Floor Trading vs Automated Systems: The Technology Shift (Chapter 27)
Previous: Competition Among Markets
In 1999, the Bangladeshi Stock Exchange replaced its trading floor with an automated system. At the same time, the New York Stock Exchange was considering where to build a new trading floor.
That contrast tells you everything about how complicated the floor vs. electronic trading debate really was. Chapter 27 of Trading and Exchanges breaks down the arguments on both sides, and honestly, the analysis holds up remarkably well even though we now know how the story ended.
The Core Comparison
Floor-based oral auctions and automated rule-based auctions are more similar than you might think. Both are order-driven markets that match buyers to sellers using similar rules. The main difference is the technology used to arrange the matches. On a floor, traders personally exchange information by shouting and signaling. In an electronic market, computers handle everything.
Since the structures are so similar, the question becomes: which technology is actually better? And the answer Harris gives is nuanced. Both have genuine strengths.
Fairness: Where Electronic Markets Win
Harris identifies two types of fairness. Operational fairness means the rules are applied uniformly and nobody cheats. Fair access means everyone has an equal chance at opportunities.
On operational fairness, electronic markets win convincingly. Automated systems do exactly what they are programmed to do. They implement trading rules without exception. They produce complete and flawless audit trails of every single action.
Floor markets, by contrast, depend on the skill and honesty of humans. And humans are imperfect. Every oral auction market has suffered trading scandals involving front running, inappropriate order exposure, fraudulent trade assignments, or prearranged trading by dishonest brokers.
Harris includes a great example about a trader named Eli who submitted a spread order for Dow futures. When he checked his execution, the numbers did not match his limit. After calling his broker, the price was magically “corrected.” Was it a typo? Did the floor trader intentionally mess up hoping Eli would not notice? On an electronic system, the story simply could not have happened.
On fair access, the picture is more complicated. Floor traders have huge advantages over off-floor traders. They can see and react to market developments seconds before anyone else. They can observe who is trading and guess why. In large futures pits, physically large traders control the best spots. Tall traders can see and be seen more easily. Traders with loud or shrill voices attract more attention.
But electronic markets create their own unfairness. They favor traders with good keyboard skills and, more importantly, traders who use computer systems to generate orders. Automated traders can monitor data feeds and respond in milliseconds. Many manual traders resent competing with machines.
The Convenience Factor
One of the biggest advantages of electronic markets is that traders can work from their desks. They can sit next to their phones, talk with colleagues, and consult any data system they need. Floor traders are stuck on the floor with limited access to information and communication tools.
Harris mentions that Microsoft stayed on Nasdaq partly for ideological reasons. The company makes distributed computing systems. Listing on a floor-based exchange would have been a bit ironic.
Capacity and Speed
Electronic systems are far more scalable than oral markets. When too many traders try to participate in an oral auction, it breaks down. Traders cannot keep track of who is quoting the best prices. They arrange trades that violate precedence rules. Futures markets even have a formal designation for this: “fast trading markets,” which is basically the exchange saying “we cannot guarantee your order will be filled at the best price because it is too chaotic.”
Automated systems handle any number of traders because computers process messages far faster than humans can. Traders do not have to track the best bid and offer. The system does it for them.
On speed, the picture is slightly more nuanced. Some floor dealers claimed they could shout a bid faster than they could type one. That might have been true. But automated systems complete the full trade cycle faster because they handle both order entry and record keeping simultaneously. In an oral market, traders must manually record price, size, counterpart, instrument, and time for every trade.
There is one area where floor trading genuinely was faster: size negotiation. When two traders want to figure out how much they are willing to trade, they take turns proposing successively higher sizes until one person says no. This back-and-forth is extremely fast face to face. In an electronic system, traders have to split orders into pieces to avoid revealing their full size, which is much slower. Some electronic systems addressed this with messaging features or hidden order sizes, but the floor still had an edge here.
The Information Exchange Advantage
This is where floor markets really shined, and it is the reason the NYSE floor survived as long as it did. Floor brokers could exchange information about their clients that electronic systems simply could not convey.
All traders want to avoid trading with well-informed counterparts. When a well-informed trader wants to sell, prices are probably too high. Anyone who buys from them will regret it. On a trading floor, brokers could ask about the other side. They could try to identify whether the counterpart is a pension fund (probably uninformed, safe to trade with) or a proprietary trading desk (probably informed, dangerous).
Large traders also cared about whether more size was coming. If you buy from a large seller and that seller has much more to sell, prices will keep dropping and you will lose money. Floor brokers with good reputations could credibly say “no more size behind this” and get better prices as a result. In electronic markets, there was no equivalent mechanism for credibly communicating this information.
Floor brokers also knew about potential traders who had not yet submitted orders but might be interested if approached with the right opportunity. This kind of informal matchmaking was impossible in electronic environments.
The Cost Structure
The cost comparison is straightforward but important. Electronic systems have high upfront development costs but very low operating costs. Off-the-shelf exchange trading systems cost about 5 million dollars. Once running, the main expenses are telecommunications and data backup.
Floor markets also have high upfront costs (building a trading floor, especially in expensive real estate) but then have massive ongoing operating costs. They need brokers, reporters, officials watching for abuses, runners carrying messages. All of these people need training. The brokers especially need extensive training in complex trading rules and procedures. Floor clients often cannot communicate directly with floor brokers, so they need sales brokers too, adding another layer of cost.
Electronic markets, by contrast, let clients access the market directly. Brokers in electronic environments mainly just guarantee and settle trades. They are not needed to operate the system.
Who Wins?
Harris does not declare an outright winner. Instead, he argues that different systems serve different clients.
Automated systems are best for active markets with high volumes of small orders. They are cheap, fast, and provide direct control. They appeal to traders who do not trust their brokers or do not want to pay for brokerage services.
Floor systems work best when traders need to exchange information before agreeing to trade. This matters most for large institutional traders. The NYSE was increasingly an institutional market, and its floor survived because floor traders provided genuinely valuable services that electronic systems could not match.
Harris also makes an important observation: the continued existence of large floor markets does not necessarily mean their technology is better. The order flow externality from Chapter 26 can keep an incumbent market liquid even if its technology is inferior. Liquidity attracts liquidity, regardless of the underlying system.
What Actually Happened
Reading this chapter in hindsight is fascinating because we now know the answer. Electronic markets won almost everywhere. The NYSE eventually went hybrid and then mostly electronic. The Chicago futures pits closed. Floor trading is a tiny fraction of what it once was.
But Harris’s analysis was not wrong. He correctly identified the floor’s genuine advantage in information exchange. What happened is that the cost and speed advantages of electronic trading overwhelmed that information advantage. And new electronic tools emerged to partially replicate what floor brokers did. Dark pools, negotiation protocols, and conditional order types addressed some of the information exchange problem.
The transition was not as clean as “screens beat floors.” It was more like electronic markets gradually absorbed the functions that made floors valuable while delivering everything else better and cheaper.
This post is part of a series on Trading and Exchanges: Market Microstructure for Practitioners by Larry Harris (Oxford University Press, 2003, ISBN: 0-19-514470-8). This retelling covers Chapter 27.