Slashdot Mirror
Knight Capital Fined $12M For a Software Bug That Cost $460M
Mark Gibbs writes "Knight Capital monumentally fouled up a software update. According to the SEC, 'Knight did not have supervisory procedures to guide its relevant personnel when significant issues developed.' In other words, not only was Knight's code management inadequate but their human management processes were just as bad. The fine for what could have been a biblical financial disaster? A measly $12 million."
The cost to them was $472 M. I *think* that will discourage them.
As a proprietary trading firm, they were working entirely with their own money. They had no external investors or whatnot (like hedge funds do). So, they made a mistake and they paid for it dearly. It's not clear to me that they should have paid any fine.
The article's whole argument seems to be made by comparing the size of the trading loss to the size of the fine, but no logical reasoning is given for why the one should have any relation to the other.
TFA sucks.
The problem with this is that they didn't know they were losing money.
The trading had gone haywire and they didn't understand why it was doing what it was doing but at the time it was happening they couldn't say if they were making or losing money.
They built up positions of billions of dollars and only once it was all unwound and settled were the losses finally known.
I can feel for the programmers and sysadmins. Maybe this was right - maybe pulling the plug out would prevent the unwinding of positions that would then make money for the firm.
I wouldn't be surprised if there weren't previous problems where programmers and admins had been criticized by management for "doing the wrong thing with hindsight" when they didn't understand what was going on. If you have that sort of management culture then the natural inclination becomes to do nothing and push the responsibility up the chain.
Similar disasters have happened in the past - one that springs to mind was Piper Alpha where the other nearby rigs continued to pump gas to it even when they could see it was on fire - because if you stop pumping it takes days and costs a fortune to get things back up and running and it might just have been an easily controllable fire.
There's a very fine line to be drawn between reacting to the unknown too soon and reacting too late. There's also a fine line between making a reasonable best guess with the facts available and just making a random guess.
http://en.wikipedia.org/wiki/Piper_Alpha
The fire would have burnt out were it not being fed with oil from both Tartan and the Claymore platforms, the resulting back pressure forcing fresh fuel out of ruptured pipework on Piper, directly into the heart of the fire. The Claymore platform continued pumping until the second explosion because the manager had no permission from the Occidental control centre to shut down. Also, the connecting pipeline to Tartan continued to pump, as its manager had been directed by his superior. The reason for this procedure was the huge cost of such a shut down. It would have taken several days to restart production after a stop, with substantial financial consequences.
God said, "div D = rho, div B = 0, curl E = -@B/@t, curl H = J + @D/@t," and there was light.
Most all Wall St firm's systems are bloody awful. There are many reasons for this. First, the true business is sales/brokerage so the engineering side, though it is a strategic asset, is often neglected. This includes putting clueless business side people in charge of IT system. Second, the boom and bust cycles of tech investment are a bad way of building tech systems. It's like not watering your garden all summer except for one day when you use a high-pressure fire hose on it. Third, as part of the boom/bust cost cutting they have no employee longevity in tech so no one understands how the mind-bogglingly complex and obscure layers of technology work. Fourth, and more recently for cost cutting, they've dispersed their dev teams around the globe so communication and teamwork are seriously compromised. Fifth, when there is a boom they try to build their systems so quickly that they take all sorts of dangerous engineering short cuts. All this adds up to engineering disaster.
What if you were an investor in that stock who had set a stop-loss at $10? Knight's wild selling would have triggered the stop-loss, and you'd lose money because of Knight's actions.
No, you'd lose money because you sold at the wrong time based on an automated trading rule rather than your own informed judgement. That is the risk you take when you enter a stop-loss order: if the drop in price is temporary, you're going to lose money. It makes no difference why the price moved. Knight's inadvertent trades are not to blame here. No one who payed attention to the fundamentals lost anything. Only those who panicked took a loss, and deservedly so.
"The state is that great fiction by which everyone tries to live at the expense of everyone else." - Bastiat