Paul De Grauwe
University of Leuven
Should we return to narrow banking? That is, should we restrict commercial banks’ activities to traditional consumer and corporate loans held on the banks’ balance sheets until maturity, allowing only investment banks to engage in securitisation and other sophisticated investments? In order to answer these questions it is good to go back to the basics of banking.
There are two characteristics of banks that make them very special. First, banks borrow short and lend long. In doing so they create a very special kind of risk, a liquidity risk. This risk is special because it is a “tail risk”. It occurs infrequently, but when it does, it has devastating effects. It is difficult, if not impossible to quantify because it occurs as a result of collective movements of distrust and panic. We have no reliable model predicting collective panics.
Second, banks are at the center of the payments system. This creates a network of borrowing and lending by banks from and to each other. This interbank market has the effect of creating an interconnectedness of risk, i.e. when one bank fails, the other banks are in trouble. When a problem appears in one bank, it is propagated into the whole banking system like an infectious disease. Thus, risks in the banking system are likely to be correlated. This contrasts with what happens in the non-banking sector. For example, when one automobile company goes bankrupt, this is good news for the other automobile companies. The latter can expand their production and increase their profits.
Securitisation, which became popular from the 1980s onwards, was thought to reduce systemic risk because it would spread the risk concentrated in one bank over many more institutions. In fact, it did the opposite; it increased systemic risk. When a bank in the Midwest bundled mortgages into an asset backed security and passed it on to financial institutions in, say, Germany, the interconnectedness and the correlation of risks in the banking system increased. Thus, securitisation amplified the inherent problem of the banking system which is that shocks occurring in one place are quickly transmitted to the rest of the system. As a result, securitisation increased the inherent fragility of the banking system.
How to tackle this problem? There are essentially two ways. In the first one, banks maintain a business model that allows them to securitise their loans in different sophisticated ways, although subject to tighter regulation and supervision than before. This is the approach towards which policymakers gravitate today. It is based on a belief that the risks created by banks can be managed and contained by imposing a series of appropriate capital and liquidity ratios, and, as has been proposed in a number of countries, by subjecting the introduction of new financial products to prior approval.
The problem of this approach is that we do not have a science of interconnected risk, which is the risk created by banks. All we have is a science of independent risks. This is the risk that arises in one place and that can be isolated, because it is not propagated to the rest of the system. This science of independent risks which has been very powerful in developing and pricing derivatives and other sophisticated financial products, is useless as a tool to manage the risks created in the banking system. It is not only useless. It is also dangerous because it will lead to a new complacency. When the new regulatory environment will be in place it will create the illusion that things are under control, while underneath the time bomb of correlated risks that can be triggered by collective movements of panic will continue to tick.
The second approach starts from the admission that we do not have the tools to quantify the risks created by the banking system. All we can do is to limit these risks by restricting the activities of banks. This is the idea behind narrow banking, which was the core principle of the Glass-Steagall Act introduced in the USA after the Great Depression and similar legislation in many European countries. In this approach commercial banks, i.e. those that hold the deposits of ordinary customers, are told that activities that increase the interconnectedness of risk shall not be allowed. Since securitisation of loans increases this interconnectedness, it will not be allowed. Thus, narrow banking aims at minimising the potential of the banking system to create correlated risks. These risks of course cannot be eliminated, but they can certainly be reduced.
An objection to the idea of narrow banking is that it will reduce credit and thus will negatively affect economic growth. The answer to this objection is that securitisation and financial innovations have led to an explosion of bank credit that has turned out to be unsustainable. It has led to consumption and housing booms that increased economic growth temporarily. The economic growth observed in the US, the UK, and many other countries during the decade prior to the crash was artificially high, sustained as it was by excessive credit made possible by securitisation. After the crash, economic growth will have to return to a lower and more sustainable level, especially in these countries that have experienced these artificially high growth rates. Such a lower but more sustainable growth rates can be achieved in an environment in which banks create fewer risks. It will also be a banking system which generates fewer profits.
Financial innovation will still occur in such a new banking landscape. Investment banks will still be able to develop new sophisticated assets. The limitation they will face, however, is that they will not be able to finance these (most often illiquid) assets by short-term funding. In other words investment banks will not at the same time be able to operate as commercial banks, as unfortunately some can today. In addition, these investment banks will not be bailed out by the state if they fail.
Bankers and their many lobbyists scream that narrow banking is terrible and will reduce innovation and growth. Don’t believe them! In fact prior to securitisation economic growth was higher in the industrialised countries. The protests of bankers against narrow banking are self-serving. When bankers make pleas to keep their business models unchanged they are not concerned at all with general welfare. Their only concern is to go back to the situation prior to the crisis that allowed them to generate high profits while making sure that part of the risk was borne by the rest of society.