‘Dark Pools’ shines light on electronic stock trading

Every time the image of the New York Stock Exchange trading floor and its clanging opening bell flashes up on morning business television, the great machine of American capitalism is on view.

Here is a trading floor — the great trading floor of the NYSE — teeming with people handling orders. Here is Wall Street at work, channeling the savings of millions of Americans and other investor capital usefully through the system.

I hate to break it to you, but this isn’t how the markets work these days. Not even close, as shown by this richly reported book, “Dark Pools: High-Speed Traders, A.I. Bandits, and the Threat to the Global Financial System.” Written by the Wall Street Journal’s Scott Patterson, it is published by Crown Business.

Only 22% of the shares listed on the NYSE are traded on the NYSE floor. Indeed, the proportion of NYSE-listed securities traded on any of the systems operated by NYSE Euronext, the company that runs the NYSE floor, is only about 33%, compared with 79% in 2005.


The rest pulses invisibly, electronically, through wires, data centers and computer servers on 50 other trading venues, including NYSE rival Nasdaq and lesser-known exchanges such as BATS Exchange or Direct Edge.

Further amounts slosh through a complex matrix of 200 broker-dealers that executes orders internally, including Knight, a big U.S. broker that has been known to handle more orders than even Nasdaq on a single day.

Then there are the 20 or so “dark pools” run by banks such as Goldman Sachs, Credit Suisse and Deutsche Bank, where — in theory, at least — asset managers can get large orders of shares done without fear of falling foul of the rapid-fire electronic traders that make up such a large part of exchanges.

The whole ecosystem is supercharged by rapid automated trading, fueled by computer algorithms programmed to buy and sell stocks at a particular moment. Stock holding periods that were measured in minutes, or days, are down to microseconds.


There are even algorithms — “algos” in the traders’ vernacular — that, like heat-seeking missiles, prowl across these multiple venues for the best price. Asset managers need this kit just to navigate the complexity.

This world is invisible to the average investor. And that is the problem with how the ecosystem of the markets — or market structure — has developed.

Through a combination of rapid technological advancement and regulation that allowed trading venues to compete freely, the basic function of markets has arguably moved even further out of the public eye than when trades were handled in noisy pits.

To be sure, the pits were hardly models of transparency and ethics. Investors were routinely ripped off as market makers demanded a wide spread between the bid and offer price for handling an order. Electronic trading has brought that spread way down and increased liquidity, with clear benefits.


Yet as Patterson reveals, the electronic markets that appeared to have democratized dealing when new trading systems emerged have themselves created a new set of problems.

The torrent of buy-and-sell orders that are spat out by computers, only to be canceled moments later as market sentiment shifts, is overburdening exchanges’ systems. And what useful purpose does this serve for price formation? And what of algos that run amok, possibly bringing the system crashing down?

The “flash crash” on Wall Street in 2010, when the Dow Jones industrial average plunged nearly 700 points before rebounding, all in a matter of minutes, should have convinced anyone who believed these were arcane considerations that this is an issue of intense public policy interest. After all, who are markets ultimately for?

Patterson takes us back to when electronic pioneers such as Island and Archipelago were challenging the cozy status quo on Wall Street.


We meet Josh Levine, an 18-year-old computer programmer with a severe buzz cut and worn tennis shoes, whose vision for Island was to make information and prices available to the masses. We also learn that placing and canceling orders was an issue even in 2001, when a study questioned the value of such “fleeting orders.”

Fast-forward to today and we read that the dark pools, which were set up as a place to avoid the slicing and dicing of automated orders, have themselves been infiltrated by the “bots.” Recently, NYSE Euronext — like many exchanges, a long-standing critic of dark pools — turned the tables with the launch of a dark pool for retail investors. Where this will end up is hard to predict.

Patterson’s prose sometimes has the overly breathless air of an airport thriller. But it is underpinned by an invaluable piece of timely journalism that should be read by regulators and anyone with a cent in the stock market. You may not save money by reading it, but you will never look at the opening bell on TV in the same way.

Jeremy Grant is a Singapore-based editor for the Financial Times of London, in which this review first appeared.