The debate over high frequency and algorithmic trading has raged since the concept was born. To its detractors, HFT traders are parasites feeding off exchange rebates whose activity threatens that stability of markets. On the other side of the coin, its advocates cite HFT trading as a trading strategy, equalled by no other in its ability to pump liquidity into markets, reduce spreads and curb volatility.

High frequency trading began its ascent upon the introduction of regulation by the US Securities and Exchange Commission for alternative trading systems, including electronic exchanges, in 1998. Its subsequent growth has been astonishing.

At the beginning of the millennium, HFT accounted for less than 10% of all equity trades in the US, a number that now stands between 60% to 70%, depending on the source. On US futures exchanges, HFT now accounts for up to a third of all trading. In Europe, HFT accounts for between 30% to 40% of equities and futures trading volume, while in Asia, the figure is about 5% to 10% of equity volume.

Since its inception, HFT has steadily increased its sphere of influence, from one market to the next, and one region to the next. It has changed the face of trading and attracted its fair share of criticism in its wake. However, its rise had stayed off the radar of most regulators until one fateful day in May 2010.

The fall guy

That day was the Flash Crash on May 6, 2010 when the Dow Jones Industrial Average fell nearly 1,000 points in less than 30 minutes, its largest intraday point loss in history. Since then, regulators have been taking a long and hard look what risks might be posed by high frequency trading, and how they can be contained.

While HFT was immediately blamed for the Flash Clash, subsequent reports have found it was not the root cause. The Flash Crash was in fact the product of a rapidly executed sell order of E-Mini Standard & Poor’s 500 futures contracts on the Chicago Mercantile Exchange, via an order originating from a fundamental trader, not a HFT firm.

The sale was conducted through an automated execution algorithm that unloaded the contracts in just over 20 minutes. The price decline that resulted, however, was then exacerbated when HFT firms resold contracts they purchased during the frenetic period – sucking away liquidity at an important juncture. So while evidence suggests that HFT firms were not the genesis cause of the crash, the trading strategy has nevertheless acquired an air of guilt by association.

Regulatory responses

Since this event, regulators have grappled with a response. In June and then again in September 2010, single stock circuit breakers were put in place to operate in the S&P 500 Index, the Russell 1000 universe and certain actively traded ETFs. Since then, if a stock’s price changed by 10%, trading has been halted for a period of five minutes.

In the joint report released in February 2011 by the CFTC-SEC Advisory Committee on Emerging Regulatory Issues, a number of recommendations were made, which included expanding the current single stock circuit breaker, as well as imposing market-wide ‘limit up/limit down’ features on trading, narrowing the band of prices able to be traded by rapidly declining stocks, and requiring futures exchanges to impose an additional tier of pre-trade risk safeguards against volatility.

Then in April 2011, the SEC, proposed replacing circuit breakers with the limit up/limit down rule, which effectively would create a 15-second moving-window price collar.

In speeches given by CFTC Commissioner Bart Chilton, he has indicated that, given the Flash Crash, limits to HFT transactions are ‘appropriate to consider’. But even if limits are broadly accepted, more specific questions still remain. Such as if there are position limits, say 10% of open interest in a market, should high frequency traders be allowed to trade 10% repeatedly over a short period of time?

Probably even more perplexing, how can regulators incentivise HFT firms to stay in the market during times of extreme volatility? Currently, high frequency market makers, although paid by stock exchanges to supply volume, have no obligation to do so. When their risk controls activate an alert, they pull quotes and sell their portfolios.

While the CFTC-SEC Advisory Committee suggested that something should be done to address this issue, no specific recommendations were forthcoming. So, while the first round of measures to address perceived inadequacies in the market system focused on trading pauses and the eliminating of unfiltered access, it is anticipated that when the final mandates are revealed in the US and Europe, they will include liquidity incentives or obligations for HFT firms.

Too expensive to regulate

Separately, the CFTC Technology Advisory Committee said its plans aimed to keep the benefits of electronic trading, but it was also seeking to check new risks that had been created as a result. The five page report highlighted three areas of the electronic trading chain which could be subject to new requirements: trading firms, clearing firms and exchanges.

However, instead of proposing a series of new rules which the CFTC would be responsible for regulating, the committee proposed using a series of checks by the three elements of the trading chain to ensure high frequency trading did not harm the market.

The committee said the problems had much to do with the sheer number of trading firms in the market, each with vastly different systems. The committee described the trading software as “complex”, while trading algorithms are sensitive intellectual property. Technology platforms also range widely, as does network hardware.

Given such challenges, the CFTC concluded: “The only way to independently enforce any sort of specific regulations on quality assurance for trading firms would be to have a virtual army of CFTC-employed quality assurance professionals who have complete access to all trading firms' intellectual property at all times.”

The report admitted that the ability to enforce new regulations would be too costly so instead recommended the trading firms be required to demonstrate to the exchange the existence of “reasonable measures” in their processes and systems before being approved to trade.

In Europe, some authorities have endorsed the idea of rest periods, which would require HFT orders to rest on the book for a minimum period of time, or to limit the ratio of orders to transactions executed by a given participant.

While within the European Union, various opinions exist – for example, the UK’s Financial Services Authority has stated that it does not believe measures such as a rest period are necessary - the next step will likely involve some politicking at the EU level between various governments and authorities, before new legislation is handed down in the latter part of 2012.

Seeking new markets

Whether you believe HFT firms had an unholy hand in the Flash Crash, or present an unnecessary risk to the financial system, it is certain that they are here to stay – the genie cannot be put back in the bottle. As US and European markets become more efficient, in part as a result of the increasing presence of HFT, the new breed of firms are spreading their wings and organising themselves in order to take advantage of anomalies in the markets.

Victor Lebreton, an independent director at Paris-based Quant Hedge, runs approximately $200m a day in HFT strategies. “HFT is really people getting ultra-organised in order to take advantage of arbitrage opportunities. So when markets are less organised, whatever and wherever they are, HFT strategies are primed to take off,” says Lebreton.

Ryan Terpstra, the chief executive officer and founder of New Jersey-based Selerity, which provides HFT firms with event data from real time sources that they are able to incorporate into their investment models, says a lot of US HFT firms are looking very seriously at establishing significant operations in Asia. “There is a lot of interest in implementing high frequency trading capabilities in foreign exchange, equities and futures markets, particularly in India and Singapore,” says Terpstra.

One issue, however, is that, while SGX recently announced it would be opening up to colocation as part of a move to encourage HFT, most Asian venues often do not allow colocation. At the other end of the colocation spectrum is Brazil.

“The poster child for colocation and futures related to high frequency trading strategies is Brazil, where many firms have already colocated. While there has already been a lot of growth in Brazil, we are still going to see much more,” says Paul Zubulake, a senior analyst at Aite Group.

The rise of alternative trading venues was one of the precursors to the increase in HFT volumes in the US. So too, regions that are today embracing alternative trading venues are seen as futures hubs for increased HFT volumes.

Hirander Misra of London-based Algo Technologies points to the move by Australian regulators to allow Chi-X Australia in May 2011 and the establishment of Chi-East in Singapore as actions that will likely attract an increased number of HFT firms. “The market is changing with the establishment of dark pool trading venues, especially in Asia where a lot of banks are now setting up operations to pump up electronic trading expertise,” says Misra. Algo Technologies is a specialist technology firm which provides advanced high speed trading solutions for firms and markets.

Mitigating risk

In the US and Europe, the emphasis is on providing trading venues that provide safeguards against events such as the Flash Crash. “We employ effective risk-, volatility- and error-mitigation functionality in place to support high frequency trading activity in a way that benefits all market participants. High-frequency traders serve as important liquidity providers to the market, which ultimately reduces execution costs for all participants, including individual investors,” a press representative from the Chicago Mercantile Exchange told FOW.

At Eurex, the aim is to implement risk limiting checks and balances that don’t impact latency such as maximum order size limits as well as balancing the equilibrium between HFT and those traders demanding longer term exposure, says Wolfgang Eholzer, who oversees the exchange’s trading systems design. Currently, 25% to 40% of the volume at Eurex is related to HFT market models.

“On the technology side, we have specific central interfaces connected to our trading platform that are good for those interested in speed, like HFT and market makers. But we also have a set of interfaces that are particularly suited to those who don’t care about latency and are interested in reducing IT costs, like some fund managers. So we are trying to offer different functionality for the different user groups,” says Eholzer.

Markets to watch

While HFT first made its mark in the equities and futures markets, it is now evolving into other areas such as commodities futures, FX and energy. Energy, which was still traded in the pit less than five years ago has grown incredibly in the area of HFT, says Zubulake at Aite Group.

“Energy is the number one area seeing a big increase in high frequency trading. It is liquid, there are different months to trade, and there are arbitrage opportunities between the gasoline contracts and crude contracts. So there are a lot of possibilities in terms of spread relationships,” says Zubulake.

Another area to watch is FX, where approximately 70% of the market is expected to trade electronically by the end of 2012. The introduction of EBS Prime and Ai into the non-bank community triggered the floodgate to high frequency trading in FX, according to Aite Group’s report entitled ‘High Frequency Trading in FX: Open for Business’, released in April 2010.

At the end of 2009, HFT accounted for approximately 25% of overall FX trading volume, a figure expected to hit more than 40% by the end of 2012. Much of this growth will come as a result of an influx of next-generation equity and futures HFT firms attempt to capture uncorrelated alpha in FX, says Zubulake.

Erik Lehtis, president of Chicago-based Dynamic FX Consulting, helps firms wanting to set up trading in FX using HFT algorithms. “There are a lot of firms out there that are in FX but do not feel they are having the best results,” says Lehtis.

Some of the biggest pitfalls surrounding HFT in FX are founded on the belief that many believe the same algorithms used in other markets can be applied, which is simply not the case, says Lehtis. “It is a fragmented market as far as liquidity is concerned. Each of the exchanges and the market data has its own quirks. Today, the best people have a really good understanding of FX itself, not just how to apply algorithms,” he says.

One of the next areas HFT will be applied is fixed income, according to Brian Schwieger, head of EMEA algorithmic execution at Bank of America Merrill Lynch. “High frequency trading has progressed from equities to futures to commodity futures and now to FX, but we believe that fixed income will be next.”

Going forward, the challenge will be to make sure regulation does not blanket all markets with the same requirements, as if that were to happen, markets like fixed income might be left out in the cold. “Fixed income has always had a very different market structure. It is an individually priced and a manually oriented trading environment. As a result, it is important that any legislation does not assume markets such as fixed income have the same characteristics as say FX,” says Schwieger.

Whether HFT moves into new regions, new markets or ever faster trading technology, it seems certain that the regulators will seek to try to shape it. Just how they will seek to do so, still remains to be fully seen. As new regions adopt alterative trading venues that are open to new and less traditional trading technologies, we are likely to see the sphere of influence of HFT grow.