union.
- Building this system with the CFTC was part of why launch took four years. She calls it probably Kalshi's most important IP.
Mental Models & Frameworks
Probabilities are not certainties
Lopes Lara stresses that a 40% chance is high: you would not walk outside if there were a 40% chance of being hit by a bus. A market that gives a candidate 60% still lets the underdog win. A poll lead of ten points maps to a market price above 90%, because the two measure different things. She says teaching this is Kalshi's main education challenge.
Bottom-up versus top-down forecasts
- Polls: an editorial board or pollster aggregates information and announces one number.
- Markets: millions of people each do research and push the price, so the number reflects what informed people believe.
- Tool: to read thousands of election markets at once, Kalshi built an index called K-Pow (balance of power) that shows sentiment trending Democratic or Republican over time.
Margin as capital efficiency
Margin lets participants trade without parking the full amount. Her example: a green energy startup worried about losing subsidies wants to hedge an election with $100 million. Someone must lock up $100 million for two years. Leverage makes it worthwhile for the other side. Kalshi has worked on margin for over two years, with regulators and clearinghouse models, and requires a separate application covering experience and network.
Trade-offs & Nuance
Speculation helps, but mechanics matter
She argues speculation drives liquidity in every market, from crypto to futures, and "there's nothing wrong with speculation." The line she draws is the mechanics: on an exchange, "the house" does not win. She also quotes DraftKings' CEO saying its latest earnings showed no cannibalization from prediction markets. Safian's pushback stands, though: users with 75% sports volume may see little difference.
Margin adds risk and access limits
Safian asked whether leverage makes Kalshi too dangerous. She calls the concern fair and says it is a worry for them too. Her answer is gating: margin is not available to everyone, and applicants are screened. Safian also noted in his closing that clickbait could influence the market itself.
Banning pushes activity offshore
She says banning prediction markets moves activity to offshore platforms like Polymarket, where names and addresses are unknown and surveillance is "impossible." Any concerns about the activity in the US would then be "100 times worse." This is her argument as a Kalshi executive.
Common Mistakes
Mistake: Poorly explained market voiding
- What happened: Kalshi listed a contract on whether Iran's supreme leader would die or be removed from power, and later refunded more than $2 million to users.
- Her view: the contract was not wrong in itself. Kalshi does not allow markets on war, terrorism or assassination, so if such an event happened, it would void all trades so no one profits.
- The mistake: the product did not explain clearly how voiding would work, which caused confusion.
- Fix: Kalshi is now far more selective about which markets it lists.
Practical Application
Treat compliance as product work
Lopes Lara's surveillance system took years to prove to the CFTC and is now core IP. When you work in a regulated space, build the compliance tooling as a product, with its own roadmap and owner, not as legal paperwork bolted on later.
Explain mechanics in the product
The voiding confusion came from users not understanding a rule. Review any feature with surprising outcomes (refunds, cancellations, voids) and add plain explanations at the moment a user takes the action.
Design incentives before features
Ask who profits when a user loses and who profits when a user wins. Kalshi's story relies on the answer. Write down the incentive for your own product's revenue model, and test whether it matches what you tell users.
Keep early-stage habits at scale
Her advice from the hypergrowth: stay small, keep speed as the priority, and keep checking whether you are hiring or adding process out of habit. Before your next planning cycle, list which processes you added in the last six months and which ones slowed shipping.
Questions to Consider
- Who profits when your users succeed, and who profits when they fail? Does your revenue model match the story you tell customers?
- If a regulator or critic grouped your product with a look-alike competitor, what mechanical difference could you point to that changes who benefits?
- Which rule in your product (a void, a cancellation, a fee) could surprise users enough that you would have to refund them?
- Where does your team rely on a top-down forecast, such as a single leader's number, that a bottom-up signal from many informed people could replace?
Bottom Line
Kalshi argues its edge is incentive design: an exchange that earns a fee, not user losses, gets honest prices and fits federal regulation. That bet took eight years of regulatory groundwork before growth exploded. The open question is whether regulators and the public will see the product the way its co-founder does.
Resources Mentioned
| Resource | Type | Why it was mentioned |
|---|---|---|
| Kalshi calibration study | Company study | Kalshi's own analysis checking whether a market's stated chance matches how often events happen. |
| Federal Reserve studies on prediction markets | Research | Safian cited them as saying prediction markets forecast better than most other methods. |
Case Studies Mentioned
Goat farmer hedges labor costs
A California goat farmer faced new legislation expected to raise labor costs. He used Kalshi to take the side that the law would pass, so if costs rose he would be paid. The lesson is that event contracts can act as insurance for small businesses that cannot reach institutional futures markets.
Chalamet's dentist chair ad
Timothee Chalamet wanted an unusual ad that made people ask "what's going on." Kalshi's team doubted it, since it ran around the World Cup, but trusted his creative intuition. Lopes Lara says it performed better than many of their other ads and competitors' ads. She frames it as a deliberate effort to tell the story differently and counter misconceptions about prediction markets.
Tools & Products
| Tool / Product | What it does | Why it was mentioned |
|---|---|---|
| Kalshi | Federally regulated prediction market exchange | The company Lopes Lara co-founded. |
| K-Pow | Balance of power index across election markets | Summarizes sentiment on Democratic versus Republican control in one number. |
| Kalshi API | Programmatic trading access | Lets AI agents trade, with the user responsible for the agent and all rules applying. |
| Polymarket | Offshore prediction market | Her example of a platform where insider trading is hard to surveil. |
People to Follow
Luana Lopes Lara
Co-founder, president and COO of Kalshi, described on the show as the world's youngest self-made female billionaire. She previously worked at Bridgewater and Citadel Securities and in personal robotics at the MIT Media Lab. She trained as a ballerina.
Notable Quotes
"It doesn't matter if you're having a good or a bad day, you need to go train." (Luana Lopes Lara, on what ballet taught her about running a company)
"There's nothing wrong with speculation." (Luana Lopes Lara)
"We were very prepared when our time came." (Luana Lopes Lara)
