Interactive Brokers
A case study in the history of brokerage and advanced automation. Interactive Brokers converts about 77 cents of every revenue dollar into pretax profit because software, not people
Original Business Case Weekly articleA few weeks ago, in May 2026, I saw that Interactive Brokers had quietly switched on a single screen where a client can trade Kalshi, CME, and ForecastEx prediction-market contracts side by side, with the system automatically routing each order to whichever venue offers the best net price [1].
What makes Interactive Brokers worth our attention is the gap between how invisible it is and how large it has become. Most people have never heard of it, yet the whole company carries an equity value north of $120 billion, it runs pretax margins in the seventies that look like a software business rather than a broker, and it is still roughly three-quarters owned by one man [3][4].
That man is Thomas Peterffy, an 81-year-old who was born in a Budapest hospital basement during a Russian air raid, landed in New York in 1965 speaking almost no English, and learned to program in part because the manuals had a vocabulary of about a hundred words while English did not [5][6]. He is now among the thirty richest people on earth, and almost no one outside finance can name him [5].
I find him interesting for a reason that has little to do with the rags-to-riches arc, which is common enough. He spent his career trying to win at trading by deleting human judgment, and the lasting fortune came from something more boring and more durable.
Rivals copied his math and his speed within a decade. What they could not copy was the discipline to take the savings that automation produced and hand them back to customers, year after year, instead of pocketing the spread the way every competitor did.
This is a case study in the most underrated edge in business, which is the willingness to look like you are leaving money on the table for thirty years.
A Hundred Words of English, 1944–1977
Your call
You are Peterffy in 1977, a programmer who has never traded and can barely speak to the people in the pit. You hold a model you believe is better than theirs. Do you risk your savings to trade it yourself, license it to an established firm, or keep your day job and refine it further? What would change your mind?
