Kalshi CEO Tarek Mansour on The Case for Prediction Markets | Ep. 48
Tarek Mansour (MIT/Goldman/Citadel/Bridgewater) on the 6-year regulatory war that built Kalshi. **Sued the CFTC in 2024 and won in October** — that ruling redefined the legal line between gambling and financial markets and is the foundation under every prediction-market product launching in 2026. The structural pitch: prediction markets are *not* gambling because the business model is transaction fees on a neutral marketplace, not customer losses against the house — so the platform's incentives align with healthy participation rather than promoted losses. Cited validation: a Fed paper now calls prediction markets 'the best gauge we have on the economy.' Real institutional use cases emerging — Florida Keys hurricane hedging (insurance has exited), Biden-era student-loan forgiveness hedging, S&P holders buying Republican/Democrat contracts to hedge election impact rather than selling the underlying. Kalshi runs at **~120 people with effectively no managerial layer** — co-founder Luana 'knows what 80-85% of the org is doing' on Slack — third lean-ops proof point in this issue.
Key points
- **The 2024 CFTC lawsuit win is the legal load-bearing wall under all of 2026 prediction-market activity.** The ruling redefined the legal test: (1) structure — open free marketplace vs house-vs-customer; (2) underlying — natural event vs artificially-created risk. The court parallel is the 1905 Supreme Court grain-futures decision (states tried to call grain futures 'gambling', SCOTUS said no — speculation is necessary for hedging markets to exist).
- **The structural argument why prediction markets aren't casinos.** Casino KPI = customer losses. Casino business *must* promote losses (algorithms find big losers, comp them, keep them coming back). Kalshi's revenue = transaction fee. Kalshi's incentive = volume = trust = healthy participation. 'I'm actually incentivised, if I want people to trade more, for the platform to be fair.' This is the policy/PR backbone the entire prediction-market sector will use over the next 12-24 months.
- **Insider-trading rules ported from equities.** If you have *direct control* over the event (politician trading on a bill they're voting on), banned. Trading on observed information (Walmart parking-lot satellite count) is *work*, not insider trading. Same line as the stock market and grain futures.
- **Fed paper validation.** Mansour cites a Federal Reserve research paper that referenced Kalshi-style markets as 'the best gauge we have on the economy.' A dispersed crowd of Main Street prediction-market users is now beating Wall Street on inflation forecasting. **This is the moment prediction markets crossed from speculation curiosity to officially-cited economic indicator.**
- **Hedging use cases that are real money:** (a) **Florida Keys hurricane hedging** — insurance carriers have pulled out, so homeowners buy Kalshi 'hurricane hits this town' contracts as the de-facto insurance product; (b) **Biden-era student-loan forgiveness** — borrowers smoothed out their cash-flow risk via the forgiveness market; (c) **S&P holders hedging election impact** — institutions now buy Republican/Democrat contracts to hedge an upcoming election rather than selling the underlying position. This is a brand-new permitted hedging instrument class.
- **Citrini scenario as the live example.** Mansour: 'Citrini's research scenario started at 11% on Kalshi, now at 33%.' Direct case where a published research thesis got priced into a Kalshi market and re-rated up 3x. Tradeable signal for hedge funds.
- **Infinite-markets thesis.** As society's complexity grows (Iran, Covid, AI, cyber), the number of dimensions feeding traditional asset prices grows. Single-asset prices have rising *entropy* — they can't capture all the inputs. Prediction markets price each dimension separately, then those prices feed back into more accurate pricing of underlying assets. Mansour: 'If you want to price Tesla, you have to price whether Elon leaves, autonomous-vehicle pace, deliveries.' Each becomes a market.
- **Kalshi runs at ~120 people with no managerial layer.** Founders work first/last, weekends. Co-founder Luana 'knows what 80-85% of the org is doing' via Slack in any 48-hour window. People self-organise to the top-N problems list (which gets re-listed dynamically). Slope > intercept on hires. Manage work, not people (Chesky frame). **Third lean-ops case in this issue (after AppLovin and the 20VC general framing).** Trade-off explicitly accepted: 'we take on more organisational chaos to avoid bureaucracy.'
- **The 6-year regulatory journey as a 'crucible moments' case study.** YC 2018 → first CFTC approval Nov 2020 (immediately rolled back by new admin) → election market blocked end-2023 → sued CFTC 2024 → won Oct 2024. 'Walking in the desert' for years. Decision to sue the regulator was the explicit anti-pattern: 20-person company suing its own government with no real product. Sequoia's 'crucible moment' framing applies. **Pattern recognition for any deeply-regulated AI/finance/health vertical: the regulator-sue is the unlock.**
Notable quotes
We decided to sue. All the bad things that were predicted happened. The audit that was supposed to be two weeks now is 18 months. Knife after knife. But the most important thing is we won.
If you're the casino, what are you going to do? Build algorithms to find the big losers and get them hooked. Our incentive is the opposite — we want a fair platform because we want volume.
The Fed paper itself says this is the best gauge we have on the economy. The people doing it are not Wall Street, it's Main Street.
What you were really trading was a reaction function — not whether Trump would win, but how the S&P would react to Trump. That's impossible to predict.
We're 120 people. There isn't really a managerial layer in the company yet. If you ask Luana what 80-85% of the people are doing, she knows because she checked with them on Slack in the last 48 hours.
Citrini's scenario started at 11% on Kalshi. Now it's at 33%. You have a market you can probe instead of listening to pundits battle on Twitter.
Themes
- The 2024 Kalshi v CFTC ruling as the load-bearing legal foundation for the sector
- Structural alignment: exchange transaction-fee model vs casino customer-loss model
- Institutional hedging use cases driving the next leg of growth
- Infinite-markets thesis: prediction markets reduce entropy in traditional asset pricing
- Kalshi's 120-person no-managerial-layer ops as the third lean-org proof in this issue
Mentioned
People
Companies
Ideas
- Kalshi v CFTC 2024 ruling redefining gambling vs financial market test
- 1905 SCOTUS grain-futures parallel
- Casino model = promoting customer losses; exchange model = transaction fee neutrality
- Insider-trading rules ported from equities (control vs observation)
- Fed paper citing prediction markets as best economic gauge
- Hurricane hedging in Florida Keys via Kalshi
- Student-loan forgiveness hedging case
- Election hedging for S&P holders without selling the underlying
- Citrini scenario priced at 11% → 33%
- Infinite-markets thesis (rising entropy of asset prices)
- Kalshi 120-person no-managerial-layer org
- Luana knows what 80-85% of org is doing via Slack
- Slope > intercept hiring
- Sue-the-regulator as the unlock anti-pattern