When insurance began to develop as an industry, it was soon felt necessary by those trying to enhance the reputation and respectability of the industry to distinguish it from gambling. The outcome of this process is that today, for a financial activity to classify as insurance and to be regulated as insurance, it has to offer products or contracts that protect against loss; gambling seeks or creates opportunities for speculative gain. More precisely, insurance hedges an open position in order to reduce exposure to risk; gambling creates or increases open positions to boost exposure to risk. There are a host of deep issues here, such as ‘what is the right metric for risk’ or ‘the risk to what: financial wealth, consumption, utility’? I will acknowledge these deep issues, ignore them and proceed.
For the insurance industry, the insurance vs gambling distinction was operationalised using the concept of insurable interest. An insurable interest is what economists would call an open position that is reduced in size by the insurance contract. In life insurance, this means that a person or a legal entity can insure the life of a third party only if the value of the life to the party wishing to purchase the insurance is greater than the value of the payout under the life insurance policy.
In property insurance, people have an insurable interest in property they own up to the value of the property, but not beyond that.
The obvious reason for limiting my capacity to take out insurance on the life of a complete stranger (whose life presumably has little intrinsic value for me) is moral hazard – what I will call micro-level endogenous risk. If the stranger’s life is insured for a sufficiently large amount, and if I can overcome the internal resistance of conscience and ‘thou shalt not kill’, I could arrange to have the highly insured stranger bumped off. When the probability of the insured-against contingency occurring is not exogenous, but can be influenced by the party purchasing the insurance, and if this cannot be verified and deterred by the party selling the insurance or by the forces of law and other and of contract enforcement, we have moral hazard.
In the case of life insurance and property insurance, the scope for moral hazard is obvious. In other financial markets that sell products that can be used either for insurance (covering or reducing open positions) or for gambling (increasing the size of an open position), similar micro-level endogenous risk or moral hazard may be found. In addition, macro-level endogenous risk can be created through the interaction of many individually insignificant agents, unless the principle of an insurable interest is extended to contingent claims in general.
Imposing an ‘insurable risk’ requirement for all derivatives
The financial sector would be decimated by the across-the-board imposition – for all contingent claims – of the requirement that an insurable risk must be present for a person or legal entity to buy or write a contingent claim. Hernando de Soto, the Peruvian economist who has done more than anyone else to encourage the conversion of unproductive informal possessions into productive capital, by registering and recording clear title to what had hitherto been informal sector possessions, estimates that there may be about $170 trillion worth of (private and public) bonds and equity and other traditional forms of credit in the world, and at at least $600 trillion (possibly as much as $1000 trillion) worth of derivatives (and yes, I know that bonds and equity can themselves be viewed as contingent claims – derivatives – also).
At the peak of the CDS market in mid-2007, there was at least $60 trillion of CDS outstanding (these are notional gross values, of course; the notional net value of all CDS is, of course, zero, as with all derivatives). The underlying bonds (against whose default the CDS provided insurance or on whose default the CDS permitted bets to be taken) were a small fraction of that $60 trillion – I have never been able to get anyone to come up with a hard number about the underlying notional value of the bonds. It is certainly likely that notional outstanding stocks of derivatives and the trading volumes would fall to a quarter or less of their 2007 levels if derivatives could be used only to buy insurance, not to place bets.
The application of the insurable interest principle to the CDS market would bar investors from purchasing default protection via credit default swaps on corporations without owning the underlying bonds. It would not be necessary to make ‘naked CDS purchases’ illegal, I believe. Just making a CDS contract unenforceable in court unless the claimant had ownership of the right amount of the underlying bonds, would suffice to discourage the trade.
The application of the insurable interest principle to short selling equity is straightforward. It certainly would rule out naked short selling – selling shares that you neither own nor have borrowed. Whether short selling stock you have borrowed qualified as an insurable interest is an interesting issue, which has would seem to depend on whether the contract under which the stock is borrowed, transfers ownership (temporarily) to the borrower.
Another application would be to equity puts (the right to sell a share of common stock at a given price at or before a given date). Requiring an insurable interest here would mean that you can buy an equity put only if you already own the equity, that is, no naked equity puts, and that you would have to sell the equity put if you sold the equity. You could hold the equity without the put, but not the put without the equity. Spread betting also would be illegal (or its contracts would be unenforceable) unless the party betting on an asset price fall either owned the right amount of the asset (or, in the case of insurable interest ‘lite’, had borrowed the asset).
Macro-endogenous risk as a reason for requiring an insurable interest
For any contingent claim it is possible to define and verify an insurable interest. But why would you want to ban the discretionary taking of open speculative positions? The first reason is our old friend moral hazard – or micro-endogenous risk. Financial markets can be manipulated. Employees of a now-defunct investment bank met in an Icedlandic bar/restaurant to discuss whether to attack the Icelandic currency, the stock of its banks and its debt, by using and abusing the CDS markets. Market abuse and manipulation through trash-and-trade strategies occur even in the deepest and most liquid financial markets.
But more important than moral hazard, which refers to conscious, deliberate individual behaviour, is macro-endogenous risk. Allowing naked CDS trading resulted in the creation of a massive gambling opportunity – a lottery of unprecedented size. This lottery worked through tickets that were traded in financial markets. Like all financial markets, the CDS market is inherently unstable. Even if all other conditions for market efficiency are satisfied as regards micro-market structure, individual rationality etc. , there is no ‘auctioneer at the end of time’, who ensures that the market cannot be corrupted, as regards its impact on resource allocation, by (rational) speculative bubbles. Furthermore, herding instincts, fads and fashions, mood swings from euphoria and irrational exuberance to fear and depression and irrational despondency, leading to market illiquidity, distressed asset sales at rock-bottom prices, are defining features of real-world financial markets. Some present and former colleagues of mine at the LSE are working on formal, rigorous models of macro-endogenous risk pathologies (see Jon Danielsson, Hyun Song Shin, Jean-Pierre Zigrand (1) and (2, for the young at heart)).
What the creation of markets for (naked) CDS and other derivatives has done is create a large number of new opportunities for taking additional risk. The lotteries or bets that are the essence of contingent claims/derivatives markets could increase allocative efficiency if they permitted the given, exogenous risk in the economy to be born by those most able to bear it. Instead that risk has ended up with those most willing but not, judging by results, most able to bear it. Opening a few additional financial markets (relative to the infinite number it would take to create a complete set of contingent claims markets), and doing so without any long-run (expectations) co-ordinating device, can create additional endogenous risk and uncertainty. It creates more opportunities for going bankrupt. Defaults and fear of defaults (triggered by market participants who were able to take open, speculative positions way in excess of what they could have achieved without these partial but inherently insufficient attempts to create additional markets for trading risk over time), are a source of macro-endogenous risk, even if no individual trader has market power or attempts to manipulate markets. Bets taken in the CDS markets, through naked equity puts, through naked equity shorting or through spread betting cause massive redistributions of wealth and income that can destroy real resources, influence the prices of ‘outside’ assets and bring down otherwise viable economic entities. And to this one has to add the non-negligible amount of resources (wages, profits and rents) absorbed by the firms that manage these betting shops.
Conclusion: time for Shari’a standards?
The principle that you cannot sell what you do not own (which also means you cannot sell debt) is a cornerstone of traditional Islamic finance, or Shari’a law-based finance (see e.g. Shari’a Standards for Islamic Financial Institutions; 1429 H -2008, Accounting and Auditing Organisation for Islamic Financial Institutions, Bahrain). During the height of the financial craze that swept the world prior to August 2007, there was some erosion of this key principle in the rulings of a few of the trendier Shari’a scholars who were willing, for a fee, to sign off on products as Shari’a-compliant, even though these instruments were substantially equivalent to standard ‘naked’ derivatives and indeed often were barely disguised clones. But one assumes that a return to basics is likely now also in Islamic finance, which never strayed that far from its origins.
The traditional Islamic opposition to trading risk without there being an insurable risk – to gambling, that is – extends well beyond the financial sphere. Indeed, in much of traditional Judaism and Christianity as well, gambling and betting are viewed as signs of moral weakness – as sins. But these moral, ethical and medical (gambling as an addiction) arguments are quite separate from the argument I have explored in this post, that risk trading without an insurable interest may be economically inefficient and destructive, not (just) for familiar moral hazard or micro-endogenous risk reasons but for macro-endogenous risk reasons.
Requiring an insurable interest to be present for risk trading to be allowed (or for such contracts to be legally enforceable) means interfering with the right of two or more parties to enter freely into contracts they all understand. At the very least it amounts to discrimination against such ‘naked’ trades by not enforcing them in a court of law. There are paternalistic arguments against gambling and betting, especially if it is addictive and can hurt or even ruin dependents. And there are conventional utilitarian efficiency arguments for interfering with the freedom to contract if there are significant negative externalities (such as the creation of macro-endogenous risk). It is less obvious that a ‘rights-based’ significant negative externalities argument can be made for interference with the freedom to contract when there is no insurable risk.
At the very least, we should have a serious debate as to whether ‘naked’ derivative trading should be declared haram everywhere.