Insurance is partly a market in securities, partly a mechanism for collective action; a means by which risks are traded, but also a means by which risks are socialised.
The French economist and businessman Michel Albert described two historical traditions from which the modern insurance industry originated. In London, English gentlemen would gather in Lloyd’s coffee house to speculate on the state of the world and the weather. They would gamble on the fate of ships. Meanwhile, in Swiss mountain villages, farmers might agree that if one of their cows died, the whole community would contribute to provide a replacement.
Although this is a caricature, it contains a large element of truth. Insurance is partly a market in securities, partly a mechanism for collective action; a means by which risks are traded, but also a means by which risks are socialised. Even today, mutual organisation is much more common in insurance than in most business activities, especially in continental Europe: the recent decline of mutuality is linked to the global spread of Anglo-American financial institutions.
The 18th century saw a degree of convergence between the traditions of market and mutuality. The development of the theory of probability and the construction of mortality tables were critical. These innovations allowed a more scientific approach to risk assessment; the actuarial profession came into being. Life assurance could be based on objectively calculated premiums.
Yet it was still amusing, and for some profitable, to gamble on the life of particular individuals. King George III, who reigned for 60 years despite his uncertain health, was a favourite subject. A book was established on the fate of Admiral John Byng, the hapless British seaman who failed to relieve Minorca and was controversially shot, as Voltaire put it, “pour encourager les autres”. But such wagers came to be seen not only as distasteful, but also dangerous. If large sums were riding on the death of Sir Hugh Walpole, someone might be tempted to hasten the outcome. Legislation established the concept of “insurable interest”. One could take out a contract of insurance only if one would suffer identifiable loss from the occurrence of the insured event.
Life assurance today is a more respectable business (although if you want to bet on the health of Kirk Douglas or Pope Benedict, you can do so online). But the underlying issue is topical again: the subject is not the death of individuals but the death of corporations. A credit default swap is, in a sense, an insurance policy on a company.
The growth of the market for credit default swaps after 1997 relies on a legal opinion by Robin Potts QC. In Mr Potts’ view of English law, such contracts are neither insurance (in which case purchases by traders who did not hold the relevant debt would have been illegal) nor gambling (in which case the contracts would, at least until the law changed in 2005, have been unenforceable).
Whatever the legal authority for Mr Potts’ view, it makes little commercial or economic sense. If someone who buys a CDS is neither insuring – protecting himself against possible losses from the borrower’s default – nor wagering – judging that the probability of default is greater than the odds implied by the market rate for a CDS contract – then what is the nature of the transaction?
Yet traders trade for no better reason than that is what traders do. It is difficult to believe that those who buy CDS protection on US Treasuries really imagine that the rate accurately reflects the probability of anything. They would have to believe not simply that the American government will default, but that when it does walk away from its liabilities, their counterparty will pay. Like the people who bought dotcom stocks or Florida property developments off plan, the purchasers are trading a security of uncertain and probably negligible fundamental value in the hope that they will be able to sell it on to someone else at a higher price.
Such asset bubbles generally collapse, although there are a few commodities – such as gold and Old Master paintings – for which dissociation of price and value seems permanent, with prices influenced only by sentiment and expectations about future prices. Holding these items is part insurance – the buyer hopes the asset may be worth something even in an apocalypse – and partly a means of attesting to one’s great wealth.
But it strains language to breaking point to describe CDS transactions as anything but gambling. The traders in AIG’s financial products division were inheritors of the amusements of Edward Lloyd’s coffee shop rather than the values of Swiss farmers. With short sales of equities, the need to borrow stock limits positions to the extent of insurable interest. But not in the CDS market. This observation leads directly to the increasingly widely held view – most recently expressed in this paper by my colleague Wolfgang Münchau – that CDS transactions should be permitted if they are insurance but not if they are gambling. The argument is in essence identical to that which led to the establishment of the doctrine of insurable interest in 1745: “It hath been found by experience, that the making of insurances, interest or no interest, or without further proof of interest than the policy, hath been productive of many pernicious practices.” Many things have changed since the 18th century, but many others have remained the same.