When Automation Met Hollywood

Before Trading Places became a cult classic movie about a hustler and a broker swapping fortunes, it began with a simple dinner bet between a flashy Hollywood producer and a methodical Hungarian engineer who was busy reinventing Wall Street.

It was the early 1980s. Thomas Peterffy was sitting across from film producer Aaron Russo in a New York restaurant, explaining that his trading floor ran so efficiently on computers that almost anyone could step in and make successful markets. Russo found that amusing and joked that even his friend, the filmmaker Melvin Van Peebles, could handle it. Peterffy called his bluff. They settled on a ten thousand dollar bet, Van Peebles joined the firm, and reportedly did just fine.

Not long after, Russo produced Trading Places. The central wager in the film mirrored the one he’d made over dinner almost exactly.

While everyone else on the trading floor was shouting orders across the pit and scribbling on scraps of paper, Peterffy was wiring up machines to outthink them. A programmer by background, he had bought a seat on the American Stock Exchange in 1977 and walked onto the floor as an independent options trader. What he saw was barely organised chaos, but he understood that chaos was opportunity if you could move logic through it fast enough.

He started by writing programs that calculated the fair price of options in real time. The problem, though, was data. In the 1970s, market prices came through ticker tapes meant for human eyes, with no direct way to stream them into a computer. So Peterffy built a device that could do it. He wired a small computer to read the electrical impulses from the ticker and translate them into signals his software could understand. Each time a trade occurred the ticker sent a pulse and his homemade circuit converted it into live data. His computer became one of the first automated pricing engines anywhere, scanning the market continuously and printing out opportunities the instant they appeared.

But then came the governance problem.

The Exchange refused to let computers place orders directly. They said every order had to be typed in by a person at a terminal. Peterffy argued that the trades were valid and accurate and the only difference was that a machine had calculated them. He even offered to use a mannequin if it helped satisfy the rule that a human had to be present. The Exchange was not amused.

So he found another way.

He wired an electromechanical keyboard to his computer. When the program identified a trade, the machine physically pressed the keys to enter the order. To everyone watching, it looked like a very fast trader hammering away. When officials eventually discovered what he had done and banned mechanical devices, he modified the circuitry to introduce random pauses and imperfect timing between keystrokes. It performed just like a person, and the orders kept flowing.

His firm, Timber Hill, became Interactive Brokers, one of the first firms to automate trading end to end. And the principles he wired together by hand are the same ones running through AI systems today. He just did it with wires instead of APIs.