Chop off those fat fingers

Comment

It seems to be becoming a bit of a habit that, come the turn of the year, a new risk management howler comes to light. In January 2004, National Australia Bank revealed that four of its forex options traders had racked up losses of A$360 million – a loss they were able to hide because of outrageously basic deficiencies in the bank's back office controls and settlement procedures. Then, at the end of that year, Singapore-listed China Aviation Oil disclosed losses of $550 million through trading oil derivatives, despite a company policy that trading positions must be closed once losses breach a stop-loss limit of $500,000.

This year, it's the turn of Mizuho Securities. And this time, an operational risk blunder could end up costing the firm as much as $340 million. The story is well known by now – in mid-December, a sausage-fingered trader accidentally keyed the wrong figures into the firm's trading system, selling 610,000 shares in newly listed recruitment company J-Com for ¥1 instead of going short one share for ¥610,000. Despite frantic efforts by Mizuho to cancel the trade, a bug within the Tokyo Stock Exchange's trading system meant the order went through.

The fact that the exchange's system failed to allow Mizuho to cancel a patently erroneous order and that J-Com stock was not suspended until the next day, causing a mad-cap day of trading as dealers and day traders profited from Mizuho's error, is bad enough. But how did the trade even get out of the door in the first place?

Why did Mizuho's trading system not automatically reject the trade? The only explanation is that there was no link to key market parameters, such as the number of outstanding shares, which would have prompted the system to spew back an error message.

The blunder is especially surprising, given that a similar thing happened in Japan in 2001. Then, a UBS trader accidentally sold 610,000 shares in advertising firm Dentsu for ¥16 each instead of the other way round – a mistake that should have prompted all banks and securities companies to double-check their own systems to ensure the same thing couldn't happen to them.

A lot of banks have been preoccupied with the development of ever more sophisticated pricing systems for exotic options, as well as the implementation of advanced internal ratings-based approaches for Basel II. But are firms forgetting the basics? Yes, trading losses, IT foul-ups and rogue trading will never be completely eliminated, but the losses of Mizuho were caused by shortcomings in trading systems so elementary, it perhaps should encourage all risk managers – and not just those in Japan – to go back and have a good hard look at some of the areas that had been taken for granted.

Nick Sawyer, Editor