Trading and Exchanges Chapter 27: Floor Trading vs Electronic Trading Systems

Chapter 27 is a fascinating time capsule. Harris wrote this around 2003, when the debate between floor trading and electronic trading was still alive. The NYSE was building a new trading floor. The Chicago exchanges were still mostly pit-based. Reading it now, knowing how completely electronic trading won, is like reading someone in 1995 carefully weighing the pros and cons of email versus fax machines.

Harris does not pick sides. He lays out why each system has strengths, and some of the floor trading advantages he describes are genuinely hard to replicate electronically even today.

The Fairness Question

Harris starts with fairness, and breaks it into two parts: operational fairness and fair access.

Automated systems win on operational fairness, hands down. Computers follow the rules every single time. No exceptions, no “mistakes,” no convenient errors. Floor trading depends on humans being both skilled enough to follow complex rules in chaos AND honest enough to follow them even when cheating would be profitable. Most floor traders were honest, but every major floor exchange had scandals involving front running, fake trade assignments, or prearranged deals.

Harris tells a story about a guy named Eli who placed a spread order on DJIA futures at the Chicago Board of Trade. The reported trade price was wrong, costing him $500. When he called to complain, the floor magically “corrected” the error. Was it a typo? Intentional theft that got caught? Nobody knows. But in an electronic system, this story literally could not happen.

On fair access, electronic systems also win. Floor traders could see and react to market information seconds before off-floor traders. In busy futures pits, your physical body mattered. Big traders muscled into the best spots. Tall traders could see over everyone. Women with high-pitched voices had an advantage because high-frequency sounds cut through pit noise better than deep voices. Harris literally discusses how your anatomy affected your trading success. In electronic markets, none of this matters.

Electronic markets have their own access issues though. Traders with faster connections and computer-generated orders have advantages over manual traders. This was a hint at the HFT arms race that exploded after the book was published.

Why Floor Trading Was Hard to Kill

So if electronic is fairer and faster, why did floor trading survive so long? Harris gives two big reasons.

First, information exchange. This is the strongest argument for floors. When arranging large trades, brokers need to know things about the other side. Is this an informed trader? Is there more size coming? Will trading with this person cause immediate losses?

On a trading floor, brokers could ask these questions directly. A broker with a good reputation could credibly say “this is the full size, no more coming” and get a better price. If they lied, their reputation was destroyed. This reputation mechanism was genuinely valuable and very hard to replicate in anonymous electronic systems.

Second, negotiation speed for large orders. When two traders want to negotiate size without revealing their full hand, oral negotiation is lightning fast. They take turns proposing bigger sizes until one says no. Done. In electronic systems, you have to split orders into pieces and play a slower game. Some systems solved this with hidden order sizes, but it was never quite the same.

Cost and Scalability: Where Electronic Crushed It

Floor trading was absurdly expensive to operate. You needed a physical trading floor in Manhattan or Chicago. You needed trained floor brokers, sales brokers, runners, reporters, surveillance officials. And you needed electronic systems on top of all that anyway for order routing and trade reporting.

Electronic systems had high upfront costs but tiny operating expenses. Harris mentions you could buy a complete exchange trading system off the shelf for about $5 million. Once running, the main costs were telecommunications and data backup. That is it.

Scalability was the other killer advantage. An oral auction breaks down when too many people participate at once. In futures markets, they had a designation called “fast market” which meant: things are so chaotic that your broker cannot guarantee filling your order at the best price. Electronic systems handled any volume because computers process orders millions of times faster than people can shout.

The Historical Verdict

Harris ends the chapter diplomatically, saying it is unlikely one structure will dominate all trading. He predicted floors would survive for large institutional trades where information exchange matters, while electronic would take over everything else.

He was half right. Electronic did take over everything else. But it also eventually took over institutional trading too, through dark pools, algorithmic execution, and electronic negotiation systems that solved many of the problems Harris identified. The NYSE finally went mostly electronic in 2006. The Chicago pits closed for futures in 2015.

The most interesting line in the chapter is Harris noting that large floor-based markets might survive not because their technology is better, but because of the order flow externality. Everyone trades there because everyone else trades there. That is inertia, not superiority. Once electronic systems got good enough and regulation forced competition, the inertia broke.

Technology transitions in finance are slower than you would expect, but more complete than incumbents believe. The NYSE spent years insisting its floor was a competitive advantage. It was not wrong about what the floor provided. It was wrong about how long those services would remain irreplaceable.

Previous: Chapter 26: Competition Within and Among Markets

Next: Chapter 28: Bubbles, Crashes, and Circuit Breakers (Part 1)