The same old folly starts a new spiral of risk
Saturday, 25 August 2007
John Kay
THE financial economics I once taught treated risk as just another commodity. People bought and sold it in line with their varying preferences. The result, in the Panglossian world of efficient markets, was that risk was widely spread and held by those best able to bear it.
Real life led me to a different view. Risk markets are driven less by different tastes for risk than by differences in information and understanding. People who know a little of what they are doing pass risks to people who know less. Since ignorance is not evenly distributed, the result may be to concentrate risk rather than spread it. The truth began to dawn when I studied what happened at Lloyd's two decades ago.
The "LMX spiral" -- whereby Lloyd's syndicates reinsured each other rather than laying off risks elsewhere -- brought that venerable insurance market close to collapse. The reinsurance contract is almost as old as insurance itself. The insurer cedes part of the premium in return for an agreement to bear losses in excess of an agreed sum. The modern innovation was to reinsure not just a single contract but a package, or even the total losses of an underwriting syndicate or an investor. These losses might themselves include claims on similar policies. As such structures proliferated, it became increasingly hard to understand the nature of the underlying risk.
Participants found comfort in two ways, A conventional wisdom told them that these sophisticated arrangements spread risks widely across the market. Historical analysis showed that very little had ever been paid out on excess-of-loss policies.
The spiral began to unwind when Piper Alpha, a North Sea oil rig, caught fire in 1988: 167 lives were lost and the rig was destroyed. The initial charge to Lloyd's of about $1.0bn was one of the largest single insurance claims ever made. The claim triggered excess-of-loss policies, which triggered other policies: the total of claims across the market was about $16bn.
The most avid participants in excess-of-loss syndicates discovered that they had, in effect, insured Piper Alpha over and over and over again. This, and analogous events, meant that a few syndicates incurred horrendous losses. The contributions required put in jeopardy the stately homes of England and the Surrey mansions of the nouveaux riches who had been attracted by the social cachet of being a name at Lloyd's.
Greed mingled with self-delusion, honest incompetence with conscious deception: it was impossible to say which had caused the crisis. Lloyd's was, I came to realise, a microcosm of what was happening in other financial markets. If trading was motivated not by differences in attitudes and preferences but by differences in information and understanding, risk would gravitate not to those best able to bear it but to those least able to comprehend it.
The ever-expanding scale of business was not a measure of market efficiency, but of inefficiency. The volume of incestuous trading within the market itself far outpaced the volume of business with the outside world. The accumulation of fees as the same risk was passed round and round raised costs to levels the business would ultimately be unable and unwilling to support.
Under syndication arrangement with FE
THE financial economics I once taught treated risk as just another commodity. People bought and sold it in line with their varying preferences. The result, in the Panglossian world of efficient markets, was that risk was widely spread and held by those best able to bear it.
Real life led me to a different view. Risk markets are driven less by different tastes for risk than by differences in information and understanding. People who know a little of what they are doing pass risks to people who know less. Since ignorance is not evenly distributed, the result may be to concentrate risk rather than spread it. The truth began to dawn when I studied what happened at Lloyd's two decades ago.
The "LMX spiral" -- whereby Lloyd's syndicates reinsured each other rather than laying off risks elsewhere -- brought that venerable insurance market close to collapse. The reinsurance contract is almost as old as insurance itself. The insurer cedes part of the premium in return for an agreement to bear losses in excess of an agreed sum. The modern innovation was to reinsure not just a single contract but a package, or even the total losses of an underwriting syndicate or an investor. These losses might themselves include claims on similar policies. As such structures proliferated, it became increasingly hard to understand the nature of the underlying risk.
Participants found comfort in two ways, A conventional wisdom told them that these sophisticated arrangements spread risks widely across the market. Historical analysis showed that very little had ever been paid out on excess-of-loss policies.
The spiral began to unwind when Piper Alpha, a North Sea oil rig, caught fire in 1988: 167 lives were lost and the rig was destroyed. The initial charge to Lloyd's of about $1.0bn was one of the largest single insurance claims ever made. The claim triggered excess-of-loss policies, which triggered other policies: the total of claims across the market was about $16bn.
The most avid participants in excess-of-loss syndicates discovered that they had, in effect, insured Piper Alpha over and over and over again. This, and analogous events, meant that a few syndicates incurred horrendous losses. The contributions required put in jeopardy the stately homes of England and the Surrey mansions of the nouveaux riches who had been attracted by the social cachet of being a name at Lloyd's.
Greed mingled with self-delusion, honest incompetence with conscious deception: it was impossible to say which had caused the crisis. Lloyd's was, I came to realise, a microcosm of what was happening in other financial markets. If trading was motivated not by differences in attitudes and preferences but by differences in information and understanding, risk would gravitate not to those best able to bear it but to those least able to comprehend it.
The ever-expanding scale of business was not a measure of market efficiency, but of inefficiency. The volume of incestuous trading within the market itself far outpaced the volume of business with the outside world. The accumulation of fees as the same risk was passed round and round raised costs to levels the business would ultimately be unable and unwilling to support.
Under syndication arrangement with FE