High-frequency trading can be defined as the use of quantitative-driven strategies to execute trades with a holding time of less than one day. In other words, firms are trading in a rapid manner, while cutting holding fees and subsequently increasing profit margins. High-frequency trading is algorithm-based and originated in the equities markets, but is currently extending its reach to other markets, including futures and commodities.
High-frequency trading occurs when traders are able to quickly identify and execute pre-developed strategies across markets and move from position to position in milliseconds. Traders are able to develop complex algorithms that look at numerous variables to determine exactly when to buy or sell specific stocks. Furthermore, some high-frequency trading shops are able to employ a strategy that relies on the ability to see certain orders a few milliseconds before the rest of the market—these are known as flash orders. Firms use these flash orders to create micro-markets around a specific stock across exchanges.
Current trends show that as many as 80 percent of orders are canceled at the last minute, which indicates firms are using high-frequency trading to determine at what price other buyers will be willing to bid on an order. For example, if an organization has 200,000 shares to sell, and it wants to determine the optimal selling point, it might first offer 5,000 shares at a certain price. If it sees that there are no immediate bidders, it can drop the bid instantly, and continue to do so in a matter of milliseconds until it determines the highest price buyers will be willing to bid on. Likewise, traders also have the ability to increase their bids instantly if they determine buyers will be willing to bid high enough. It has become a controversial practice in the financial services industry—and accounts for approximately 50 percent of all US equity trading—but some speculate that high-frequency trading is still largely misunderstood.
The Major Players
There is a common perception that high-frequency trading is reserved for the biggest and best firms—those that have virtually unlimited capital and resources. This assumption, however, is largely untrue. In fact, approximately 10 percent of firms that use high-frequency trading are small entities with fewer than five people, according to BNY ConvergEx Group.
Another common assumption is that firms require a great amount of capital to execute high-frequency trading strategies, when in fact, that is not necessarily the case. Yes, there are initial capital requirements, including technology, but infrastructure and application costs will decrease over time. Additionally, co-location offers firms a low-cost alternative to hosting and maintaining their own infrastructure. Data centers such as Equinix house many firms that use high-frequency trading.
Like its size, a firm’s location often has no bearing on its ability to execute successful trading strategies, including high-frequency trading. However, firms may find it advantageous to maintain relative proximity to the exchanges they frequent the most. While firms can realistically trade any market or exchange, they may not be able to do so as effectively after surpassing a certain distance.
Most importantly, firms should ensure that their connectivity to their datacenter is secure, particularly if there is great distance between the co-location site and their office space. Trends show that the majority of high-frequency trading firms are located in New York, Chicago and London at the moment, while other regions continue to grow, particularly in areas like Singapore, Sydney, South America and South Africa.
The Importance of Strategy
When it comes to high-frequency trading, the most important element for a firm is its strategy. Because the practice of high-frequency trading is based upon the speed of execution, strategies are constantly changing and evolving. Firms should, ideally, be changing their strategies every few weeks in order to stay ahead of their competition.
Thinking outside the box and tapping into new markets can be a way for firms to differentiate themselves. A firm’s strategy will also dictate the specific infrastructure and connectivity that it requires. Each strategy has different latency requirements, which could force the firm to host its infrastructure within the exchange itself; alternatively, it may only need to be connected to the exchange. Additionally, the strategy will also provide some information on the potential return of the fund through the back-testing of the strategy. This will provide investors with some comfort and fund managers with an understanding of the reason behind the capital investment for the infrastructure.
The practice of pre-trade risk management technically exists in high-frequency trading circles, but it has not come close to being fully adopted. An effective risk management system should monitor a trader’s activity in real time, and be able to identify when a trade exceeds risk boundaries that have been preset by the firm. Pre-trade risk management is affordable and easily available to high-frequency trading firms. So why aren’t they using it? Because it takes time. Traders have only milliseconds to execute their strategies, but a risk management system can often add latency when it computes the firm’s risk profile. Regardless of time, however, traders may be forced to invest in risk controls if investors continue to push for greater transparency across the board. It might not make for the fastest trades, but risk management seems like a smart move when it comes to business protection. Also, as pre-trade risk management systems become more prevalent, or even required, everyone will be facing the same increased latency which will level the playing field.
Regulations: What the Future Holds
Regulatory bodies like the Securities and Exchange Commission (SEC) have taken steps to better regulate the financial industry in the wake of the country’s economic meltdown and the Flash Crash, and high-frequency trading has been one of the targets. The SEC has taken actions to investigate flash orders, which are often a part of high-frequency trading. More recently, the SEC has proposed a ban on sponsored access, a process through which brokers give their level of access to high-speed traders, allowing them to buy and sell stocks without identifying themselves. The SEC has stated its fear that such anonymous trading threatens market stability. But high-frequency trading firms that use sponsored access maintain that it adds liquidity to the market. The SEC is also worried that with sponsored access, a firm that does not perform pre-trade risk management would be able to create an order that potentially bankrupts the company and severely disrupts the market—the reason why high-frequency trading was originally blamed for the May 6 upheaval.
The ultimate question remains: should high-frequency trading be regulated? A “yes” or “no” seems premature at this point, simply because the SEC should take its time in understanding all facets of high-frequency trading and focus not only on the potential negative impacts, but also on the positives that result from firms using it. Until there is a clear definition of high-frequency trading, the SEC cannot properly regulate and impose restrictions on it.
What’s next? On the regulatory front, it’s hard to know what the SEC will choose to do next. But for traders still employing high-frequency strategies to get ahead, innovation is key. As processing power continues to rise, and markets continue to gain liquidity, traders must invent new algorithms that can generate revenue quickly and efficiently in an ever changing marketplace.
Matt Bretan is director of strategic consulting at Eze Castle Integration, a Boston-based provider of technology and consulting services to hedge funds and investment management firms.
Adam Sussman joins Anthony Malakian to talk about Liquidnet's acquisition of OTAS, machine learning and AI, and what the buy side wants from analytics platforms.Subscribe to Weekly Wrap emails