Red-Blooded Risk: The Secret History of Wall Street

Middle Office

In the early 1990s, I moved from being a front-office business head and risk manager to being a middle-office risk manager. This was a term we invented at the time; there never was a physical middle office. The basic problem of all financial institutions was that information flowed only one way, from the front office to the back office. The broker tossed the trade ticket over the wall, and never wanted to see it again. If things didn't add up, the back office was supposed to make sense of it somehow. On rare occasions, for intractable problems, a senior back-office person would approach the broker to ask him to correct an obvious mistake on his ticket. The back-office guy could expect to get yelled at. He also expected to get yelled at every time paperwork was late or wrong, even though the delays and errors were usually caused by the front office. Because front-office mistakes were headaches only for the back office, the front office never learned to do things right.

This problem was exacerbated when financial institutions started to computerize. Now a lot more data was flowing a lot faster, but it still flowed only one way. The information technology (IT) department designed computer systems to replace trade tickets. The systems never rejected a trade. If the program refused to accept clearly erroneous input (say, a stock trade with the name and ticker symbol of the stock missing), the trader went ahead and made the trade anyway and blamed the stupid computer system ...