High-frequency trading has come under attack for recent stock market crashes and for threatening the long-term integrity of the market. Nicholas Pratt listens to the case for the defence
“A high-tech, high-speed poker game is playing out in the stock market, and billions of dollars are at stake. The adversaries are high-frequency traders and big investors such as mutual funds.”
The impression given by many such articles is that the integrity of the markets is under threat because sensible long-term investors are being unfairly outdone by technologically tooled-up predators that are able to second guess their trading intentions. These same aggressive and speculative short-term opportunists are also creating unwanted volatility in the market and when their automated algorithms run out of control, the markets go into a tailspin. Action has to be taken to protect both the stability and integrity of markets and to instil some sense of fair play among investors, it seems.
The complaints reached a crescendo following the ‘flash crash’ incident on 6 May when US stock markets briefly plunged into freefall. Nearly 21,000 trades were cancelled on various exchanges because they were deemed erroneous and it caused widespread panic among investors and traders and a 1,000 point drop in the Dow Jones industrial average, the largest single-day decline in Dow Jones history. Thirty minutes later, however, the market corrected itself and a crisis was averted.
Pointing the finger
A number of subsequent inquiries into the flash crash have pointed the finger at high-frequency trading. In a statement to an SEC panel, Jeff Engelberg, a senior trader at US mutual fund Southeastern Asset Management which has $35bn (€22.7bn) of assets under management, complained about the advent of firms engaged in low latency, colocated, active trading strategies that require little or no fundamental knowledge of underlying securities.
As a result of this trend, when the flash crash occurred, many market participants simply did not know what businesses they fundamentally owned. This ignorance made a bad situation drastically worse, states Engelberg. “Our psychology has fallen victim to a notion that, without qualification, technology and speed should be embraced. Contrary to the claim that speed reduces risk, the introduction of excessive speed and unrestrained technology destabilised the markets and made them wholly indecipherable on 6 May.”
Engelberg’s statement goes on to draw a distinction between those long-term investors that rely on a “logical and defined fundamental research process” as opposed to trading “borne out of a short-term profit motive” and calls on the SEC to “consider which market participants best foster the capital formation process through resilient, stable equity markets and ensure markets are absolutely conducive to those strategies”.
Unsurprisingly, the trading firms, investment banks and technology providers that work in the algorithmic trading arena have come out in defence of high-frequency trading and its part in the flash crash. While high-frequency trading may have the potential to add to volatility in a flash crash scenario, it can also be a balancing force because algorithms essentially act on logic rather than instinct. They are designed to remove emotion from the trading process and it was this devotion to logic that was responsible for the rapid self-correction of the market, say some.
“I think if regulators were to examine the trading activity that happened on 6 May they would find there was a lot of fear there,” says Scott DePetris, chief operating officer at trading software vendor Portware. Much of this fear has been centred on high-frequency trading via the perception among some investors that high-frequency traders have access to information that the rest of the market is not privy to. “Whether this is true or not is almost not important – the perception is as important as the reality because it is what is driving investors’ behaviour. So it is important that we kill that perception.”
More transparency around trading activity would help significantly, says DePetris, so that investors and regulators can see exactly what activity there is. Consolidated tape would be a good start, as would the establishment of central data repositories into which brokers are obliged to report all trades of a certain size and the electronic tagging of particularly sizeable high-frequency traders.
“I think all of the negative connotations around high-frequency trading come from this lack of transparency – it is the fear of the unknown,” says DePetris. “If the market becomes more informed and educated about how high-frequency traders interact with order flow, it may help to kill these negative perceptions. Right now though I think it is a public relations problem more than anything.”
The small, independent firms that specialise in high-frequency trading are not typically known for devoting large amounts of their revenue to public relations and have a very different media profile to other investment houses, which has helped to make them an easy target for criticism, says Tony Nash, head of execution services at investment bank Execution Noble.
Nash believes the accusations that high-frequency traders have an unfair advantage are based on misconceptions. “Becoming a successful high-frequency trader costs money – amongst other factors you have to construct and maintain complex models and build a fast network so the barriers to entry can be significant. As for high-frequency trading being unfair, no-one is stopping anyone else getting in on it.”
In fact the benefits to the wider market that come from high-frequency trading are often forgotten, says Nash. “They provide a larger amount of liquidity then they are given credit for. These illiquid markets would be far worse without their presence.” High-frequency trading also lowers trading costs, tightens spreads and makes the markets more efficient, argue others.
There are also some misconceptions about the extent to which algorithmic technology is used for predatory, short-term trading. “There is an awful lot of attention on the competition to be the fastest, but for many traders, algorithms are used to minimise market impact more than they are used as an aggressive weapon,” says Giles Nelson, chief technology officer of Progress Software, a technology vendor specialising in the provision of algorithms.
The road to regulation
Nevertheless, it is inevitable that some regulatory action will be taken. The flash crash highlighted the need to put some measures in place to guard against the risk to the system should algorithms run amok, but it also highlighted the difficulty of enforcing changes at the execution venue level in today’s deregulated and fragmented securities market where the levels of technological sophistication vary markedly.
For example, the volatility triggered by the mysterious rogue trade on 6 May was exacerbated by the fact that some venues, like the New York Stock Exchange, had so-called ‘circuit breakers’ that automatically suspend trading when a certain level of volatility is reached, while others did not.
Consequently, for that 30-minute period the US equities market was in disarray and a number of algorithmic trading programs were working off an incomplete set of market data. The SEC is likely to introduce mandatory circuit-breakers for all venues but it is less clear what other steps can be taken, particularly in Europe where the Committee of European Securities Regulators (Cesr) announced in June that it intends to examine the area to “better understand any risks” it presents.
“Regulators need to give guidelines as to how to safeguard execution, but they cannot be too specific or prescriptive because each venue has a different way of doing things and the regulators cannot keep pace with the rate of development at these venues,” says Jim Downs, founder of US-based trading technology firm Connamara Systems. “You could bring in measures such as pre-trade risk, limiting order sizes and limiting losses, outlawing algos that sweep the market. But most firms will trade at multiple venues so it is questionable what effect single-venue measures will have.”
There is also the competition that exists between venues, thanks to legislation like MiFID, particularly those that specifically target high-frequency traders by promoting their low levels of latency. “This was something not thought through when MiFID was introduced,” says Rob Boardman, chief executive of agency broker and trading technology provider ITG Europe. “How do you police the market when stocks are fully distributed?” Above all, though, regulators need to avoid making any kneejerk reactions, says Boardman, particularly when it comes to the technology. “It is always dangerous to
put a lid on ingenuity and you cannot uninvent things.”
If regulators really wanted to address any potential threats to the market’s integrity, they might be better advised to focus on the lack of transparency that pervades the high-frequency trading market, says Execution Noble’s Nash. In today’s fragmented market, where you trade is becoming a much more important consideration and buy-side firms want to be able to identify the venues that attract high-frequency traders. The ability to manage the list of venues they execute their trades on will therefore become an increasingly important factor. “This would reduce the likelihood of being gamed but would not remove it entirely,” says Nash.
“But if buy-side firms could be more informed and empowered regarding the various venues, they will be able to control their trading to a greater extent.”
©2010 funds europe