7th August 2012
As this chart from Felix Salmon on the Reuters blog shows, high-frequency trading (HFT) volumes have grown exponentially since 2007. Yet following the huge trading loss suffered by Knight Capital questions over the benefits of these strategies are growing.
The first line of defence likely to be raised by supporters of HFT is how investors will be able to trade in and out of stocks as easily as they are currently able to without large algorithmic traders.
To an extent this argument is self fulfilling. The market participants with the most pressing need for liquidity are the high-frequency traders themselves who need to be able to buy and sell multiple times in a given second.
For small or medium sized investors the benefits of being able to trade by the millisecond are less apparent. In fact those following a buy-and-hold strategy may actually be inconvenienced by the presence of traders looking for pricing anomalies rather than fundamental value.
Kevin Murphy, part of the value investment team at Schroders, wrote in August last year that because the computer models behind these trades are unable to take individual company balance sheets into account they are prone to "move shares up and down irrespective of their starting valuation or their prospects".
If these trades represented only a small part of the overall market then this impact would be relatively marginal. Unfortunately for value investors, however, they currently make up over three-fifths of overall market trading volumes.
This is a problem for a value investor, Murphy says:
"Thus when the computers stop trading for whatever reason, it can lead to a dearth of liquidity and share prices move very erratically. When the intraday spreads of huge companies can be plus or minus 20% – regardless of whether they have published results – that is obviously a sign things are not quite working properly because no business's value changes that much in the course of a day. This goes beyond speculation and is in danger of missing the point of investment, which is about buying low and selling high, not buying fast and selling faster."
Algorithmic models need these value traders to set the price of assets, taking advantage of data lags to skim profits from small price differentials. As such the strategy has been likened to "picking up pennies in front of a steamroller" – although given the meagre return from any one trade it relies on a combination of volume and speed to drive profits.
The 2010 "flash crash", much like the current Knight Capital debacle, clearly demonstrated that these algorithmic models entail significant tail-risk. Yet the lesson was one that perhaps should have been learned over a decade ago with the collapse of Long-Term Capital Management (LTCM).
Worryingly for supporters of HFT the case of LTCM cannot be not blamed on a problematic bit of code caused by human error in an otherwise functional trading model. The flaw was the belief that financial markets could be modelled to such a degree that all potential risk could be calculated and mitigated.
Before August 1998 Russia had not defaulted on its debt since the revolution of 1917. It therefore seemed a safe bet to assume that it was unlikely to do so in the immediate future. But it was exactly that – a bet.
As the New York Times wrote in 2008 on the tenth anniversary of its collapse:
"Thus, in August 1998, the hedge fund calculated that its daily "value at risk" – meaning the total it could lose – was only $35 million. Later that month, it dropped $550 million in a day."
What could investors expect from a post-high frequency trading world?
Following Joy's suggestion that the primary purpose of machines is to make life easier for people, algorithmic trading might seem to meet the definition. If, however, the frequency of "flash crash" events are increasing then the risk they pose may outweigh the efficiencies provided.
This is not only a matter of process but also one of trust. The vast majority of small and medium sized investors, who should be responsible for setting the prices of assets, have seen huge swings in the value of their assets over a single day irrespective of the underlying strength of the companies. How are they to trust that they will reap the benefit of buy-and-hold investing when they come to sell their assets?
If we define liquidity as buyers always being able to find sellers (and vice versa) then the ideal marketplace is one in which there is a multitude of investors with various investment priorities. Just as the erosion of trust in banks has lead to a sharp increase in peer-to-peer lending, perhaps a collapse in confidence over equity trading could precipitate demand for a true peer-to-peer forum for share trading.
Failure to act will only increase the likelihood of more Knight Capitals over the next few years, as the global economy struggles to regain momentum. As Joy says:
"We are being propelled into this new century with no plan, no control, no brakes. Have we already gone too far down the path to alter course? I don't believe so, but we aren't trying yet, and the last chance to assert control – the fail-safe point – is rapidly approaching."
More on Mindful Money
To receive our free daily newsletter sign up here.