less technophilia - have faith in fools

link Leader Column, Apr. 99, Treasury Management International, London

Technophilia
link 1st Quarter 2000, Global Trading, London

link

By Warren Edwardes *

In 1991 Warren Edwardes set up Delphi, a London-based consultancy dedicated to consulting and training in financial product innovation and risk management. Prior to that he was on the board of Charterhouse Bank as Director, Financial Engineering in the Capital Markets Division. He graduated in Probability in 1974 and then in Econometrics and has worked for the Equitable Life Assurance Society, the UK Government Actuary's Department and the Treasury and/or Asset/Liability Management Units of Barclays Bank, British Gas and Midland Bank. For TMI Warren Edwardes wrote “Hedging Yardsticks”, Jul./Aug. 1993 and "Derivatiphobia" Dec. 1998. "How to capture The Big New IDEA" will be published later this year.

This article forms the basis for a chapter in "Key Financial Instruments: understanding and innovating in the world of derivatives" 4 February 2000, Commissioned by Financial Times Prentice Hall ISBN 0273 63300 7 London  link

Translate this page! Systran by Alta-Vista Translation Services

"Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in informatiom?" T.S.Elliot

Perhaps the treasury management pendulum has swung too far in the direction of technophilia and away from common sense. Value @ Risk and other measures have become the be all and end all of risk management. The assumptions behind Nobel Prize winners' models are infrequently questioned. But it is not a question of barrow boy uneducated trader versus physics PhD. It is just that the balance has swung too far towards blind faith in calculations that boards do not adequately understand, and more dangerously, are too embarrassed to demand justification. It is incredible that the entire case by Kidder Peabody against Joseph Jett was that he used the management accounting system designed by them to generate performance bonuses. There should be more emphasis on the big picture. Bank and corporate boards should hire more fools. The fools should have a brief to question everything. If only Korean banks had had a few fools to question loans to chaebols.

Sophisticated technology on its own is no solution. It has to be correctly interpreted. And the limitations of the output must be understood. In ("Derivatiphobia", TMI Dec 1998, Princeton Economics, 2nd Quarter 1999) I related how what was then the very latest technological device, radar, had alerted its operator in Pearl Harbour, Hawaii. The machine correctly detected a squadron of planes approaching; but they were assumed to be American. Similar human error and over-confidence in technology was the cause of the bombing of the Chinese embassy in Belgrade on 7 May 1999. Three days before Nato bombed the embassy, a military intelligence officer told colleagues that the wrong building had been chosen. The intelligence analyst noticed surveillance photographs of the building proposed for a Nato strike did not appear to coincide with the intended target, the Yugoslav Federal Directorate of Supply and Procurement (FDSP). The analyst did not know that the building in question was in fact the Chinese embassy. Apparently, a series of errors caused the bombing. First, CIA staff were using old maps and databases which did not show that the embassy had moved location. Then an analyst made a mistaken assumption about the FDSP's location based on its address. Finally a second analyst noticed discrepancies between the FDSP's supposed location and that of the target building. This middle ranking officer called a another middle ranking officer in Europe and conveyed his concerns, and at the same time he tried to fix a meeting within the CIA to relay his concerns. But he failed to arrange the meeting at the CIA. He then went off for scheduled training and when he returned on 7 May he found that the raid was planned for that night. According to The Washington he was told: "The bombers are in the air. It's too late".

Probabilistic models have to be understood for what they are. It is a truism that one can drown in a lake three feet deep on average. Mathematical ability and familiarity with computer screens is not enough to make a good dealer. Some years ago a bank’s HR Director observing that dealers spent all day in front of computer monitors wondered "Who else does that?". The bank hired several bright air traffic controllers. They soon picked up the technicalities of foreign exchange. But as air traffic controllers they had been trained to avoid risk at all costs. They just sat frozen in front of their screens.

According to a study by Elizabeth Brannon and Herbert Terrace of Columbia University, (Ordering of the Numerosities 1 to 9 by Monkeys, Science Oct 23 1998) even monkeys can grasp the concept of numbers. Two rhesus monkeys were taught to count using touch-sensitive screens of images that each contained one to four objects. The monkeys were taught to touch the screens in order, from the image with one object to the image with four objects. If the monkey touched all four pictures in the right order, he got a banana pellet. If the monkey made an error, the screen blacked out. The rhesus monkeys were more than happy to make mathematical decisions and make errors in the pursuit of banana flavoured pellets. And the air-traffic controllers just would not trade.

But the Long Term Capital Management story suggests that it is not just corporate boards of directors that need to do their homework. How many senior bankers really understand the risks that they are carrying? There is a touching blind faith in the computer outputs produced by the rocket science mathematical modelling of financial derivatives. Common sense "foolish" guesstimates are not always made. I was recently asked to analyse a complex structured cross-currency swap with embedded third currency and term options. An investment bank offered my client $0.3 million to close the deal. I guessed a value of $6 million. Using a scientific calculator I produced a valuation of $6.55 million. Developing and testing a simulation model, I refined the answer to $6.42 million. No high technology was needed to realise that my client’s bank was attempting "ignorance arbitrage".

Innovation by banks in new financial instruments is necessary as it is in any industry to avoid being left behind in the competitive marketplace. And there will be a number of genuinely useful products that meet customers’ needs. These needs may be fiscal or regulatory; or the packaged return or cost of a bond issue or loan plus a structured swap may be more favourable; or because they have been engineered to generate a specifically required exposure; or because they neatly hedge an asset or liability exposure. But if the bank selling the products or corporation buying these instruments does not understand why they are entering into such transactions and how to measure and control the risks generated by them, it would be better to steer well clear of them.

He who asks is a fool for five minutes, but he who does not ask remains a fool forever - Chinese Proverb

Let me update an old story. I heard that in view of the emerging markets financial crisis the dealing room of a major bank is being down-sized. There will be a total crew of three: a computer, a man, and a dog. "What is the dog going to be dealing?" I asked. "Don’t be a fool" I was told. "The dog is the back office. He is there to make sure nobody touches the computer". "And the man?" I asked. "Stupid question. He is the middle office. His job is to feed the dog."

Corporate presidents should follow the mediaeval emperors. They should have faith in fools.

Let us be thankful for the fools. But for them the rest of us could not succeed. -  Mark Twain

"Statistics prove that more English and Spanish footballers should play football abroad."

Profile History Social & Charitable

Delphi Risk Management: Delphi Delphi & Delphi are the Innovation, Communication & Risk Management arms of Delphi Risk Management Limited 

If you have reached this page directly from a search engine visit Delphi's