High Frequency Trading

What is High Frequency Trading?

High frequency trading (HFT) programs execute sophisticated intuitive algorithms that generate rapid-fire trades at blinding speeds across multiple markets and securities for purposes including market making, arbitrage and implementation of proprietary trading strategies. HFT programs extract extremely thin profits from massive volumes of transactions numbering in the millions on any given day. HFT trading accounted for nearly half of all the volume traded on U.S. stock markets in 2012. Since HFT trading programs rarely hold positions for extended periods of time, they maintain a Sharpe ratio north of 10. From the flash crash of 2010 to the Knight Capital Group implosion in August 2012, they have generated much controversy globally for generating too much unnecessary volatility and the ramifications when they go awry. Saturation, competition and looming regulation has peaked out the industry since then. To curtail HFT programs, Italy subjected a 0.02% tax on equity round trip transactions lasting under 0.05 seconds in September 2013.

How HFT and Algorithms Work

High frequency trading requires the lowest latency possible to maintain a speed advantage over the competition including retail traders. Sophisticated algorithms are at the heart of these programs. The algorithms are the instructions for reacting to market conditions based on highly intuitive signals. The complicated coding is the DNA of the programs, which can consist of millions of lines of code. Some of the largest HFT firms have claimed profit periods consisting of over 1,000 consecutive trading days without a single losing day. The advantage of speed, access, capital and virtually no holding time make it one of the most risk adverse forms of trading, when everything functions properly.

Latency

Latency refers to the amount of time it takes for data to reach its endpoints including market quote data from the exchanges to the trade fills. Low latency indicates faster speeds. Firms invest top dollar for the most powerful hardware and infrastructure to gain a slight edge on competition in terms of speed. During the speed wars of HFT in 2012, firms were trying to gain an edge on the speed of data by running a straight cable connecting Chicago to New Jersey that would reduce roundtrip latency to 12.8 milliseconds from 13.1 milliseconds compared to the average roundtrip time of 14.5 millions. Since then, other HFT firms have adopted air-based transmission using microwaves cutting the roundtrip latency times as low as 8.5 milliseconds. Speed arbitrage is considered a sophomoric strategy for HFT trading in terms of sophistication as competitors aggressively seek other forms of an edge.

Algorithms

Simply put, algorithms are sets of instructions to be executed when various conditions are triggered. Take that and extrapolate it over millions of lines of code to construct the complex DNA of a high frequency trading program. Algorithms integrate complex quantitative pricing, execution and portfolio models to implement trading and management strategies. High frequency trading algorithms are aptly named due to the low latency aspect of executing them. However, algorithms are becoming more commonplace without the low latency requirement. Even retail traders are getting in on the game utilizing routing algorithms embedded directly into trading platforms. Retail traders are able automate their strategies with a growing number of third-party services offering algorithm leasing and programming services. Traders should be prudent in researching these products before taking the plunge. Automated trading programs are becoming a crowded field where over-hyped claims and fraud can be very prevalent, so buyers beware.

How Algorithms Affect the Market and Traders

For intraday traders, high frequency trading programs are a double-edged sword. Advocates argue that HFT programs help provide more liquidity to the markets, but intraday traders attest the opposite holds true. They argue that HFTs actually shrink liquidity as their speed allows them to front-run orders regularly to skim profits, at the expense of the trading counter parties.

Traders have noticed how quickly the bids and asks can disperse when trying to sell or buy a larger sized position (IE: 5,000 shares) of a stock. This is a sign of thin liquidity, which is common for thinly traded stocks. However, this phenomenon happens regularly on widely traded stocks.

On the flipside, the rapid price movements are a boon to traders positioned on the right side of the move as HFTs leapfrog each other to capture liquidity thereby spiking prices higher or lower much faster than normal. This is why a stock can generate several hours worth of price movement in just a matter of minutes during the first hour of trading. This type of volatility is a Godsend for momentum traders. However, the activity to literally switch off in a heartbeat, which can also trap traders with little liquidity to exit positions without market impact. Market impact is the keyword here. High frequency trading programs magnify the market impact of trades, which institutions try to avoid, while traders try to exploit.