Light-Speed Trading

How Algorithms are Turning the Investment Industry on its Head

Special IssueColin Harris — Published December 3, 2013

Using algorithms and powerful servers to make the fastest trade decisions possible, high-frequency trading has become an essential tool for the power-player investment firms—with Forbes reporting in September that 50 to 70 per cent of trades made in world markets are HFTs.

It’s changing the investment business, but the jury’s still out on whether it’s for better or for worse.

The proliferation of computer-based trading doesn’t come as a surprise. It’s existed in some form or another for more than a decade, and follows a trend seen in many industries—computers taking jobs once held by people.

What regulators are now faced with is the question of how fast is too fast—whether there is a point where this rapid trading can cause real damage to the financial system.

HFT encompasses not just one strategy, but rather the usage of a variety of technological approaches to make quick trades before a human investor can even process the data.

This includes using algorithms to make snap sales based on new information less than a second after markets open, and buying stock for a penny more than other bidders and then quickly selling it again for a higher value that is still less than the seller’s asking price.

HFTs can even automatically bid a slightly higher price than human traders instantaneously after discovering the competition.

HFTs also means physical changes to trading, like having private servers on the market floor for the lowest possible latency when transferring data.

The use of HFTs has declined since its peak in 2009, after the New York Times reported in the previous year that HFTs had earned about $21 billion—before anyone outside of Wall Street knew even what they were. Since then, some investors have gotten wise to some of these HFT strategies.

But it really caught the attention of regulators after the 2010 Flash Crash, when the dropped by nearly 10 per cent and recovered in mere minutes. Once the algorithms noticed a drop due to bad human trading, they ceased activity—and their absence created a much larger dip in the index.

The Investment Inventory Regulatory Organization of Canada will be releasing a report on the impact of HFTs next year.

HFTs have become the majority of the trades made in markets, but make up an even higher amount of the bids offered. Critics say this creates the illusion of high confidence in a stock, and increases the cost of trading due to the increased volume of messages transferred—between 90 and 95 of all quotes come from HFTs according to Forbes.

Proponents of HFTs argue that these tools reduce the difference between what the seller is asking for a stock and what bidders are offering (known as the “bid/ask spread”), which lowers the fees that go to traders to find a middle ground and thus indirectly means savings for the investors.

Jeffrey G. MacIntosh argues in the Financial Post that this offsets any increased costs of HFTs caused by flooding markets with stock quotes.

After the economic crash of 2008, HFTs were a way for big banks to start making large profits again. But it raises questions of the balance between fairness and efficiency, with large banks and investment firms able to pay for slightly faster information about the state of American markets, and paying a premium to have their servers share floorspace with market servers.

In this technological arms race, regulators are constantly playing catch-up. In the high-tech industry, innovation moves at light speed. And considering they’re only shy by a few milliseconds, soon the trades themselves may, too.