Man against machines: Trading company shares

Never before has the proportion of equity trades that go through the New York Stock Exchange been as low as it is today. In the past few months it has fallen to fewer than one third of all trades. Instead, orders are regularly routed though a new generation of trading platforms that do not always have the same requirements to show what investors are wanting to buy and sell. While television news channels still report the daily stock market moves from the floor of the exchange, the action is increasingly taking place in “dark pools” managed by computers, often designed by broking firms themselves, away from recognised exchanges.

Even more profoundly, the number of orders that come from traditional fund managers, who have decided to buy or sell a particular company share after thinking over its investment potential, has also never been lower. In the US, more than 50 per cent of all trades are now generated by computers, many inside hedge funds or other new trading firms, who are buying and selling small numbers of shares at very high frequency across these splintered trading platforms. By putting in orders every few milliseconds, they harvest millions of tiny profits that quickly add up to a bumper payday.

Computer trading is revolutionising the equity market in ways that are profound, but the fact regulators are only now catching up strikes many as profoundly scary. Under political pressure, Mary Schapiro, the chairman of the Securities and Exchange Commission, said the agency would consider banning one practice that has grown up since the rise of the machines ? “flash trading”. This is where exchanges and other share-trading platforms offer information about a customer’s order to buy or sell shares to a select group of traders for a half a second before revealing it publicly. The idea is to find a corresponding seller or buyer to fill the order before the exchange has to give up on the trade, but critics say that split second is enough to give the select group enough secret data about the likely direction of a share price for their computers to use to make a quick penny.

The emergence of flash trading has been one of the unanticipated consequences of the explosion in high-frequency trading, which has been enabled by a succession of regulatory changes since the emergence of the first electronic trading in the 1980s. The trend went exponential in 2006 when the NYSE introduced electronic trading too; by 2008, the average volume of shares traded in the US every day had doubled, and the number of individual orders was up by 150 per cent.

“Well over half of average daily trading volume is now accounted for by high-frequency traders,” says Sang Lee, of the consultancy Aite Group, who likens the development of sophisticated trading programs to an arms race. “This takes a lot of commitment: from sheer brainpower to get these programs written, to super-fast computers and even co-location facilities so you can put your servers ever closer to those of the exchange, so you can get market data fractionally faster and act faster on that data than other players.”

Chuck Schumer, a New York Senator who applied political pressure to Ms Schapiro, said he feared what iniquities might have grown up during the explosion of new trading methods. “It is also important to make sure flash orders aren’t just the tip of an iceberg lurking in the dark reaches of the market,” he said. “There is a lot of mystery about what goes on in dark pools and in the realm of high-frequency trading generally.”

He’s right about that. Wall Street, always awash with conspiracy theories, loves to gossip about the conflicts of interest that big firms have in running both their own proprietary trading programs and some of the biggest “dark pools” where customer orders are routed.

But high-frequency trading and dark pools are quite different manifestations of the emergence of computer-based trading. Computer algorithms have long been used to work out how best to implement a customer’s buy or sell order, perhaps by breaking it up into smaller parts so as to mask the demand and prevent it moving the market. Increasingly, the computers are showing that it is best to trade in anonymous “dark pools”, which, unlike exchanges, don’t have to show orders coming in and only have to report trades after they are done.

The SEC ? and the Financial Services Authority in the UK, where both high-frequency trading and dark pools are so far less well established ? are investigating what impact these dramatic shifts are having on the equity markets. In the US, they have undoubtedly brought trading costs down overall, as the spread between the cost of buying and selling a share has narrowed, but many players believe it has come at the expensive of market stability and has opened up much potential for abuse. Worst of all, if all trading is done in the dark, splintered between numerous trading platforms, how will anyone know the price of anything?

Announcing the SEC’s investigation in June, Ms Schapiro pointed out that other irony: “If dark pool volume were to continue to expand indefinitely, their success could threaten the very price discovery function on which their existence depends.”

View full article here

VN:F [1.9.22_1171]
Rating: 0.0/10 (0 votes cast)

Tags: , , , , , , , , , , , , , , , , , , ,

Ezine Article Board


This author has published 5774 articles so far.

Comments are closed