Taking a global look at BOT trading he's a longish but scary...

Taking a global look at BOT trading he's a longish but scary article. Bot is a new cyber monster.

Markets are inherently inequitable, inevitably favouring some types of investors and transactions over others. A question being studied with increasing intensity and concern by regulators around the globe is whether some inequities are more inequitable than others.

The ASX today released a review of algorithmic trading and market access arrangements. While the review started in July last year it gained a particular focus after the Australian Securities and Investments Commission last year identified a number of areas of focus that it wanted included in the review.

ASX and ASIC arent the only market operators and regulators trying to come to terms with the growth in algorithmic trading generally computer-generated trading based on mathematical algorithms and designed to place orders automatically to either reduce market impact and execution costs or create trading profits. The US Securities and Exchange Commission is also reviewing its equity markets and trading technologies and practices. Algorithmic trading is far more prevalent in its markets than on the ASX.

While there is algorithmic trading occurring in our market it isnt of the same scale or complexity as that occurring in the US and elsewhere.

However, the ASX believes that will change with the likely granting of licences to alternate market operators. Last year the federal government announced it would transfer responsibility for supervising trading on the ASX and other platforms to ASIC widely seen as preparation for the imminent introduction of competition for the ASX.

The reasons the ASX proffers for the relatively slow take-up of algorithmic trading so far, and its prediction that it will increase more rapidly in future, are its current monopoly and the likely emergence of competitors.

The experience in the US and elsewhere is that where there are multiple trading platforms and fragmented liquidity traders use algorithmic trading to capture arbitrage and trading opportunities that arent available in a single market.

Such trading isnt necessarily a threat to market liquidity or integrity. Indeed, the ASX says it can add to liquidity, while adding the rider than the extra volumes generated by algorithmic trading dont necessarily increase liquidity.

There are, broadly, two types of algorithmic trading. One is designed to improve execution of orders by minimising the impact the orders themselves have on the market reducing the extent to which a big buying order might drive up the price, for instance. The most commonly used algorithm in this market, the ASX says, is one designed to achieve or better the volume-weighted average price of a security over the course of a day.

The other, which the ASX terms situational', is designed to profit from changes in data, information and events. It is about trading strategies rather than efficient execution (although situational algorithms would obviously be tied to executional algorithms to produce efficient execution of the strategies).

It is the latter type that is bemusing and concerning regulators and legislators. They are worried that traders using algorithms are undermining the transparency of markets, adding to their volatility, introducing new sources of risk, increasing the potential for market manipulation, exacerbating inequities and changing the nature of markets from facilitating efficient capital formation and risk transfers to forums for pure short-term trading. It isn't clear whether or not any of those things are occurring the sophistication of the technologies and strategies and the opacity of the trading makes it almost impossible to come to any definitive conclusion although the potential for abuse and damage to markets is obvious.

One of the types of trading that is most disconcerting for the US regulators is high-frequency trading. Thats perhaps not surprising, given that it is now estimated to account for as much as 70 per cent of the trading volumes in US equity markets.

That kind of trading is predicated on exploiting tiny discrepancies between markets and trading platforms by churning numerous but relatively small trades in a stock. Trading times can be measured in milliseconds and, as the ASX review noted, has the potential to distort "price discovery" it is trading driven by arbitrage opportunities not any attempt to assess a stocks fundamentals and the trading could create confusing volume and price signals.

The distortion could be magnified and become manipulation or, as the ASX terms it, "micro-manipulation" if high-frequency trading is also pursuing a momentum strategy, where the trading can by itself magnify the arbitrage opportunity and impact on the market. Given that algorithmic trading can both place and cancel orders in slivers of a second (and it could be the same orders that is both placed and cancelled) the potential for manipulation is obvious.

Competition between platforms for liquidity exchanges offer rebates to high-frequency traders in the US to win their volume could add another layer of distortion and inequity.

The emergence of techniques like 'flash orders' where some platforms allows traders computers a millisecond preview of an order ahead of the general market and the growth of dark pools of liquidity, where large orders are transacted out of sight of the market before being booked after the event, adds to the opacity and potentially inequity between types of market participants.

Some US platforms also allow their brokers to grant naked or unfiltered access to trading platforms to the algorithmic traders, which speeds execution further. The perceived advantage and potential profit to be gained by shaving fractions of a second off execution times can be seen in the spread of co-location, where traders actually locate their systems within the exchanges, with the exchange's active support. Whether it is in the interests of markets to give direct and priority access to unvetted and unsupervised participants is doubtful.

While the SEC and ASX reviews demonstrate some concern about some aspects of algorithmic trading, the ASXs recommendations to ASIC are somewhat vague and directional rather than prescriptive. They (understandably but self-interestedly) focus on the threats to market equity and integrity that will emerge with the multiple platforms that ASIC will regulate in future.

It is true that participants in markets have never been equal institutions and wealthy individuals have always had preferred access to information and opportunities and lower transaction costs than mums and dads.

The core issue raised by the more complex and exotic types of algorithmic trading for ASIC, the SEC and their fellow market regulators is how potentially inequitable, opaque and vulnerable to invisible manipulation are they prepared to allow trading in equity markets to become.

Another way of looking at that question would be to ask whose interests, if there were a potential conflict between the interests of market participants, should the authorities prioritise traders or investors.
ASX Stock Chart for ASX

So our market regulators have a big complex job ahead.