Knight in soiled armour

Last week the markets experienced another little jolt, similar to the flash crash in 2010. A computer glitch at Knight Capital caused its systems to send out incorrect orders, causing huge price swings in stocks and driving the company nearly to bankruptcy.

Bugs

There is now talk of changing regulations (again) and everyone wonders what Knight Capital did wrong. They had a bug. I cannot be certain, but my guess is that whatever software they were using had been tested. But with sophisticated systems bugs always slip through (it’s inevitable). Unluckily for Knight their particular bug caused a lot of trouble and damaged the company’s reputation, possibly irreparably.

It’s (almost) all in the mind

What is interesting is that the company is set to survive. Knight was bailed out not by government, but by a handful of its competitors – perhaps they know this could just as easily have happened to them. They were not just being helpful, though. They saw an opportunity to buy a large stake of a good business (potentially good, in any case) very cheaply. The major obstacles to Knight’s continued survival were a lack of capital and loss of confidence. Both were addressed by the bailout. The very fact that these companies were willing to fund Knight will give the market assurance that the business was worth saving.

Reputations

This bailout comes at a price to current shareholders (who are the ones who should pay – not taxpayers and not clients) whose holdings are diluted. Perhaps, slowly, clients will return. The key reason to think this may happen is the reputation of Knight’s CEO, Thomas Joyce, which seems to have been both battered and uplifted in the debacle. Unlike Bob Diamond, CEO of Barclays and other banks’ top brass, Joyce’s integrity is not in question. He has been called a hero for managing to get hold of the much needed financing. Knight also, it seems, absorbed most of the losses – shielding their clients.

What is in question is Knight’s risk management and software testing. And after an incident like this, I think this is liable to become too strict rather than too relaxed. They cannot afford another incident – that would almost certainly end them, if not through a direct loss, then due to a loss of confidence.

Operational risk 

Knight’s problems highlight, once again, that the major risk in business, any business, is operational. It is unpredictable, its costs can be little or gigantic. Even with good risk management procedures (which are a must – and there is no reason to believe Knight’s were not adequate) mistakes will be made. The regulatory reactions to this will (probably) be firstly to increase the amount of capital that companies need to hold so they can absorb operational losses, secondly to mandate more stringent risk management, and thirdly to demand more detailed reports to the regulator (in this case the SEC). All of these things have their costs.

People will, of course, be very interested in knowing what exactly caused the software malfunction and how. And what Knight will be doing to prevent it from happening again. But what the error was hardly matters – it was random. Next time it will be something else. Hopefully responses will focus less on the specific nature of the problem that arose and more on the general nature of operational risk, which has a tendency to pop up in unexpected places.

References

• Kisling, W., & Mehta, N. (2012). Joyce Puts Knight Survival Over Shares in Rescue Deal. Bloomberg. Retrieved August 7, 2012, from http://www.bloomberg.com/news/2012-08-06/joyce-puts-knight-survival-over-shares-forging-400-million-deal.html
• Pratley, N. (2012). Knight Capital’s computer “glitch” shows dangers of desire for faster trading. The Guardian. Retrieved August 7, 2012, from http://www.guardian.co.uk/business/nils-pratley-on-finance/2012/aug/06/knight-capital-computer-glitch-trading?newsfeed=true
• Reuters. (2012). Knight Capital handed $400m lifeline after trading debacle. The Guardian. Retrieved August 7, 2012, from http://www.guardian.co.uk/business/2012/aug/06/knight-capital-400m-lifeline
• Sapa-AP. (2012). Knight Capital’s $440m computer glitch. Times Live. Retrieved August 7, 2012, from http://www.timeslive.co.za/scitech/2012/08/03/knight-capital-s-440m-computer-glitch
• The Economist. (2012). Desperate times. The Economist. Retrieved August 7, 2012, from http://www.economist.com/blogs/schumpeter/2012/08/knight-capital?fsrc=scn/fb/wl/bl/desperatetimes
• Touryalai, H. (2012). Knight Capital: The Ideal Way To Screw Up On Wall Street. Forbes. Retrieved August 7, 2012, from http://www.forbes.com/sites/halahtouryalai/2012/08/06/knight-capital-the-ideal-way-to-screw-up-on-wall-street/

Another day, another embarrassment

It seems that Wall Street never misses an opportunity to embarrass itself. The most recent scandal is a multi-billion (read milliard in strictly correct British English) dollar trading loss at JP Morgan. The exact value of this trade (or rather collection of trades) will only become known when it is finally unwound. Details of what happened are sketchy and likely to remain so.


Why don't we know what happened?

At least partially we don't know because JP Morgan won't tell us. In fact the market probably already knows far more than JP Morgan would like. As long as the trade is not completely wound down, it is bad for the bank if others know exactly what it has been up to, because then they can more easily trade against it. In fact the market has already done so, magnifying the loss. More fundamentally, however, we don't know because nobody knows, perhaps not even the trader, known as the “London Whale” who initiated the trades. It is somewhat frightening to think that such things can happen without the CEO knowing, but one man can only know so much and there is just too much too know in an organisation so large. Ironically, JP Morgan ended up betting against itself – one unit of the company took the opposite position to the loss-making trade, offsetting part of the losses.


The department of mysteries

The unit of JP Morgan where this loss took place, the Chief Investment Office, appears to be a bit of a mystery. It has been described as being "charged with managing the bank’s idle cash to earn a profit while minimizing risk". Well, if it really were doing that, it should not have taken such huge bets, not under any circumstances. Other reports say it was meant to hedge risk. This means it was supposed to enter into trades that would minimize possible losses on other positions. From what little is known of the trades involved, though, they seemed more like pure speculation (although some of the losses were offset by other positions in the company). I can imagine that the trades were in a kind of gray area – hedge-like enough to put through as legitimate and too profitable (at least momentarily) to really question. What is clear, though is that the operations in the CIO and its links with other departments failed and they failed horribly. This was not really a trading loss. It was an operational loss, caused by the failure of proper risk management, proper oversight and proper communication.


Whose fault is it?

A lot of blame has been dished out. The trader involved has taken a lot of blame. It does not, however, appear that he was a rogue trader. He was making ill-advised bets, yes, but his trades were signed off, it seems. He may even have thought he really was hedging risk (rather than creating it). The person at the head of the CIO, of course should (and has) taken much blame. And so has Jamie Dimon, JP Morgan's CEO. He deserves blame, not for not being aware of these trades (he cannot know what every individual trader is doing), but for ignoring reports about these trades when they surfaced before the loss became evident.


The problem with complexity

A part of the problem, perhaps, lies in complexity. The trades that caused this huge loss were, it seems, very complex and interconnected. Built up over a number of years, they could have grown in complexity to the point where even their progenitor did not fully understand them. JP Morgan itself is a frighteningly complex institution (I admit I looked at its wiki page and did not even try to unravel its intricacies). Human minds cannot handle such complexity. We need to cordon off specific elements, simple enough to comprehend, to specific people and groups. The people at the top must handle the big picture, which can be made simple enough, but they cannot know all the details. In such an environment things get lost, things are forgotten, and errors are made. Ultimately accountability is hard to assign because no one (or two, or even three people) knew enough to have done something. This, of course, plays right into the "too big to fail" debate because bigger institutions are more complex. However, it is not clear to me that forcing these institutions to split into smaller ones would help (they became this big for a reason, so being big does have advantages).

Is it really a problem?

On the one hand this loss represents an egregious failure within JP Morgan. On the other hand, we should expect such losses (operational and otherwise) to occur from time to time. If you invest in stocks you don't always expect them to go up. If you have a large institution you don't expect everything to always go like clockwork. What regulators have attempted to do is to ensure that banks have enough capital to absorb these losses. At least, in this case, the loss was easily covered and only made a little dent in the bank's balance sheet. The instinctive response of more and stricter regulation may not be the right one, not if it increases complexity even further.


The final curtain

The details of the trade and the loss, perhaps, are not all that important. The point is that yet another financial institution has (for not the first time in its history) proven that it cannot be trusted. JP Morgan was in the thick of the speculation in 1930s and now we have this loss. I imagine it was not blameless in the raging financial crisis either. I plan to work in the financial sector. I hope I do not one day have to say "We know we were sloppy. We know we were stupid."

Some references

• Campbell, D., 2012. JPMorgan Faces $4.2 Billion Trading Loss. Huff Post Business. Available at: http://www.huffingtonpost.com/2012/06/05/jpmorgan-chases-trading-l_n_1570578.html [Accessed June 11, 2012]. 
• Campbell, D., 2012. JPMorgan Faces $4.2 Billion Trading Loss, ISI Forecasts - Bloomberg. Bloomberg. Available at: http://www.bloomberg.com/news/2012-06-04/jpmorgan-faces-4-2-billion-trading-loss-isi-forecasts.html [Accessed June 11, 2012]. 
• LaCapra, L.T., 2012. JPMorgan trading loss shows danger in bank size -Volcker. Reuters. Available at: http://uk.reuters.com/article/2012/06/07/uk-jpmorgan-volcker-idUKBRE85615720120607 [Accessed June 11, 2012]. 
• Macke, J., 2012. JP Morgan Trading Loss: Still More Questions Than Answers. Yahoo! Finance. Available at: http://finance.yahoo.com/blogs/breakout/jp-morgan-trading-loss-still-more-questions-answers-154010320.html [Accessed June 11, 2012]. 
• Schwartz, N.D. & Silver-Greenberg, J., 2012. JPMorgan’s Trading Loss Is Said to Rise at Least 50%. New York Times. Available at: http://dealbook.nytimes.com/2012/05/16/jpmorgans-trading-loss-is-said-to-rise-at-least-50/ [Accessed June 11, 2012]. 
• Smith, A. & Thomasch, P., 2012. JPMorgan $2 billion loss hits shares, dents image. Reuters. Available at: http://www.reuters.com/article/2012/05/11/us-jpmorgan-trading-idUSBRE8491H020120511 [Accessed June 11, 2012]. 
• Wikipedia, 2012. 2012 JPMorgan Chase trading loss. Wikipedia. Available at: http://en.wikipedia.org/wiki/2012_JPMorgan_Chase_trading_loss. 
• Zuckerman, G. & Burne, K., 2012. “London Whale” Rattles Debt Market. Wall Street Journal. Available at: http://online.wsj.com/article/SB10001424052702303299604577326031119412436.html [Accessed June 11, 2012].