Yesterday market maker Knight Capital lost $440m in the space of 45 minutes due to a technical problem in new software that had been implemented the night before, which led to an algorithm going bad and bidding up shares.
The consequences of this could potentially be devastating. Knight Capital is not the type of firm that could shrug of this kind of loss à la JP Morgan and right now it is desperately exploring every strategic alternative that might prevent it from going under.
So far, the signs aren’t encouraging. Ratings company Egan Jones downgraded Knight Capital all the way down to CCC on the basis that its net revenues were down, there are likely to be few buyers in the market for it, that it currently doesn’t have enough capital to continue and won’t be able to raise enough without a buyer.
Then reports started to emerge that Knight Capital’s biggest clients were moving their business to other firms as the extent of the damage became clear. More recently questions have been raised about whether Knight Capital will be able to prevent competitors coming in for their key staff members.
Knight speaks out
Thomas Joyce, chief executive of Knight, took to the air on Bloomberg TV to defend the firm, declaring “we’re open for business”.
The emphasis from Joyce was that Knight had moved to protect their clients as soon as they discovered what he described as a “software bug” and that none of their interests had been harmed, with Knight taking it on the chin themselves instead.
“All of our clients respect what we did yesterday and were very happy with us because once we realized that we had a problem, we alerted them and got them out of the way,” said Joyce.
Indeed, at times Joyce’s attitude appeared alarmingly c'est la vie. “Technology breaks. What happens next is what matters…..You can’t immunise people against making mistakes, you can’t keep people from doing stupid things….That’s what happens when you have a culture of risk,” was his response to this monumental failure of Knight’s technology.
That no one will get hurt by this except Knight seems unlikely, as does the idea that Knight’s clients are happy with the firm.
Back to the algo debate
However, Joyce does make a valid point that there is room for error in any one individual system. The failure here though does not only lie with faulty code in the software but with the systems that were supposed to be monitoring the firm’s activity and this has implications that go far beyond Knight Capital and equity markets.
The Flash Crash, BATS’ failed IPO launch and the Facebook IPO fiasco are all examples of where technology gone wrong with a significant and detrimental effect on markets.
HFT and algo trading has been a source of constant and heated debate in the derivatives industry and Knight Capital will be held up by many as an example of why regulators need to take a tougher line with firms that engage in this type of trading.
Rik Turner, senior analyst of financial services technology at Ovum, argued that the debate that will inevitably ensue from Knight’s losses about whether technology has elevated risk in trading to unacceptable levels is missing the real issue.
“To argue that modern trading in the financial markets is overly dependent on technology suggests that there is an alternative, which there really isn’t,” he said.
While Turner conceded that there could be a case for arguing that the automation of trading has increased risk he claimed that the reality of the situation is that this software is here to stay.
“The challenge is to monitor and manage its performance and to deploy circuit breakers for exceptional circumstances like that when the market goes crazy.
“Knight Capital’s shareholders should certainly be asking about monitoring systems for the company’s software, not to mention its insurance policies against operational risk,” he added.
Checks and balances
While an individual piece of software can malfunction there should be enough checks and balances in the system to prevent it from doing $440m worth of damage and crippling a company.
The majority of firms in the industry run reconciliations between the front and middle office. These reconciliations should be done in real-time to prevent this kind of damage and the technology is certain out there that allows this.
Joyce played down the suggestion that Knight’s failure was proof that electronic trading is dangerous and instead claimed that the software problem affected its infrastructure rather than market making or trading.
While it will perhaps take some time before the exact nature of the technological fault is made known to the rest of the market, it definitely looks right now as though Knight’s infrastructure was to blame.
Inadequate checks and balances failed to pick up a major problem that in just 45 minutes was able to bring one of the major equities market makers to its knees.
The next few days will be make or break for Knight Capital and in the meantime other firms will hopefully be looking to their own systems to ensure that they don’t suffer the same, entirely preventable, fate.