VoxEU Column Financial Markets

Let banks be banks, let investors be investors

Simplicity and transparency, two major causalities of recent financial market changes, are essential to restoring trust in financial markets. This column suggests that distinguishing two types of financial intermediaries – client servicers and capital managers – would be a big step in the right direction. Today’s lack of distinction means one set of regulations is applied to the two very different functions.

Simplicity and transparency – two vital ingredients of trust in financial dealings – are casualties of financial market changes in the recent years. To restart the financial system and the world economy we need to re-establish them, a point acknowledged in the recent G20 communiqué.

Restoring simplicity and transparency is a task for world governments. I propose a straightforward method for doing so. Here is the sound-bite version: let banks be banks, let investors be investors.

Essential distinctions among financial intermediaries

Intermediaries in financial markets (and in particular, in securities markets) necessarily fall into one of two categories:

  • Client servicers
  • Capital managers.

The client business is defined by proximity to clients. It consists of selling access to markets (primary, i.e. capital raising, and secondary, i.e. brokerage), and all its ancillary activities such as research and advice (on capital raising, mergers and acquisitions). The client business also includes supplying derivative contracts tailored to client needs, so these are hedged in the books of the client servicer. Finally, client business includes borrowing and lending cash, as well as borrowing and lending securities performed only to fulfil client requests.

The capital management business is completely different. The simple and robust criterion for distinguishing capital management from client servicing is that the franchise of the capital management business is not counterparties, but shareholders. The client business gains in value if it provides good service to its counterparties. The capital management business has no special regard for counterparties; it gains in value if it provides good investment returns to shareholders. Capital managers may invest their shareholders’ moneys on a leveraged or un-leveraged basis and provide exposure to the whole spectrum of financial risks. While their transactions may be the same as those of the intermediaries in the client business and therefore their balance sheets may look very similar, their objectives and their risk profile is completely different.

A dangerous muddling of functions

What is wrong with the financial system now? That there is no clear distinction between these two functions, and as a result regulations and functions have grown too far apart.

Banks and investment banks mix client business and capital management business – despite the fact that the proximity of these two businesses gives rise to very serious conflicts of interest. In addition, the capital management business within banks and investment banks is not transparent (who knows how much money comes from proprietary investments?), and is not subject to efficient capital constraints.

Consider this simple example: a bank can hold risk-weighted assets up to 12.5 times its capital (the capital requirement is 8% of risk weighted assets, equal 1/12.5). Is this a constraint appropriate to the low-risk world of client business (which, as explained above, by definition systematically hedges away financial risk) or the potentially much higher risk of investment management? In the past year or so, the largest dollar losses caused by the financial turmoil have come from banks. By definition, these large losses imply that the banks that have incurred in them have taken excessive risks. The hypothesis that an inappropriate regulatory framework led these institutions to take on excessive risks is by now the most credible one.

Capital management is also performed by asset managers and hedge funds (in addition to the banks mentioned above). But hedge funds are institutions that are subject to very different rules than asset managers, despite the fact that they do the same business. The very fact that most hedge funds are corporations based in offshore centres illustrates the lack of a satisfactory institutional setup for an industry that in recent years has become the most dynamic part of the financial system.

How do we let banks be banks and let investors be investors?

The answer, in a nutshell, is to create a regulatory regime (or better, adapting current regulations) that clearly distinguishes between the two types of intermediaries – client businesses and investors – and provides a consistent set of rules for each.

This idea may look similar to recent proposals to re-instate Glass-Steagall legislation prohibiting banks from dealing in securities or to return to narrow banking (see Paul De Grauwe 2008). It is in fact quite different.

Consider Glass-Steagall first. It can be argued conceptually and it has been shown empirically that the presumed conflicts of interest between lending and capital markets (issuance of bonds) activities are insignificant.1 In a financial system where securities have become, appropriately, a dominant medium of transactions, it makes little sense to separate out securities activities from the rest of the activities of client businesses. Proponents of narrow banking, by contrast, argue for ways to eliminate liquidity transformation.2 I do not think this should be an objective of a well-working financial system, because liquidity transformation is a socially productive activity and, as the recent experience has shown, liquidity risk never goes away from securities markets. There are, however, ways to mitigate and control liquidity risk, as I discuss below. In conclusion, banks and broker dealers should only be that; they should not be investment managers.

Consider now the capital management business. The second fundamental pillar of a well-working financial system is a consistent institutional framework for capital managers. All capital (or investment) managers, mutual funds and hedge funds, should be subject to the same rules (which of course recognize the different degrees of risk of different asset classes and investment styles). Thus, all hedge funds should be onshore, and subject to the appropriate regulations (like other asset managers, prudential rules and investor protection rules). Such rules are not expected to prevent professional hedge fund managers to carry out their business as they have done so far. My impression is that those who claim that hedge funds were much less hit in their investments by market turmoil because they were offshore and unregulated are overstating the potential perverse incentives caused by investment management rules.

Dealing with financial instabilities

It is apparent that both categories of intermediaries, client businesses and capital managers, may be subject to financial instabilities that are caused by liquidity mismatches and can be aggravated by leverage. Banks are subject to runs. Even money market mutual funds can be subject to runs (the recent $540 billion bailout of US money-market funds is a good example) and, a fortiori, those hedge funds that use leverage and invest in relatively illiquid securities. These events may spread in the financial industry, causing systemic risks.

The avoidance of systemic risk has to come from the combination of three things:

  • Appropriate capital structures for every type of intermediary;
  • Full knowledge of intermediaries by public institutions responsible for financial stability; and
  • Appropriate ex-post intervention by authorities.

The fact that an appropriate capital structure is the first defence against runs is almost a tautology. The choice of appropriate capital structures does not need to be mandated, though regulations could for example allow different investors to access investment managers depending on their business and the safeguards provided by their capital structure.

We should support the creation of a new and comprehensive system of disclosure of all risk positions, on- and off-balance sheet, to authorities in charge of market stability. This information should allow them to aggregate and analyze the macro issues and identify the areas of weakness, as well as to communicate to individual institutions their fragilities.

While this is a big and complex task, it should be greatly facilitated by the clear separation of the different classes of intermediaries described above. Does this not look like a simpler financial system than what we have today? 


Paul De Grauwe (2008). “Returning to Narrow Banking”, in B. Eichengreen and R. Baldwin, What G20 Leaders Must Do to Stabilise the Economy and Fix the Financial System, a VoxEU.org Publication, 2008. 

Randall S. Kroszner and Raghuram G. Rajan (1994). “Is the Glass-Steagall Act Justified? A Study of the U.S. Experience with Universal Banking before 1933”, American Economic Review, American Economic Association, vol. 84(4), pages 810-32, September.


1. Proponents of the separation of lending and securities businesses were inspired by the hypothesis that lenders may offload unwanted credit risks by issuing bonds by the same borrower to the public. The many defaults that occurred in the early 1930s were at the origin of that hypothesis and of the legislative initiatives in the US. However, the hypothesis does not take into account the fact that capital markets activities are repeated, and that selling bonds that subsequently default causes large damage to such activities. In addition, for empirical analysis showing that in the early 1930s United States broker-dealers’-issued corporate bonds were not significantly less risky than banks’-issued corporate bonds, See Kroszner and Rajan (1994).

2. For example, De Grauwe (2008) suggests that matching the duration of assets and liabilities of banks would amount to a return to narrow banking. While this rule would eliminate interest-rate risk from banks balance sheet, it would not eliminate liquidity risk, which is more specific to the nature of the assets and liabilities held. In other words, it is not the sensitivity of assets’ and liabilities’ values to interest-rate movements that determines their liquidity. 

105 Reads