(The blog post is cowritten with Rabee Tourky of The University of Queensland.)
Recent events have demonstrated that the financial sector is exposed to systemic risk. That is, the possibility that many financial institutions fail at the same time. The concern is that should this negative outcome realise, there will be a dramatic fall in liquidity that will impact on the real economy.
There are many reactions to this exposure. We have seen the moves by the government to guarantee various forms of lending. We have seen direct forms of intervention to sell securities (residential back mortgages, foreign currency, etc). And we have seen unplanned reductions in interest rates. While all of these moves seem justified given the crisis mode we find ourselves in, it is hard not to be concerned that we are exposed to another risk: that some poor government policy choices might find themselves thrown into the mix.
One such area of policy exposure is our attitude towards bank mergers. Already we have seen the current crisis has been put forward as a motivation behind the proposed take-over of BankWest. But there is also a risk that mergers may become the preferred policy prescription to deal with financial institution difficulty or failure.
We need to be extremely cautious and sceptical about such moves. Mergers, in our already highly concentrated banking sector, will reduce competition pressures – particularly, in certain regional markets and loan segments. Any other reasons will have to outweigh these.
Some have argued that bank mergers reduce systemic risks. The argument rests on the notion that a larger bank holds a more diverse set of assets. However, it is not clear that this argument entirely convincing.
Let’s think through the issue of systemic risk carefully. It is reasonable to suppose that there are ways of insuring against the possibility of one independent financial firm defaulting. However, insuring against endemic default requires entirely different approaches and is something that is probably difficult to anticipate and analyse. Classical insurance theory tells us that we can mitigate risk if a large number of agents with exposure to independent risk get together and pool their risk. A problem that is associated with the kind of system-wide collapse that we have recently seen elsewhere is when financial institutions pooling their risk are exposed to correlated risk. This correlation appears in at least two ways, first the kind of assets that financial institutions hold and second the potential for contagion.
The idea behind prudential regulation of the banking system is that governments should distinguish banks from non-banking financial institutions and that they ought to regulate banks. The traditional situation is that such regulation takes the form of rule of thumb methods including capital requirements. The main idea is to reduce the probability that a given individual bank fails. However, it seems that prudential regulation does not explicitly address the likely correlated exposure to risk that financial institutions face. So while a bank might be diversified internally, if all other banks hold a similar set of assets, there is no inter-bank diversity and there is exposure to systemic risk.
Therefore, it is quite clear that prudential regulation should take into account exposure to systemic risk arising from a lack of inter-bank diversity and correlated intra-bank diversification. This is an argument that prudential regulators should actively encourage inter-bank diversification and provide incentives and adjust their rules accordingly. Capital adequacy requirements do not address this problem because making all banks hold bonds does not reduce the asset holding correlation amongst them.
This logic has direct implications in regard to our attitudes towards mergers from a prudential perspective. A merger between two banks may not lead to the merged bank holding a portfolio of assets with less correlated returns; inter or even intra correlation. That depends on their portfolio and the portfolios of the remaining banks. Moreover, given that a merger is not necessarily taking place in the absence of options, the acquirer is most likely to prefer to merge with a bank that is a closer competitor.
Systemic risk arises when the returns of assets held by banks are highly correlated. A merger need not change this fact. The assets are there, they are combined and so the total portfolio of assets held by the banking sector does not change.
In this situation, a merger between a distressed institution and another may even increase systemic risk. What it does is force the bank with greater liquidity to direct funds to the distressed institution and cover its liabilities. It is no different than what a central bank might do by injecting liquidity into the distressed institution. It is just that this way, it is off the balance sheet of the central bank and the government and on to the backs of consumers who face a lack of competition in the future.
It appears to us that to reduce systemic risk one would need a large number of financial institutions pursuing diverse strategies in the market place. Whether it is in the types of customers that it lends to, different emphasises in production sectors or geographic diversity, these have the potential to reduce systemic risk. When it comes down to it our four pillars policy looks very much like a four clones policy. And that is our concern.
We believe that our financial authorities and the Australian Competition and Consumer Commission should examine very closely proposals for unsolicited mergers between financial institutions on the basis of some reduction in systemic risk. Moreover, care should taken in dealing with any distressed or failing institutions by a forced marriage as this may only move the mortgages around while at the same time mortgaging our competitive future.