Thomas Peterffy owns 75% of a brokerage worth $160 billion, and he is putting money into land because he puts a 20% chance on the United States taking private property away.
The usual hedge against that fear is gold. He rejects it, on the grounds that a government confiscating property will simply make owning gold illegal and require it to be handed in.
"I think land is the logical asset."
Peterffy grew up in communist Hungary, where his family's land was taken and then, 45 years later, given back — because land was the one asset the state had not managed to destroy.
The full interview is covered here so you can skip it. 47 minutes of audio, 15 minutes of reading.
Here are the 13 predictions that matter.
👤 Guest: Thomas Peterffy, Chairman of Interactive Brokers, who founded the firm in 1978, still owns 75% of it, and was described by the host as a digital trading pioneer
🎙️ Host: Wilfred Frost, lead presenter at Sky News and CEO of Paradine Productions, who anchored CNBC's Closing Bell until 2022
📰 Published: 16 September 2026 on YouTube (The Master Investor Podcast) and the show's own feed
🔴 YouTube | 🟣 Apple Podcasts | ⏱️ 47 min | ✅ Time saved: 32 min
Key Takeaways
Peterffy expects prediction markets to become the majority of Interactive Brokers' earnings
He built a play-money version a decade ago, was told it would cost him a banking license, and watched Kalshi and Polymarket take the idea
He puts a 20% chance on democratic socialists taking power and seizing private property, and is buying land as the hedge
Gold fails the test, he said, because owning it can simply be made illegal
Nvidia at 27 times earnings with 70% earnings growth expected is a low valuation, in his reading
The hyperscalers will have to write down much of the computing hardware they are buying
None of them will stop, because each needs to be one of the two survivors
He is against slowing AI development, because China will not slow down
The level of interest rates barely matters to Interactive Brokers, because it lends and borrows at a fixed spread around the benchmark
What worries him is leverage he cannot see — the over-the-counter book at the banks, not margin at his own firm
Current AI models cannot forecast, because their weights are frozen at training and cannot take in new information
1. The Professional Trader Bet
Frost opened on the performance: roughly 40% a year of underlying earnings growth since 2022, and a share price to match. Peterffy's explanation goes back to a bet placed 33 years ago.
"So when we started the brokerage business some 33 years ago, we did that with the idea of providing a platform to floor traders on the various equities and options and commodities exchanges to continue their business as the floors go electronic."
He was early by a decade. "Conversions happened between the year 2000 and 2010 and that was the time when we began on boarding our first target customers."
The target customer was specific: "So these were people with capital from half a million to $5 million and they conducted a regular daily trading business." Other brokers were handing customers order tickets to fill in; Interactive Brokers had to run several limit orders at once across different products, and support arbitrage, pair trading and shorting.
Only after that did the platform open to everyone, which brought in hedge funds above that range and retail clients below it, but he said the professional trader is still the client whose needs set the roadmap.
His theory of the growth is that sophistication sells the product: "So that is the reason we can say that the more investors know about the ins and outs of the mechanics of investing the more people realize that in order to maximize their returns it is not enough to pick the right investment strategy but they also must pick the right platform to implement it."
Frost's framing of the guest, for context: "My guest today is a titan of markets, a digital trading pioneer who founded Interactive Brokers in 1978 and built it to sit at a market cap of $160 billion today." He added that "Thomas still owns 75% of the company and therefore is worth over a hundred billion dollars himself."
2. Rate Levels Don't Matter
Frost put it to him that net interest income drove the earnings surge as rates came off the floor, and asked what happens if they go higher still. Peterffy said the level is close to irrelevant to the firm.
The firm lends and pays at a fixed distance either side of the benchmark rate, so the spread survives whatever the benchmark does.
"So we don't care if fed funds are 5% or 10% or 20% or 1%."
3. Hike to Prove Independence
Asked whether the Fed should raise rates the following day, Peterffy separated what he thinks should happen from what he thinks will.
His theory of the inflation is that it is an oil story and the oil is a war story. Shutting down Iran's ability to attack the Strait of Hormuz would, on his account, do more to bring inflation down than a rate rise.
"It may be better to run a hotter economy while we are still fighting the war so that we can easier cope with unfavorable economic fallouts." Once the war ends, he said, you can see whether lower oil prices bring inflation down on their own.
He acknowledged the bond market is calling for higher rates and said he is unsure how much that should count.
"Well, I think the Fed will probably raise rates, although I'm not, as I said, I'm not I would not be in favor of that, but they will have to do it just to demonstrate that they are independent."
4. Gradual Yields Are Fine
Frost described the long end as flashing red: "Clearly the longer end of the bond market at the moment is really flashing red. 5% on the 10-year just crossed higher still for the 30 year and it's a global issue." Peterffy was unmoved.
Asked directly whether the pace of rising yields worries him, he said it does not.
His condition is the speed, not the level: "Well, if they rise very sharply suddenly, then that could cause all kinds of problems. But as long as it's gradually going up as it is currently, it's okay with me."
Pressed on whether the moves since August still count as gradual, he said they do.
5. Nvidia at 27x Is Cheap
Frost asked whether one theme — AI — can carry equity markets for another five years the way it has for the last five. Peterffy said the valuations are not the problem.
His arithmetic is the growth rate against the multiple: "So if you look at a company like Nvidia 27 times earnings and they stated expectations of increasing earnings by 70% that is a relatively low valuation."
He expects a milder version of the same everywhere else, as AI spreads through the economy — not a bet on AI companies but on AI raising productivity and earnings at ordinary ones.
The frontier is not the dependency: "And even if the frontier models would stop progressing at this moment there are so many open-source models all over the place and adopting them by all the other companies will result in huge increases in earnings and productivity."
"So I'm basically very optimistic."
6. Compute Gets Written Down
Frost noted Peterffy's reputation on Wall Street for capital discipline — "Thomas, you're quite famous in the business which Wall Street has always celebrated for not wanting to waste money to being very careful with the money you invest and making sure the return on capital is attractive" — and asked whether the hyperscalers have been reckless.
His model of their behavior is a two-survivor race. Each believes only one or two will prevail, so each has to buy all the computing capacity it can.
"So at this point, none of them are willing to give up on not being number one, right? So they have to keep buying compute and that eventually will contract the price of compute. So it basically is going to benefit tremendously everybody else and these companies of course will have to write down much of the compute that they bought."
The claim has two halves, and the second is the investable one: cheaper computing power for everyone outside the race, paid for by write-downs inside it.
7. Don't Slow Down for China
Asked about chief executives calling for regulation and a slowdown, Peterffy separated the two and rejected one.
"China is not going to slow down."
"So this is basically a competition between the two hemispheres", and on the slowdown itself he was flat: "I would like us to win."
On the risk of a catastrophic AI security incident, he was dismissive: "I don't really believe in that. Hacks, they've been hacking systems for a very long time and it hasn't really caused substantial problems."
On his own firm's exposure, the standard is containment rather than prevention: "Of course, we're continuously worrying about it and try to take all the steps that we must take to make sure that we are not going to be hacked. But I mean, you can never be 100% sure that's not going to happen. You have to make sure that even if you do get hacked, it's not going to cause too much damage before you discover it."
8. Prediction Markets Win
The longest stretch of the interview is about a product Interactive Brokers launched in 2024 and a decision it made ten years earlier.
Interactive Brokers built a prediction market a decade ago and never released it for real money. Consultants working on the firm's banking license application found out and warned that going ahead would cost the license.
"So we actually came out with it, but we came out with it with phantom money." That version, he said, was real enough for the founders of Kalshi and Polymarket to see the idea and build it for real. "so now they are way ahead of us of course."
When Kalshi got a CFTC license, he tried to buy the company: "And so I thought that well to get licensed by the CFTC would take about two and a half or 3 years. So I thought I'd rather buy them than do it myself. And I tried to buy them but they wouldn't entertain an offer. So we went to apply for our own license and we got it."
The strategic split is sports against finance. Most of the volume at the venues ahead of him is sports betting, which he said Interactive Brokers was never interested in — and which now faces state opposition and a Supreme Court case he expects to be decided in the late spring of next year, over whether sports bets are swaps or bets requiring state licensing.
His rationale for the product is that prices alone do not answer an investor's questions: "So in our view the economy is the major determining factor for the stock market, right?" Companies sit inside regional economies, which sit inside national ones, inside the global economy, inside a social and climate environment — and a stock cannot be valued without a view on all of it.
He said the financial markets are not yet deep enough to represent genuine consensus, but that they will be, as more businesses with real exposure to these questions start hedging them.
His prediction for the expert class is that they stop selling opinions and start trading them: "Right. They should take their own positions based on what they think and the prediction markets will tell the world what the experts think."
Making markets in illiquid contracts leaves the firm exposed, he acknowledged, and it is not an attractive business today. He expects it to take off regardless.
The forecast that carries the section: "I think well I don't know exactly 10 years but in the fullness of time it's going to be the majority of our earnings will be derived from prediction markets."
9. The Midterms Market
Frost recalled that on his previous appearance Peterffy thought the market was underpricing Republicans holding both chambers, and asked where he stands now.
"Well, it's still I mean the odds don't look good for people who are Republicans and it does it certainly doesn't look good."
On whether that market is deep enough to be believed: "Yes, the political markets are certainly deep enough to reflect the concern." He added that he thinks they are more accurate than polls.
His reading of what they now say: "Well, no, they are pointing to the House of Representatives becoming Democratic and the Senate remaining Republican. We shall see what happens."
10. The Leverage He Can't See
Frost quoted back a line Peterffy had given a Bloomberg podcast — that leverage is the single biggest risk that keeps him up at night — and asked how often he thinks about it.
Not at his own firm, he said. Interactive Brokers evaluates customer leverage continuously and liquidates immediately on any margin violation.
"But it's generally the leverage in the entire system that worries me." Once other firms have problems, he said, it permeates the markets and everyone ends up with one.
The reason he cannot size it: "Part of the problem is that we don't know what's out there, right?" There is no central place where it is all aggregated.
His diagnosis points at the banks' over-the-counter books rather than exchange-traded products. Interactive Brokers deals only in exchange-traded, centrally cleared products; a bank writes one contract with a counterparty and hedges with another, and when one counterparty wobbles nobody can work out the exposures.
"It's a spaghetti."
Asked whether the moves in US, Japanese and UK yields and in the yen have been big enough to catch someone offside, he declined to raise the alarm. There is always a little worry, he said, but he does not want to overstate it and does not think about it much at present.
11. Why AI Can't Forecast
Frost asked whether AI makes Interactive Brokers' technology lead easier for rivals to copy. Peterffy said no, and then made a broader claim about what these systems cannot do.
Interactive Brokers has made the major AI models available inside its platform, where customers can connect them to their own portfolios. He treats that as an advantage rather than a threat.
His limit on AI in securities analysis is architectural: "So the problem is with the current AI models is that the weights are all frozen." They are set at training time and cannot be updated one at a time, because changing one part of the vector space destabilizes the rest.
"So they are not enabling new information to come in and change any of the weights." The consequence, on his account, is that the systems cannot learn — and learning is what forecasting earnings requires.
He rates them as good at assembling and sorting facts and poor at projecting events forward, which is the gap he thinks prediction markets fill better than models do.
Frost's counter, which Peterffy did not contest, is that models could eventually take prediction-market prices for growth or recession odds as inputs to their own forecasts.
12. The Trust Charter Play
Interactive Brokers is applying for a bank trust charter rather than a full banking license. Frost asked whether that means courting duller clients.
The purpose is custody: only a bank can custody the assets of ETF issuers and mutual fund managers.
It is not a sleepy business, in his telling, because custody feeds securities lending: "Well, it's not so boring because it has great repercussions for our ability to lend shares to customers who need them, right? Because much of our interest income comes from short people who are shorting stocks for whatever reason they do that and that is substantially important for our performance."
Asked what he can offer that Northern Trust, State Street or JPMorgan cannot: "We offer everything that they offer because as a matter of fact we do it better."
His example is price transparency. Interactive Brokers publishes its short inventory and its lending rates online, so a client can see a 2% rate at his firm, call their bank quoting 3%, and get matched — after which, he said, many of them send him some business anyway.
13. Buying Land for Heirs
The interview's last substantive stretch is the one the episode is named for, and Peterffy's reasoning is historical rather than financial.
"Yeah, so as I have grown up in communist Hungary and my family used to own a lot of land before communism came in and then of course they took it away from us. They took everything away." Real estate, land, companies and businesses all went.
What came back 45 years later was land, and only land — specifically the agricultural land out in the country that the state had not been able to build on or ruin.
He expects the same sequence in the United States: "I am buying land so that after the democratic socialists take everything over and they ruin all the stuff they will eventually find out that it doesn't work and they will have to go back to protecting people's private property."
Asked what odds he puts on capitalism being eradicated, at least temporarily: "I think currently chances I would put them at 20%." He said he sees politicians of that kind winning power and expects more of them.
Gold does not work as the alternative, on his reasoning, because it cannot be taken with you and because ownership can simply be declared illegal and the metal required to be handed in.
Frost said he had assumed the land buying was an inflation hedge and was struck to find it was not. Peterffy confirmed the frame is inheritance, not inflation: "I think land is the logical asset."
Bonus Insights
On the re-rating from about 12 times earnings five years ago to about 30 times today, Frost asked whether the multiple worries him. It does not, because he does not follow it: "I think about the business. I think about, what kind of new things we can come up with and provide to our customers and how can we grow the business because the stock price, there's nothing I can do about it."
His closing advice is procedural rather than financial: "But you always have to have a plan and you always have to work on it and you have to know every morning you get up. You have to know what you are supposed to do that day and you have to consult your plan if you're temporarily confused." The plan, he added, stays open to revision when new information arrives.
He finished by telling listeners to learn as much about AI as they can and work out how to use it, which he called the thing to come.
Frost declined to re-run Peterffy's life story, having covered it on the guest's previous appearance about a year earlier, and pointed listeners back to that episode instead.
Peterffy's bottom line is that the things most investors are worried about — rate levels, AI valuations, bond yields — are not what he is hedging; the tail he is actually positioning for is the confiscation of private property, and the business he expects to define his company is a market for forecasts rather than a market for prices.
Products, Companies & Tools Mentioned
Interactive Brokers (Peterffy's firm, founded in 1978, which he said is worth $160 billion and of which he still owns 75%)
Nvidia (His example of a valuation that looks high and is not: 27 times earnings against stated expectations of 70% earnings growth)
Kalshi and Polymarket (The prediction markets he said took the idea after seeing Interactive Brokers' play-money version; he tried and failed to buy Kalshi)
Northern Trust, State Street and JPMorgan (The incumbent custodians he intends to compete with once the trust charter is granted)
Forecast Trader (Interactive Brokers' prediction-markets product, launched in 2024 and expanded in 2025)
Watch the full episode:
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: