the geeks of wall street

Article in last weeks BRW magazine, here is a summary of some of the most criticle comments

The merging of mathematics and fast computers with share trading is creating a market suitable only for the biggest players.

Software based trading (often refered to as algorithmic trading) already has had big influence on the market. "really, its come out from nowhere in the past three years to the point where it accounts for about 50% of the volume traded and weve seen its effect on the market as the size of each trade decreases but the frequency of trades has increased", Frino says (Alex Frino, professor of finance at the university of sydney and chief executive of the Capital Markets Co-operative Research Centre)

The chief executive of market monitoring software group Smarts, Andreas Furche, believes that most of the institutional trading on the Australian market is now channelled through execution algorithm software as it enables investors to execute an order very efficiently.

"virtually every order hitting the market from institutional trading is today fed through automated trading platforms,--- the software generally splits the orders up into small order lots, often the minimum lot size"
Furche says.

The second and more recent development is sophisticated algorithmic software that draws on market data to create and execute orders automatically. Referred to as situational algorithms, these are favoured by traders who seek to gain in short term speculation.

Complex, costly and hard to track, this second generation of algorithmic trading platforms potentially will cause headaches for market regulators who have to maintain a level playing field for investors, according to Frino.

"What markets and regulators are struggling with at this stage is determining what constitutes an illegal trade in the realms of algorithmic trading " Frino says. "Because timing is so crucial to these systems, some brokers position their servers in a facility close to the markets servers. The proximity shaves tenths of a second off the time it takes to execute a trade, but that makes a difference".

Despite the argument that the speed and fluidy of algorithmic trading leads to a more efficient market, the risks of the rapid uptake of algorithmc trading cannot be understated according to Frino.

"Regulators have no real mechanisms to robot-gone mad scenario,where a single software failure triggers a systematic response across the market," Frino says.

The cost and complexity of the systems has also made it impossible and , Frino argues, potentially dangerous, for small investors to participate in the market now dominated by institutional investors armed to the teeth with the best money software can buy.

" I dont understand why the so called mum and dad investors are even in the market," Frino says . "The risks are just too high and ultimately they are competing against machines that spot and take advantage of an opportunity within a split second. Even if they buy software, it will be out of date within a matter of months because all the trading houses have a team of developers who are all trying to beat each other."