Lee Porter, a Hong Kong-based professional in the financial services industry, was in New York when the so-called flash crash by the Dow Jones Industrial Average (DJIA) happened on May 6, 2010. Like many traders and investors on Wall Street and elsewhere that day, he was befuddled by the index’s swift 9% plunge and equally astounding recovery from most of its losses within minutes. “Everyone was like ‘Wow! What happened there?’” says Porter, recalling his initial reaction to the incident. US regulators subsequently blamed the DJIA’s tailspin on a combination of factors, including an usually large and quick sale of certain futures contracts by a mutual fund as a hedge for an existing equity position. High-frequency trading also took the flak for exacerbating the losses, as the mutual fund’s sell order sparked rampant automated selling by super-fast computerised traders, some of which shut down, draining liquidity from the market. Among other things, circuit breakers and increased policing of high-frequency trading firms by regulators have since been in place on Wall Street to prevent similar occurrences. Yet, more than two years after the flash crash, regulators the world over and market participants such as Porter are still wary of rapid-fire computer-driven trading. “When you have technology advancing at such an incredible rate — you’re talking about people trading in micro seconds now — it’s obviously an arms race for technology,” says Porter, head of Asia-Pacific operations at Liquidnet, an independent dark-pool operator. While some high-frequency traders do inject liquidity into the market by acting as market makers, others are “almost toxic”, he says. “They go in and try to take advantage of any indication in the market that there’s a buy-sell imbalance so that they can trade ahead of it. They’re not fundamental investors. They typically don’t care that much about the management [of the company] or, for example, its long-term profitability. As long as they can buy and sell it very, very quickly and make a quick profit, that’s all they care about.” Since the DJIA’s 2010 flash crash, several smaller but no less worrisome incidents involving high-speed computerised trades and software malfunctions have surfaced. One was the near-collapse of US broker Knight Capital Group in August this year, arising from a programming glitch that caused its computers to execute millions of faulty trades. Knight Capital ended up losing some US$440 million ($536 million). Regulators take action Concerns like these prompted regulators in Australia to introduce new marke trading rules last month. Among the changes are so-called kill switches that can be used to immediately stop computer-generated, or algorithmic, trading in the event of sudden or adverse market movements. High-frequency trading accounts for about 30% of daily trades on the Australian Securities Exchange (ASX). In Singapore, high-frequency trading accounts for about 30% of derivatives trading volumes on the Singapore Exchange (SGX), which identified this form of trading as a growth initiative for its equities business shortly after its proposed takeover of ASX failed last year. SGX has yet to roll it out, though, saying it will do so only after appropriate safeguards are in place. Given the caution in other markets, however, the bourse operator is likely to drag its feet before it allows high-frequency trading in Singapore. The Hong Kong Exchanges & Clearing has said that while rapid-fire electronic trading strategies will eventually make their way to the city-state, regulators in Hong Kong are still studying markets in the US and Europe before they push for changes to the local market structure to attract computer traders. In Germany, where high-speed trading is estimated to account for up to 40% of total trades in the market, new rules have been proposed to mitigate related risks. These include requiring traders to register with the country’s financial regulator, pay a fee if they use high-speed trading systems excessively, and disclose in detail their trading practices if called upon by the state. No real benefits? At a trading and technology conference in Singapore last month, the trading heads of several fund houses downplayed the purported benefits of high-frequency trading. One of them, Jonathan Evans, managing director at JP Morgan Asset Management (Japan), said trading using super-fast computers does not boost liquidity in the market. “It provides no real opportunity to get real liquidity,” Evans said. “It’s a net taker of liquidity from the market. High-frequency firms don’t hold inventory. They don’t bring anything to the market and everything is closed out at the end of the day.” In fact, high-frequency trading can paint a false picture of the level of liquidity in the market, noted Kent Rossiter, another speaker at the conference. “It kind of hides what the real situation is. You can look at the trading volumes and see it does a million shares a day. In reality, of the million shares you’re seeing, there are not really a million shares out there for you to take. There are probably only 600,000 or 700,000,” said Rossiter, head of Asia-Pacific trading at RCM Capital Management, part of Allianz Global Investors. According to Liquidnet’s Porter, high-frequency traders who are in the game solely to make a quick buck and have no intention for market making offer no real value to anyone. “You have to question the value of high-frequency trading when many of these traders are opportunistic, predator-type, trading 10 shares at a time, buying and selling, and buying and selling. What value are they providing in the marketplace by buying and selling 10 shares at the click of a finger?” he says. Some of Porter’s peers maintain, however, that there is room for high-frequency trading. Chris Jenkins, Asia-Pacific managing director at TORA, a registered broker- dealer in Hong Kong and a dark-pool operator, says “to tar all high-frequency trading as bad is probably inaccurate”, even if investors such as buy-side institutions are naturally concerned about the risk of being beaten by super-fast computerised trades. “I understand that certain large institutional buy-side firms do not want to interact with high-frequency trading. They don’t want to be gamed by someone on the other side who might be trading against them. I understand that,” Jenkins says. “But there are lots of other people who are more than happy to interact with high-frequency trading. Some high-frequency trading is very relevant, certainly on the ‘lit’ exchanges, which can benefit from it because it puts liquidity out there,” he says. ‘Lit’ pools, like traditional stock exchanges, show various bids and offers for stocks, and are in effect the opposite of dark pools. “In some markets, high-frequency trading is probably about 40% of what’s going on in the market in terms of daily volumes. That’s a lot of liquidity to not have. Even if it’s 10% or 20%, it’s still a lot of liquidity to not have,” Jenkins adds. “Liquidity, or the lack of it, has always been an issue in Asia.” The way Jenkins sees it, regulators sometimes are looking to solve problems that are not necessarily there. “The market should be able to cope with a variety of trading stances and parties, whether it’s slow or fast, mums and dads, hedge funds and so on. All of these people should be able to engage in the market according to how they want to trade. Let’s not throw the baby out with the bath water.” New study Recent findings from an independent research outfit in Australia appear to back Jenkins. The Sydney-based Capital Markets Cooperative Research Centre (CMCRC) carried out an empirical study to find out whether there was any relationship between high-frequency trading and market abuse. Using data from theLondon Stock Exchange and Euronext Paris over five years from 2006 to 2011, CMCRC concluded last month that high-frequency trading does not correlate with an increase in market manipulation. “The debate on high-frequency trading has become almost hysterical in some regions. Yet, it’s characterised by an excess of opinion and deficit of proof,” saysAlex Frino, CEO of CMCRC. “Because orders aren’t tagged as such, you can’t look at any one order and say, ‘That’s high-frequency trading.’” While high-frequency trading has increased considerably between 2006 and 2011, Frino says it does not deserve the bad press it has received. “In an environment of such low returns, everyone is going to be looking for a scapegoat. However, high-frequency trading and its relationship to market fabric is very complex and needs to be analysed as such before any conclusions can be drawn. It’s not good enough just to have an opinion, when regulations are being drawn up that will affect the way markets work around the world.” Dark pools Whatever the case, with trading volumes on stock markets worldwide declining and regulators stepping up efforts to prevent super-fast programme trading from spiralling out of control, prospects in the near term for high-speed trading firms do not appear favourable. Yet, traditional traders and investors hoping for a more level playing field should be mindful that alternative trading venues such as dark pools are still popular in some markets. “When institutions find liquidity, they are going to take advantage of it because it’s a low liquidity environment now,” says Liquidnet’s Porter. “Across Southeast Asia, our volumes this year are up over 20%. You compare that to volumes in any typical market, which are down between 10% and 25%. That tells me we’ve been winning some market share.” The average size of a trade on Liquidnet, which connects its users to 41 markets across five continents, is more than US$1 million, compared with the average of US$6,000 for a trade on SGX and between US$6,000 and US$7,000 on ASX, according to Porter. Still, going by SGX’s experience, dark pools may not be everyone’s cup of tea. SGX had a go at setting up a dark pool in 2010 through an equal joint venture with Chi-X Global. The intention was to allow money managers to anonymously trade large blocks of selected stocks from Singapore, Hong Kong, Japan and Australia on an electronic platform. But the joint venture folded in May this year because of a lack of critical mass.