Easy access to libraries of financial data and lightning fast exchange connection speeds have spawned a new breed of investment strategies that are estimated to account for up to 20 per cent of traded volume on the Australian sharemarket.
While not as influential as in the US, the ASX has experienced "exponential growth" in remote computerised trading in the past six months, according to Justin Webb, director of hedge fund Helix Partners, partly owned by Westpac subsidiary Ascalon Capital Managers.
In the US, computer-driven algorithmic trading accounts for an estimated 70 to 90 per cent of New York Stock Exchange trading.
Algo trading, also known as "black box" trading, is the computer-automated execution of share trades based on algorithms, or mathematical models, that determine the correct time, price and quantity of shares or other assets to buy or sell.
Different strategies include high-frequency trading (HFT), which can execute multiple trades a second, statistical arbitrage, index arbitrage and trend momentum.
They are usually housed in quantitative hedge funds and investment banks, and are managed by "quants" schooled in maths, physics and behavioural psychology. So instead of traders watching a handful of prices on a screen, computers watch thousands from around the world looking for trade signals.
Proponents of algo trading cite increased market liquidity, which leads to more efficient price discovery, with benefits for investors big and small.
Critics claim high processing speeds allow computers to "front-run" small traditional investors, and that when liquidity dries up or computers malfunction, the withdrawal of HFT leads to "flash crashes".
The biggest flash crash occurred in May last year when the US S&P 500 index fell 573 points, or about 5 per cent, in five minutes, before reversing sharply. However, many investors with stop-losses in place lost money as few trades were reversed. The NYSE experiences "flash crashes" in individual stocks regularly.
Mr Webb says algo trading on the ASX is clearly evident, especially from more simple strategies. "When value opportunities arise there are signs of greater capital swings when linear strategies are applied, and these are exploited far more rapidly than in the past," he said.
That meant market laggards were quickly snapped up and buy-on-the-dip strategies could lead to sharp reversals.
Algo trading has been used in US markets for decades. It was a key factor in the 1987 Black Monday Crash and the $US2 billion collapse in 1998 of hedge fund Long Term Capital Management.
Traditional fund managers have also used algorithms to feed orders on to the ASX for more than a decade. Deutsche Bank's head of electronic trading, Ben Radclyffe, says Australia is "the next frontier" for expansion of algo trading.
"It looks like US models will work here and the cost of trading is relatively cheap," he said.
As an indicator of market depth, Australia has about 100 stocks with an average daily trade of $10 million-plus, while the US has 1700. China has 1600, but exchange rules make HFT impossible there.
Mr Webb says the global financial crisis has "jilted thinking" towards more sophisticated portfolio management, and Australian super fund asset allocators are seizing on algo trading as a way to diversify risk and improve returns.
However, there are fears that riskier strategies are not differentiated enough, leading to record high global market correlations and self-fulfilling, ponzi-like strategies incorporating equities, commodities, derivatives and credit instruments 24 hours a day across all regions.
In 2008 after the US sub-prime mortgage crisis surfaced, stocks rallied higher despite signs of severe distress in credit markets.
The buying was driven by a number of major US quant funds that were all forced into indiscriminate short-covering of equities that were supposed to be a hedge for mortgage-related losses. Once the funds covered their shorts the equity bear market resumed.
