Skip to content

Podcast

#402 Thomas Peterffy- The $80 Billion Founder Who Automates Everything

Founders

Source ↗ ← All highlights
  • From Wartime Budapest To Early Hustles
    • Thomas Peterffy escaped post-war Hungary and learned capitalism from his grandmother’s library.
    • Early hustles like cutting gum and scavenging scrap steel taught him to monetize scarcity and survive. Transcript: David Senra And so Dom writes, I’m at the $108 million Aspen house of Tomas Petrfi. Once inside, I find Petrfi hunched over his chair. His piercing eyes look up at me in confusion as I was introduced. I thought we were doing this over Zoom, he said. My stomach dropped. Where have you come from, he asked. London. That’s crazy, he said, as if I’d come by boat. To a man who has spent the past 60 years automating as much of his business as possible, to the point where Interactive Brokers has 71% profit margins, my journey to meet him was an absurd Misallocation of resources. Peterfi is the 23rd richest person in the world. He pioneered automated trading and built one of the largest options market makers on earth. He is the reason that you can trade stocks in your pajamas. His second act, Interactive Brokers, is worth over $100 billion. I was surprised that he was surprised that anyone would travel to hear his story. He settled down in his chair and began to explain how he got his start in life by dragging a metal bathtub through the rubble of post-war Budapest. I was born during a Soviet bombing raid, Petrfi began. It was September 30th, 1944, and the Red Army was pushing into Hungary. He remembers nothing until he was five, by which time Hungary belonged to the communists and his father had vanished, having divorced his mother and fled the country when Petrfee was Two. I often remember my mother crying, and I’d ask, Mom, why are you crying? And she’d reply, we’re going to starve to death. And she was dead serious about that. (Time 0:00:40)
  • Automation Multiplies Individual Productivity
    • Peterffy discovered that breaking work into repeatable steps lets machines do the heavy lifting.
    • Automating repetitive calculations cut tasks from 20 minutes to 30 seconds. Transcript: David Senra Routine calculations could take up to 20 minutes. Yet in the corner of the office sat a $3,000 solution that nobody wanted to touch. The Olivetti Programma 101 weighed 20 pounds and looked like an oversized cash register. It was one of the first desktop computers. When Petrify volunteered to tackle the unused machine, nobody objected. I figured it would be easier to learn than English, he said. The computer performed basic arithmetic and printed results on paper receipts. It could also store simple programs. The first night, Petterfi took the manual home, and he was relieved to find it contained only 100 English words. The rest was equations and diagrams. The machine’s logic immediately appealed to him. Break each calculation into steps, record those steps on a card, feed the card into the slot, enter the surveyor’s numbers, and receive an answer. Petrify began writing his first programs. When a card finally worked, he labeled the function and added it to his growing stack. Within weeks, he had built a library of programs for the office’s most common calculations. What had taken 20 minutes by hand now took 30 seconds. (Time 0:04:43)
  • Automated Quotes Beat Trader Intuition
    • Peterffy built a system that printed bid-ask quotes for traders and sent runners to the pits.
    • He used math-driven pricing rather than trader intuition to capture consistent profits. Transcript: David Senra Petrfi arrived at Jarecki’s office armed with a book called Gold that he had been using to learn about commodities. Jarecki explained his observation that the price of silver was volatile, but stayed within defined boundaries. He wanted somebody to write a program that could model what would happen if they bought every downtick and sold every uptick, profiting from silver’s nervous energy. To answer Jarecki’s question, Pederfee needed data. So he goes down to COMEX, which is a commodity exchange, and this is what he finds. He found a prehistoric setup. Reporters seated in a circular pit, dictating prices through radio headsets to clerks on scaffolding who scrawled numbers on the walls. At day’s end, the prices were copied to paper. So once I learned about Tomas Pederfee, not only did I read this profile, but I read every interview I could find with him and every other profile about him. And in another profile, I discovered one of the most important sentences that I think is key to understanding Pederfee and what makes him personally interesting to me. And he said, on Wall Street, I feel like I’m Alice in Wonderland. Nothing makes sense. Everything is mixed up and different than the way that I think it should be. And so Pederfee’s assignment is very simple. Trade silver and try to make money. And Pedrafi has no idea how to do this. He says, I had a horrible time. How do you decide when you’re going to buy? And how do you decide when you’re going to sell? He had no framework for decisions. No system beyond intuition that he did not possess. Jarecki also had a larger vision, but we will see not as large a vision as Pedrafi had. Jarecki wanted his nascent business to become a bullion dealer, quoting continuous silver and gold prices to banks and traders in New York, London, and Hong Kong. The exact problems that a computer could solve best. Peterfi designed the system from scratch. So there’s another quote in one of these other profiles that I read about Petrfi where he talks about that he does not consider himself a trader. He says, I’m a computer programmer and so are all the most important people in my company. So it says, Petrfi designed the system from scratch. Petrfi’s programs ran the data through proprietary equations and printed fresh bid-ask quotes on green bar paper. Runners then grabbed the sheets and raced them to the trading pit where clerks gave live prices through hand signals. Other firms relied on their traders intuition. Peterfee built a machine that ran on math. (Time 0:07:50)
  • Hacking Market Data From His Desk
    • Confined by a knee injury, Peterffy hacked a Quotron feed with an oscilloscope to read prices in real time.
    • Feeding live market data into his algorithms let him spot profitable options faster than competitors. Transcript: David Senra The second one is going to be Interactive Brokers. That same year, Petterfi tore several ligaments in his knee through a series of accidents. He found himself unable to stand on the trading floor for long periods. Confined to his office, Petterfi spent hours watching his Quatron machine, a beige box that pulled up one stock price at a time over a dedicated phone line. He asked Quatron to sell him the data feed, and when they refused, he helped himself, cutting the wire and attaching an oscilloscope. I had to look this up. An oscilloscope is an electric test instrument that visually displays the variation of voltage over time. So listen to this. I want to buy your data. You say no. And it says he helped himself. He essentially hacked the system, cutting the wire and attaching the oscilloscope. The oscilloscope sat on his desk like a small television with a grid hashed over its face. When he attached the probes to the severed line, green traces swept across the screen, displaying the electrical pulses carrying each stock price. This is insane. Every number had its own signature in spikes and dips. He studied the patterns, matching each trace to the prices appearing on his Quotron. Maybe they were right. He was a mad scientist, or he is a mad scientist. Soon, his computer was being fed price changes across the entire market in real time. This is the early 1980s. With that stream of data, his algorithms could spot profitable options trades faster than anyone else. (Time 0:15:27)
  • Creative Workarounds To Exchange Biases
    • He hired attractive women as traders to bypass exchange biases and get orders filled.
    • When specialists caught on, he invented handheld trading devices to keep his edge. Transcript: David Senra And in that clubby world, Petterfy was not one of the boys. His solution was calculated. He hired six tall, beautiful women to trade for him. The specialists who had ignored his bids suddenly to fill his new employees’ traits, which Pedderfy was delivering by phone. Everybody loved the women, he said. We were making money hand over fist. And so there’s, I know Dom Cook, the writer of this profile. And so I asked Dom to send me as much background information on Pedderfy as possible to help me make this episode. There is a story that did not make it in the profile, which is nuts. And it’s exact. It’s about this time in Petterfee’s career. So it says in 1982, Petterfee was out to dinner on the Upper East Side with a friend. When they walked into the restaurant, three men at a table near the entrance spotted his friend and invited them over. All three of the men sitting at the table worked in show business. Peterfi knew none of them. One was Aaron Russo, the film producer. Another was Melvin Van Peebles, the filmmaker. Russo turned to Peterfi. So what do you do? Peterfi explained that he was a trader, but that he had injured his knee and couldn’t stand on the floor anymore. So he hired attractive women to execute his trades. They took instructions over the phone and relayed them to specialists on the exchange floor. You mean anybody could do this, Russo asked? Petterfi shrugged. Theoretically, yes. Russo put his hand on Van Peeble’s shoulder, the filmmaker, okay? You mean Melvin here could do it? I think so. I’ll make you a $10,000 bet, said Russo. You hire Melvin. If he lasts a year, I’ll pay you. Petterfee agreed. Melvin went through Timber Hill’s two-week training course, learning to take Petterfee’s instructions and relay orders to specialists. Then he was sent to the American Stock Exchange floor, where he quickly gained popularity. He spent a full year trading for Timber Hill and did a fantastic job. Petterfee collected Russo’s $10,000. Here’s the insane part. A year later, Russo produced the movie Trading Places, starring Eddie Murphy and Dan Aykroyd, which was about a wealthy broker and a street hustler whose lives are switched as part Of a bet by two rich financiers. The film earned $120 million in its first year. That takes place exactly where we are in the story. Back to the story. The honeymoon ended when the specialists finally grokked what was going on. They delivered an ultimatum. If Pefferty wanted to keep trading, he would have to become a market maker, maintaining constant bid and offer prices instead of cherry picking only the most profitable options to Trade. Market making required split second responses to price movements. But his traders took their orders from algorithms running in his office. How could they make markets without direct access? And as he does with almost every problem throughout his career, he comes up with a novel solution. In 1983, 27 years before Steve Jobs unveiled the iPad, Petterfi invented the first handheld trading computer. He built rectangular boxes, each about the size of a hardcover encyclopedia. Inside were rows of transistors and circuit boards powered upon a crude touchscreen. Each morning, Petterfi lined up the devices along his desk, plugged them in, and uploaded fresh market data and options prices. Then he handed them off. On the floor, when specialists demanded quotes, his traders glanced at their screens, answered with prices, and tapped again to log the trades. But the cycle required constant feeding. After five trades, the devices had to be updated. The clerks would sprint the two blocks between the American Stock Exchange floor and Petrfi’s office, carrying the computers in satchels. He would upload the trades, recalculate exposures, feed in new prices, and then send them racing back. (Time 0:17:08)
  • Mechanical Typist To Keep Automation Alive
    • When Nasdaq forced keyboard entry, Peterffy built a camera-and-robot-typing rig to automate order submission.
    • The mechanical fingers typed buy and sell orders faster than any human could. Transcript: David Senra Now, the way he built the very first automated trading system in Wall Street history is going to be very familiar to you and I. He’s going to hijack the data. So a NASDAQ employee who is making a routine visit to one of its fastest growing clients’ office sees that there’s no human in sight. Unbeknownst to Nasdaq, Petrify had hijacked the terminal’s data line. For most traders, the terminal was just a screen and a keyboard, a way to type in orders one at a time. But Pederfee had wired it into his own computer, pulling live prices straight from the feed, running them through his algorithms and sending trades back out through the same cable. The Nasdaq employee gave him one week to make it right. All trades, he insisted, had to be entered by a keyboard and typed one after another, just like everyone else. And so go back to his quote about Alice in Wonderland. He says, on Wall Street I feel like I’m Alice in Wonderland. Nothing makes sense. Everything is mixed up and different than I think it should be. And so in Petrify’s view, he has NASDAQ trying to make his system worse. Listen to what he does here. This is probably the funniest thing that he ever does. This guy is so crazy. Petrify and his team worked every night for a week. They mounted a camera above the terminal screen to read the prices, then built a frame of metal arms and tiny motors suspended above a keyboard. When the computer spotted a trade, signals fired through his invention, and the metal fingers began to type like a mechanical spider. When the NASDAQ man returned, he found an office transformed. The suspicious silence was gone, replaced by the violent percussion of automated typing. Pedderphy’s creation attacked the keyboard in bursts. Rat-a Pause. Rat-a Pause. Rat-a Each sequence spelling out buy and sell orders faster than human could think, let alone type. (Time 0:22:53)
  • Build Infrastructure Before The Market Exists
    • Interactive Brokers anticipated a market shift to screen-based trading before mass demand existed.
    • Building infrastructure early positioned the firm to dominate as exchanges automated. Transcript: David Senra In 1993, he launched Interactive Brokers. This is one of the craziest lines in the entire profile. For the better part of the 1990s, Interactive Brokers was an elegant solution to a problem that didn’t yet exist. The infrastructure was ready. The technology was sophisticated, but the market, particularly in the United States, remained stubbornly analog. Around the turn of the millennium, the great automation of American exchanges began to accelerate. NASDAQ had been born electronic, but now the most traditional bound exchanges were surrendering. Even the New York Stock Exchange was giving way to the hum of computer servers. Wall Street was becoming a screen-based business. The floor traders who had once mocked Pedervi’s folded sheets and handheld computers found themselves staring into obsolescence. This worked to Pedervi’s advantage. They knew that I was an honest business person, and they needed a way to continue their business on a computer from their office. So they became our customers. (Time 0:25:53)