Competition Within and Among Markets: Who Wins the Trading Game (Chapter 26)
Previous: Internalization and Crossing
You might think there is one stock market. There is not. There are many. And they are all fighting each other for your orders.
Chapter 26 of Trading and Exchanges tackles one of the most fundamental tensions in market design: the battle between consolidation and fragmentation. Should all trading happen in one place? Or should multiple venues compete for business? The answer is more complicated than you would expect, and understanding it explains a lot about why modern markets look the way they do.
The Order Flow Externality
Here is the core concept that drives everything in this chapter. Harris calls it the order flow externality, and it is basically a fancy way of saying that liquidity attracts liquidity.
Think about it like a crowd forming around a free sample table at a grocery store. One person stops, which attracts another person, which attracts more. In markets, every trader who joins a particular venue makes that venue more attractive to the next trader. More buyers means it is easier to sell. More sellers means it is easier to buy. Everyone benefits from everyone else being there.
This creates a powerful gravitational pull toward consolidation. Markets naturally want to merge into one big pool. And in a simplified world where every trader is identical, that is exactly what would happen. Everyone would trade in the same place because it would be cheapest and easiest.
But traders are not identical. And that is where things get interesting.
Why Markets Fragment
Markets break apart because different traders have fundamentally different needs. Harris identifies several key dimensions where traders differ:
Size matters. Large institutional traders are terrified of revealing their orders. If the market knows a pension fund wants to buy 5 million shares, front runners will pile in ahead of them and drive up the price. Large traders prefer venues where they can control who sees their orders. Small traders, on the other hand, want maximum exposure. They want everyone to see their order so it fills quickly at the best price.
Information asymmetry. Well-informed traders want to hide in the crowd. They prefer anonymous consolidated markets where nobody can identify them. Uninformed traders, paradoxically, want the opposite. They want markets that can identify them as uninformed so dealers will give them better prices. This is exactly why payment for order flow exists. Retail orders get routed to dealers who offer slightly better prices because those orders are predictably uninformed.
Patience. Impatient traders want guaranteed execution right now and will pay the spread to get it. They love quote-driven markets with dealers always standing ready. Patient traders are willing to wait for a better price and prefer order-driven markets that enforce time precedence so their limit orders actually get filled.
Access. Some traders have direct access to exchange floors and information that off-floor traders cannot see. Traders who feel disadvantaged naturally gravitate toward venues where they have a more equal role.
These differences mean that no single market structure can perfectly serve everyone. So markets fragment. Multiple venues emerge, each catering to different types of traders.
How Fragmented Markets Hold Together
So if markets fragment, does price discovery fall apart? Not necessarily. Harris explains three mechanisms that keep fragmented markets working as a unified whole.
First, traders in each segment watch what is happening in other segments and adjust their orders accordingly. If a stock is selling for less on one venue, traders everywhere update their prices.
Second, smart order routing means traders send their orders to whichever venue currently offers the best deal. Buy-side order management systems constantly scan multiple venues to find the best execution.
Third, arbitrageurs specialize in moving liquidity between segments. When prices diverge across venues, arbitrageurs buy in the cheap market and sell in the expensive one. Their trading enforces the law of one price across all segments.
These three forces are powerful enough to make fragmented markets function almost like consolidated ones, as long as information flows freely between segments and some traders can choose where to route. Harris makes a key point here: even if some traders are locked into one venue, the market stays unified as long as other traders can move between venues.
The Network Externality Problem
The order flow externality is really a type of network externality, similar to what telephone companies and social media platforms experience. The more people on the network, the more valuable it is to each user.
Network externalities create massive barriers to entry. Once one market gets big, it is incredibly hard for a new market to gain traction. Nobody wants to be the first trader at a new venue. Even if the new venue has objectively better technology or lower fees, it cannot compete without liquidity.
Harris uses the spectacular failure of Optimark to illustrate this point. Optimark was a brilliant trading system that let traders express complex preferences about price and quantity. Institutional traders were excited about it during development. But when it launched, nobody sent orders. After burning through 406 million dollars, it shut down.
The lesson is harsh but clear: being better is not enough. You have to solve the chicken-and-egg problem of liquidity.
ECNs found a clever workaround. They integrated with Nasdaq so that limit orders placed on ECNs were effectively exposed to the entire Nasdaq market. This allowed them to piggyback on Nasdaq’s liquidity while offering their own services. But this trick did not work against the NYSE because the order-routing mechanics were different. That is why ECNs successfully competed against Nasdaq but struggled against the NYSE.
The Externality Problems
Competition among markets is not purely beneficial. Harris identifies several externality problems.
The most important involves secondary precedence rules. Time precedence protects limit order traders from being front-run by quote matchers. But time precedence only works within a single market. If another venue does not enforce time precedence, quote matchers can just go there. This creates a “race to the bottom” dynamic. Markets that want to protect limit orders cannot do so because competing venues undercut them with smaller tick sizes and no time precedence.
This is exactly what happened with decimalization in the US. Smaller markets like the American Stock Exchange and ECNs moved to smaller price increments first because they had few limit order traders to protect. Larger markets had to follow to remain competitive.
There is also a regulatory services externality. Some markets invest in surveillance, insider trading enforcement, and price continuity. These are public goods that benefit everyone. But traders can avoid paying for them by routing to markets that do not provide these services. So unregulated competition tends to produce less regulatory oversight than is socially optimal.
The Consolidation Wave
Beyond the fragmentation of trading in individual stocks, Harris describes a broader trend of market centers merging. Two big waves of consolidation happened. The first followed the telegraph and telephone. Regional exchanges that once served every major city consolidated as communication costs dropped and traders could send orders to distant markets.
The second wave started in the 1990s, driven by the order flow externality, economies of scale in electronic trading systems, and the loosening of cross-border regulations. European markets were consolidating rapidly. And this trend has only accelerated since the book was published.
What This Means Today
Harris frames the debate perfectly. There are two types of competition in markets, and policies that help one can hurt the other. Competition among traders for the best price works best when everyone trades in one place. Competition among venues to provide the best services requires multiple markets. Regulators have to balance both.
This tension has not been resolved. It shows up in every modern debate about payment for order flow, dark pools, best execution standards, and market data fees. The concepts Harris laid out in 2003 remain the essential framework for understanding these issues.
The takeaway is that market structure is never neutral. Every rule, every design choice, shifts the balance of power among different types of traders. And whoever controls the rules has enormous influence over who wins and who loses.
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 26.