High-Frequency Trading, Explained

High-frequency trading is a technique that involves specialized software and algorithms, high-end computers, low-latency internet access and up-to-the-moment market data in order to outpace all of the competition and allow for unique strategies not otherwise possible. 

High-frequency trading, or HFT, is a system in which algorithms and software make multiple trades per second and which offers a slew of benefits not available to regular traders. It has existed long before cryptocurrency and is believed to make up as much as 80% of the volume in certain asset markets. Now, however, it is becoming a major factor in the world of decentralized assets as more and more institutional investors begin to take notice.

The principle behind most HFT strategies requires that those executing them are among the first to do so — and sometimes, all it takes is a fraction of a second to turn a profitable move into an unprofitable one. This has evolved into some specific techniques that this type of traders use to stay slightly ahead of the competition. This can certainly involve the latest and most powerful equipment, but it also requires the development of very clever ways of working inside the slight variables present throughout the market and exploiting them in millisecond turnarounds. 

Techniques and strategies can include colocation, market-making, arbitrage, pinging and news-based trading. Each option comes with costs and benefits, and not all strategies are always available to traders, depending on market conditions. Of course, there is still some debate as to whether it is ethical to create an uneven playing field. To be sure, some of these strategies are illegal or highly regulated inside traditional markets. It is true that cryptocurrency is still in a bit of a “Wild West” phase, but this is changing quickly and traders should be aware of their local laws before engaging in this type of activity.

Colocation refers to the practice of placing a trading server as physically close to an exchange’s data center as possible — sometimes in the same facility — to get the minimum possible delay in the transmission of market data.

On regular exchanges, every user is susceptible to risks associated with communication delays. For most retail investors, these slight delays in processing won’t affect much; for the high-frequency trader, however, milliseconds can be everything. One obvious way to have an edge is to use the best equipment, but another is to place your own trading server physically close to the data center that powers your chosen exchange. In some cases, users literally set up shop in the area directly surrounding a data center — often, though, these exchanges offer to house private servers on-site for those interested. Customers can even have a cross-connection directly to the main server, and those who implement this option circumvent having to connect over the internet, which significantly cuts transmission delays. This practice is called “colocation,” and it can easily shave enough time off of data transfer latency to give a meaningful edge to anyone employing it. 

In cryptocurrency, there are already some exchanges that are offering this type of connectivity. For example, exchanges such as HitBTC, Gemini and ErisX have had colocation options available to clients for over a year. This can allow HFT strategies to be utilized by firms not physically close enough to the data centers they need access to. Generally speaking, colocation is a major building block for performing this type of trading, as it will always give an edge over those who are farther away from the access point.

Market-making is a common strategy in the trading world. This is when a trader with sufficient resources places both bids and asks into the same market, which provides liquidity and effectively turns a profit based upon the spread.

Market-making in regular trading is usually provided by large firms and is generally seen as a positive practice to keep important markets sufficiently liquid. By providing ample orders on both sides of the order book, makers ensure other clients can always move funds. It is common for a major exchange to have contracts with one or more market makers, and the practice is generally seen as healthy for the entire system. There are already a small batch of firms popping up that are offering market-making services for cryptocurrency.

However HFT market makers are often smaller private firms that have no contracts with the exchange. They are again leveraging their improved performance to ensure that it is their bids and asks that are making the market. While this effectively performs the same service as major public firms, smaller market-making firms are less reliable, as they hold no business connection to the exchange.

Arbitrage is the act of taking advantage of a price difference on the same asset across multiple markets.

Occasionally, a single asset will have an inconsistency in its price on different exchanges. The act of identifying and exploiting these differences is a common source of profit among traders. These opportunities happen even more frequently in cryptocurrency — due to increased volatility — but are quickly eliminated by regular market forces, as all traders should have access to these opportunities. 

However, HFT practitioners can identify and take advantage of these situations first. By using software designed to spot these discrepancies in real time, they can then create orders in response, hundreds of times faster than a regular trader. There are multiple ways to employ the arbitrage strategy, but all involve spotting variability in a market and jumping on it first. It is fair to point out that the more efficient traders get at profiting off of arbitrage, the less profitable it actually becomes, on average.

Pinging is a process where HFT users will use a series of smaller orders made very quickly to hunt out larger orders that have been “segmented” — or broken up into small parts — so as not to affect the market price too much. 

This method essentially tests a range of prices, using small orders, to find out the high and low range that a big mover is trying to sell for. Because the means of detecting these orders are encoded into algorithms, they can hunt for these opportunities several times a second. The software can then use its increased speed to purchase the asset immediately on the low end of that range and sell it to the “prey” at the high end, netting an instant profit. 

Admittedly, this technique mostly feeds on big movers and is quite often used in places known as “dark pools.” Dark pools are either private exchanges or forums that don’t report their order book in real time. Regulations usually demand that transaction info be released, but it can be delayed long enough that big institutional users are able to make large trades without immediately affecting the market. However, these systems are primed to be victimized by high-frequency traders who can use the “pinging” technique to sniff out and trade against other dark pool users. This can cut into the profits of the traders who aren’t high frequency and undermine the benefits of using a dark pool.

News-based trading is simply the act of buying or selling in response to news releases concerning market assets.

Most traders have used the news at one point or another to inform their trading strategy — it’s almost impossible not to. For most, the news is a reasonably level playing field because the general public should be gaining access to this information at about the same time. Using information that has not been publicly released is considered “insider-trading” and is generally illegal, though still rampant.

The way that HFT gets ahead here is that it can utilize modern software to analyze news stories in the seconds after they break and then instantly begin placing orders in response. The software is smart enough to determine not only which asset is being discussed but whether the news is positive or negative. While clearly not insider trading, this could still give a massive edge over any trader who has to process the story and decide how to respond by themselves.

Some propose that the practice gives a clear, unfair advantage to traders who have the means to enact it. However, supporters feel that it helps provide liquidity and stability to the cryptocurrency market.

As has been laid out, high-frequency trading definitely unevens the playing field against anyone not involved in it. A traditional trader could never hope to succeed in the same strategies, as they could never match the speed of a high-frequency trader. This effectively puts the vast majority of retail investors on a lower tier, which doesn’t sit right with many. While the practice is already common in legacy finance, it stands to have an even bigger impact on the future of cryptocurrency, with its 24/7 trading and loose regulations (for now). Those who see decentralized assets as a means to reimagine money feel this is another form of elitism that is contaminating a free market.

One other risk to markets is a possible increase in volatility and exposure to flash crashes. Basically, the algorithmic nature and raw speed of HFT systems mean that certain market conditions can occasionally cause a major sell-off in just seconds. This adds a heightened element of risk into the market, which can hurt unwary traders.

Proponents take another approach, though. They make the case that since HFT acts to add higher liquidity on exchanges where it is used, regular traders are more likely to find matching orders and move their money quickly. Also, the efficiency of price discovery improves, which should tighten the spread across bids and asks — and in fact decreases overall arbitrage opportunities. The case could also be made that it is just this increase in efficiency and liquidity that should decrease the aforementioned risk of flash crashes. 

Lastly, while those who don’t use high-frequency trading could never beat high-frequency traders at their own strategies, there are other options available to traders who don’t rely on speed. It can be argued that these other techniques actually benefit from the relative stability that this practice brings in.

While still in its infancy, a few traditional HFT firms have begun offering digital assets to their clients. Even more, at least one cryptocurrency exchange is giving users the tools they need to get started.

In recent years, multiple traditional HFT firms have begun to get involved in decentralized assets. The biggest one is Cumberland Mining, a subsidiary of Chicago-based firm DRW. Other institutional companies getting involved include Jump Trading, DV Trading and Hehmeyer Trading. That being said, currently, there are not too many cryptocurrency exchanges that have the speed or tools built in to allow for effective HFT. There are a few options users can turn to, however. As mentioned, exchanges like Gemini and ErisX offer colocation, but others such as Coinbase and HitBTC provide colocation as well as specific tools for high-frequency traders. With APIs such as FIX and Streaming, these sites ensure that the speed of trade execution depends only on network delays, which is generally less than 1 millisecond.

By combining powerful software with on-site hosting, users potentially have everything they need to experiment with the general strategies outlined here. It also helps that some of these exchanges offer a demo trading mode, where users can set up and run their experiments with the same real-market data but without risking actual coins. Considering that most of the other options available are traditional financial firms, using a true digital asset exchange may be more appealing to cryptocurrency enthusiasts.

While high-frequency trading may not be for everyone, it can definitely offer new techniques that tech-savvy and adventurous traders may enjoy. It’s true that knowing and keeping up to date with regulations will be essential as the cryptocurrency space evolves. However, it is highly unlikely that HFT as a whole is going away anytime soon. There’s a lot more to learn, but with the right tools and techniques, anyone has the chance to employ this method.