High Frequency Trading is a term used to describe a group of trading methods that incorporate high levels of automation into the buying and selling of securities where large volume trades are executed in seconds or fractions of seconds.
One way to think of high frequency trading is to think about the difference between writing a letter and sending an instant message. When you write a letter, you try to cover a lot of bases by conveying as much detail and meaning as you can because what’s happening at the time you write the letter could easily be different once your friend receives it. To reduce the effects of this time delay, you involve your computer or your smartphone and send an instant message. Tweeting or texting require a lot less when it comes to word choice because you can send thousands of messages in sequence and in the end, your friend will know what you’re talking about. High frequency trading strategies attempt to do something similar for securities traders, as computers are used to make trades at specific points based on conditions that are predetermined by the trader.
So We’re Letting Computers Handle Our Money?
In a way, yes. Most transactions that are processed today are handled electronically, simply because computers can more efficiently process large amounts of data at speed. High frequency traders use complex computer models to analyze a vast array of market conditions in real-time and make trades that can occur in seconds or between seconds depending on what a given trader is looking for. So in this sense, computers are the processing agents for the buy and sell orders but human traders are supplying the when, where and how much.
What Does High Frequency Trading Look Like?
High Frequency trading is generally defined by the following characteristics:
- uses computer software to analyze incoming market data and conduct trades based on this data.
- trades remain open for very short periods, usually from hours down to fractions of seconds. Trading occurs very rapidly often trading thousands to tens of thousands of trades each trading day
- High frequency traders usually do not have any positions open at the end of a given trading day.
- Relies heavily on the processing capability of computers to find, analyze relevant market data.
There are also different types of high frequency trading:
- Market Making: places buy or sell orders on various securities in order to earn it’s bid-ask spread, thus creating a new market by acting as a counterparty to incoming market orders.
- Ticker tape trading: uses fast processing computers to analyze ticker quotes and volumes to glean information that has yet to be reported.
- Event arbitrage: using recurring events and past price history to forecast short-term market moves and profit accordingly
- Statistical arbitrage: an HFT strategy that attempts to take advantage of predictable deviatiions in “stable statistical relationships” between securities.
Why Trade At Such High Speeds?
Let’s say that you wanted to buy 300 shares of Apple stock at its last sale price of $628.71 per share, that’s a $188,613.00 investment. To turn a profit you, the investor, want to wait until the share price rises to say $635.00 that way you can sell your 300 shares back for $190,500.00 or even more. While investing and charting techniques could give you enough insight to determine that at some point, Apple shares might reach or exceed that $635.00 level, there’s no way to know exactly what day that might happen. Once your buy order is submitted, your money is active on the exchange and at risk until your shares increase in value to the level you desire.
High frequency trading strategies attempt to minimize the amount of time money is active in the market by buying and selling securities over fractions of seconds. Reducing the time your money is in the open market decreases its overall exposure to market forces since it literally is there and gone again in too short a time for market dynamisms to significantly impact the investment positions in your portfolio. Because the transaction times are so short, high frequency traders look for very small returns in their positions, usually pennies on the dollar or fractions of pennies (or whatever currency their position trades in). The idea is that a large number of small gains can result in a net profit at the end of each trading day.
What Are The Risks Posed By High Frequency Trading?
With any system operated by machines or with high levels of automation, the most prominent risk is for software malfunction, corruption, or manipulation and high frequency trading systems are no exception. The computers used for this type of trading are subject to human errors in software design or mistakes in execution on the parts of the traders, and these errors can create serious complications for major stock indexes. According to The New York Times, high frequency trades make up about 60 percent of the near 7 billion shares exchanged on U.S. stock indexes (about 4.2 billion shares) every trading day so glitches in these systems can move markets dramatically.
On May 6, 2010 traders at Waddell & Reed Financial, a mutual fund, ran a program intending to sell $4.1 billion worth of futures contracts. The algorithm sold over 75,000 contracts back onto the market over some twenty minutes, increasing its sales as prices fell. The sales from Waddell & Reed prompted other high frequency traders to sell their holdings of “E-Mini S&P 500 futures” just as aggressively which led Waddell’s algortihm to accelerate its selling in response. The various computer systems of high frequency traders exchanged these contracts 27,000 times in 14 seconds (that’s over 1,900 times a second) though only 200 were actually being bought and sold. The selling activity spilled over into the stock markets when the investors began buying up the futures contracts at the cheaper prices. Automated trading systems in the stock markets shut down after noticing the uptick in buying and selling, and this led to extreme disruptions in the share prices on major indexes. Some companies saw their stock price drop to pennies, while other companies traded to highs of $100,000.00/share. Volatility in the markets continued until computers on the futures exchange intervened and paused trading for 5 seconds to normalize the market. As a result of the electronic mishap, the Dow Jones Industrial Average lost 600 points over the course of about five or ten minutes.
The Knight Capital Group is a financial services firm that specializes in elecrtonic market making and computerized trading and on August 2, 2012 the company lost $440 million dollars thanks to a trading error from one of its automated systems. Computers at Knight executed sales of the companies stock that it purchased during the previous day of trading resulting in losses that far exceeded its available capital and revenues from the previous quarter. Knight Capital shares lost 95 percent of their value in two days of trading and the stock closed at $0.258 cents a share that Thursday. The issue was caused by new software that had just been implemented, and the errors not only sent Knight Capital into insolvency, but caused unnatural fluctuations in share prices across the financial sector.
Anything Else We Should Know?
In light of incidents like the Flash Crash and Knight Capital, regulatory authorities in Europe and here in the U.S. are proposing that new rules be added to monitor or even ban outright the use of computerized trading techniques. If such bans are adopted, particularly in U.S. markets, the effects of less computerization could harm the individual investor in the short term, but improve market confidence in the sustainability and resilience in major global exchanges.