What stands out prominently in the analysis of the ongoing financial crisis is the criticality of interconnections among balance sheets in the deregulated financial system. The network structure of a web of claims and counterclaims among investment banks is what has made many of them ?too interconnected to fail?. It is, after all, an effect of the network that when one of them is under threat, financial intermediaries fear to do business with each of the others, generating a liquidity crunch instantaneously.
Bipartisan discussions through this last weekend in the US Congress demonstrate how expensive it is to build back confidence up to a level where the system has any chance of unfreezing at all. It is clear that the nature and degree of connectivity in the network structure of the financial system should be a central concern of financial stability policy.
Analysis of networks is a fast growing area in academic research in economics. There is a useful amount now known about surprising economic and social implications that follow from the precise structures of ?social? networks in which we are embedded. Somewhat surprisingly, applications of network analysis to financial systems are still in relative infancy. Such research suggests that in a highly connected system the likelihood of widespread default might be lower, but the potential impact would be higher. As might be expected, the losses of a failing bank can be more widely dispersed among others, but greater connectivity increases the chances that a bank that survives the effects of an initial default will be exposed again to other defaulting counterparties in successive rounds.
The network structure of the financial system generates what economists call an externality. In current circumstances, the overall cost to the economy from the distress of individual banks is much higher than the sum of the individual costs. The financial network evolves organically. The growth of overall connectivity of the system is beyond the control of any one bank. The responsibility for monitoring the network and factoring in implications into policy cannot rest with anyone other than financial regulatory agencies and the central bank.
If investment is all that investment banks do, any losses made by a bank are not a direct threat to the financial system. The problem arises when the investment arm of the bank is bundled up with the market making part of its business. This is the part that enters into bilateral contracts based on derivatives. These contracts used to enable all sorts of risks to be hedged, and the larger part of their value lay in finessing the contracts terms (principal, maturity and various contingencies) individually for the buyer. This is the part of the business that interconnects balance sheets. If (because of losses on its investments) a bank is unable to fulfil its obligations as a counterparty in the contracts it has entered into, the threat of triggering a cascade of defaults becomes a threat to financial stability and to confidence in the entire system.
Policy agencies can and do monitor the level of capitalisation of the banks from a financial stability point of view, and they monitor the degree of concentration of banking business, from a competition policy point of view. However, it is not clear that in the US or the UK they were monitoring the network structure of the deregulated anglo-american financial system. Indeed, it is not clear that there is any practical way in which they would have been able to track the evolving network as connections sprouted and spread through heterogeneous and sophisticated credit backed contracts.
Going forward, the proposal that OTC contracts be subject to exchange trading with the exchange serving as the counterparty to all traders (Ajay Shah, 27 September) has much to recommend it. This will of course mean that only a few types of standardised, and hence more transparent contracts, will exist.
?The author is reader in economics at Cambridge University?s Judge Business School