For those of us who know and understand high frequency trading, it has become laughable to read what the main street people and reports have to say on the subject. It's even more when they trot out "experts" and never mention the guys from NANEX. I have contributed many times about Flash Trading (CC has a dedicated page here). This latest "correction" was started by algo's, added to by general market news around the world, and perpetuated by the algo's ability to suck up liquidity where it is needed and to amplify where it shouldn't be. What makes me laugh is to hear the talk of regulation. For those with iPads that watched "Money and Speed", you heard Eric Hunsader talk of the SEC's inability to analyze quotes within a second. The SEC blamed Waddell and Reed for causing the Flash Crash. NANEX was able to have W&R send over their data and this is what it looks like (the red dots are W&R trades). The people should understand this stuff, don't, what else is new? The lawyers at the SEC are behind in the race and for good reason, they are out of their league and the pay that is required to recruit the talent needed (along with the technology) is something the SEC is unable to grant, since they can't catch any illegal activity and plunder the goods like the true pro's, the US Government.
From FT:
The good news about the extreme market volatility of the past month is that there has been no repeat so far of last year’s notorious Flash Crash, in which $1,000bn was wiped off share values in half an hour. Yet that is no reason to be complacent about the activities of high frequency traders who are engaged in a high-tech arms race to reduce trade execution times to millionths of a second.
For conventional investors there is a long-standing concern about the uneven playing field on which they compete with traders who can afford to co-locate their high-tech kit expensively beside exchanges to reap a time advantage. While it is true that the bigger fry have always had a natural advantage in capital markets against the small, it is a legitimate goal of regulation to try to level the playing field as far as possible. And it would be good to have a regulatory verdict on whether predatory algorithmic traders are artificially manipulating prices at the expense of slower or less sophisticated market participants.
I am also profoundly suspicious of the practice of “pinging”, whereby traders send out and cancel a multiplicity of orders until they manage to identify hidden pools of liquidity from which they can extract a predatory profit. The orders entered by such traders are now a huge multiple – even hundreds of thousands – of the trades actually effected.
Small wonder Lord Myners, the UK’s former City minister, expressed concern last week that high frequency trading was so remote from the true function of the capital markets. It is high time we had a definitive verdict from the securities watchdogs on whether the unholy alliance between these traders and the managers of exchanges and trading platforms acts as a stealth tax on ordinary institutional and retail investors. For the formal exchanges in particular, which have seen their market share collapse as a result of liberalisation and the arrival of new trading platforms such as dark pools, the increased volumes generated by high frequency trading have been a boon. Note, in passing, that the New York Stock Exchange’s share of the trade in NYSE-listed securities fell from 80 per cent in 2005 to just 24 per cent in early 2011.
Yet the real concern is systemic risk. Carol Clark of the Reserve Bank of Chicago told the story in a recent paper of a US trading firm that became insolvent in 16 seconds in 2003 when an employee who had no involvement with algorithms switched one on.* It took the company 47 minutes to realise that it was bust and to call its clearing bank, which knew nothing of the situation. Since then volumes have gone through the roof and intra-day volatility has conspicuously increased.
The social justification offered by high frequency traders for their role is that their de facto market making activity has caused bid-ask spreads to shrink, to the benefit of all investors. Liquidity, they add, has greatly increased as a result of their activity. Yet the reality revealed by the official report on the Flash Crash was that this was no more than fair-weather market making. Many firms significantly scaled back activity and as a group they were net sellers of stock at that time. As the Bank of England’s Andy Haldane noted in a recent paper, far from mitigating market stress these traders appear to have amplified it. Moreover, they have substantially squeezed out conventional market makers. Those that are left, says Mr Haldane, are at an acute informational disadvantage in situations of stress, “The result,” he adds, “is a potentially double liquidity void”.
Another worry is that the emergence of these traders in increasingly fragmented trading platforms makes for more complexity and contagion, which are classic ingredients of systemic disaster. And the greater the volatility they generate, the bigger the potential flash crash. Speed, which is what their game is all about, simply increases the risk of liquidity feast and famine.
The difficulty here is that the seemingly reasonable pursuit of profit by individual traders is collectively harmful. So regulation clearly needs to address the potential systemic consequences. Yet regulators are always one (or more) step behind market practitioners. And now for the first time in financial history machines can execute trades far faster than humans can intervene.
If that sounds a nightmare, it probably is [sic]. Mr Haldane points out that while regulatory intervention could feasibly have forestalled the 1987 stock market crash, regulators at the time of the Flash Crash might have blinked – literally blinked – and missed their chance. Yet he believes a combination of market making commitments to provide liquidity, circuit breakers and minimum resting periods for trades could resolve the regulators’ dilemma. Let us hope he is right and that such rules are in place before the next flash crash occurs.
