Morgan Stanley Fined $300,000 for $10.8 Billion Trade Mistake
Financial firms deal with very large amounts of money, most often electronically. Instructions are incessantly sent to execute a particular payment, loan, insurance, security, derivative or financing transaction. Funds are received here and disbursed there, many millions of times each day. Over time, significant policies, procedures, systems and controls have been instituted to help ensure that both sides complete each particular financial transaction as expected. With very high regularity, this is indeed the case. However, when the processes, people or systems fail, the losses to a financial institution can be quite significant.
As an example, in late 2005 the securities arm of Japan's second largest bank, Mizuho Financial Group, took an order to sell some stock. The preceding day 2,800 shares in a small company had been issued in an initial public offering. However, instead of selling one share of this new stock at 610,000 yen as the customer wanted, the order was somehow passed on to the Tokyo Stock Exchange as an instruction to sell 610,000 shares for one yen each!! The price and quantity amounts had been reversed. Chaos ensued. Dealers and investors scurried to buy the cheaply offered stock. Meanwhile, Mizuho first tried to cancel the order and then scrambled to buy back all of the stock that mistakenly had been sold short. The total loss from this data entry error was 41.5 billion yen (or approximately $350 million).
Now the regulatory arm of the New York Stock Exchange has fined Morgan Stanley $300,000 for a failure to provide sufficient inhibitors and blocks within its trading system that led to a massive botched trade. According to Reuters,
According to the regulator, on the morning of September 1, 2004 a Morgan Stanley trader received instructions from a customer to unwind a portion of an existing swap, with an affiliate of the firm acting as the counterparty. To hedge its position, the affiliate took a short position in the shares of common stock underlying the basket. The mistake came about when the trader entered an order to buy 100,000 units of the basket to cover a portion of the affiliate's short position, not realizing that the tool used to create the basket had a built in multiplier of 1,000. The result was a basket with a value of $10.8 billion instead of $10.8 million.
Over 81 million shares with a market value of $875 million were executed before the firm cancelled the order. NYSE claims that the order cause significant market disruption as the accuracy of the erroneous orders was verified. NYSE charged Morgan Stanley with fault for the error claiming that the order entry system used by the trader had the capability to set trading limits, which would have prevented, or at least limited the impact of the error. The firm neither utilized the functionality nor told the traders that it existed.
Interestingly, NYSE claims that the firm then did not have adequate controls in place to validate order accuracy and to establish limits on block orders that exceed certain parameters. That has now changed. Each trader on its swaps desk now has preset trade limits and other safeguards have been added.
Both the Mizuho and Morgan Stanley events are examples of "fat finger" operational risk losses where incorrect keys were pressed with no malicious intent. Other financial firms, particularly those involved with the dynamics of trading rooms, have suffered smaller, but nonetheless significant, losses. Normally, such data errors are caught by internal processes, people reviews or system checks. Questions like "Does this order size make sense for this customer?" or "Is this order price 'close' to the previous trade price?" often are quite effective in preventing operational errors. Hence, frequently no operational losses result. However, as new financial products, services, activities and transaction methods are introduced, data errors will likely occur again and more “fat finger” operational losses will result.
Toomre Capital Markets LLC encourages the reader to pause for a moment and to consider where in their operations they may be exposed to "fat finger" loss events. What processes, people and system checks are in place to prevent the mistake from proceeding further? Surely no organization wants the adverse publicity like both Mizuho and Morgan Stanley have suffered as a result of these innocuous events. Reader thoughts and comments are welcome.