FYI | Nov 06 2017
High frequency trading is leading to average trade sizes on the ASX collapsing to around one tenth of the average size of twenty years ago.
– Market up, volumes down
– Algorithms at play
– HFT on the rise
By Greg Peel
Deutsche Bank highlighted some interesting statistics this morning regarding trading on the Australian Securities Exchange.
The Australian stock market rose 3.7% in October to mark an 11% gain year on year. The value of all trading year to date was $496bn, down -1% from the same period last year. Capital raisings were up 1%. The average daily value of trading is down -3% to $4.2bn.
The number of individual trades placed on the market is up a remarkable 21% year on year. The average size of individual trades is $3,762. This may not strike anyone as remarkable until one considers that twenty years ago, the average trade size was around $34,000.
What has changed?
Well first we might consider that coming out of the early nineties recession, retail interest in the Australian stock market was at a low ebb. The bulk of trading was left to institutions, being large, well-known fund managers, who traded in sizeable blocks. But with the privatisation of Commonwealth Bank in the early nineties and then the first tranche of Telstra’s privatisation in 1997, Australians began to rekindle an interest in share market investment.
By the early noughties it became apparent that aforementioned large fund managers were sucking the life out of superannuation returns through excessive fees. There began the inexorable rise of self-managed super funds. Between renewed retail interest, and individuals managing their own super, it makes sense that average trade sizes would reduce from the 1997 level.
In the 21st century, ASX trading became ever more computerised and computer power ever greater. And faster. So began, as Deutsche Bank puts it, “the inexorable rise of the bots”.
High frequency trading does not have a lot of fans outside of HFT operations themselves. HFT uses algorithms to, in effect, train computers to trade by themselves without human intervention based on signals the market is sending out at any given moment. HFT is fundamentally ambivalent. Whereas a human investor might decide to buy and hold BHP for a period on a belief that stock is set to rise, HFT simply nips in and out of the market on an incremental basis deploying only small levels of funds with the intention of closing out that trade just as quickly.
In and out, all day long, and always square when the closing bell sounds. The intention is a lot of increments add up to a lot of money over a period of time.
HFT is no guarantee of a profit. There are days in which net HFT trading exceeds 50%, meaning HFT exponents are trading against each other as well as genuine investors. By default, there must be losers among them as well as winners. And as for those mysterious algorithms, there is no mystery at all.
Algorithms simply quantify and convert into computer logic exactly the same short-term trading tactics humans have used in stock markets, or any market, for decades. It’s just that they operate that much faster. In milliseconds, faster than the human eye can ever notice on a screen.
As computer power continues to increase exponentially, so, too, will HFT. And presumably that will lead to even lower average trade sizes. But the more the number of HFT participants grows, the lower the profit pool, given more will be trading against each other.
Does HFT disadvantage the genuine investor? On a day to day basis I’d say not really. What is often noticeable nevertheless are frequent sharp moves in the first half hour of trading on the ASX – the “opening rotation” in which one by one trading opens in individual stocks – which more often than not sharply reverse once human interaction kicks in. These sharp moves tend to suggest HFT feeding on itself, as algorithms fail to recognise the difference.
This might become an issue if we see another 1987, just as “portfolio insurance” (or more correctly, the market-makers on the other side of put option positions) exacerbated the market fall back then. But then, so too might the growth of exchange-traded fund (EFT) investment, which could well lead to a similar result in a similar scenario.
The one thing we do know about HFT is that it’s not going to go away.
Find out why FNArena subscribers like the service so much: "Your Feedback (Thank You)" – Warning this story contains unashamedly positive feedback on the service provided.
