Volume 46, November/December 2011, Number 6 · pp. 298-299
On the Necessity of Separating Investment and Commercial Banking
Perhaps once or twice in a century secular trends in innovation and structural change collide with institutional arrangements and regulations, creating conflicts in policy objectives and market volatility that are capable of bringing down the global financial system. While the current crisis is global in nature, Europe has its own special brand of institutional arrangements that are being tested in the extreme: a monetary union amongst countries with very different structures that are subject to asymmetric real shocks, and the presence of a banking system that permits the co-mingling of capital markets and OTC derivative products with traditional banking on a large scale. The inconsistencies of the monetary union have resulted in a sovereign crisis which is interacting with the structure of banking through financial price volatility in a most dangerous manner. With respect to banks, policy has allowed too much leverage and has so far failed to take seriously the need to separate retail from investment bank (IB) activities.
At its core, the cause of the crisis has been the under-pricing of risk. Excessive risk in banking can always be traced to two basic causes: first, to too much leverage, and second - for given leverage - to increased dealing in high-risk products. By having nothing to say about the ratio of risk-weighted assets to total assets, the Basel Tier 1 rule regulates very little at all. Systemically important banks are permitted to use their own internal models and derivatives to alter the very risk characteristics of assets to which the capital weighting rules apply. The Basel rule as constructed - and so widely supported by the banks - cannot control the two forms of risk at the same time.
The OECD has consistently argued from the outset of the crisis that the two key reform requirements are the introduction of a leverage ratio to control the first form of risk and the separation of IBs from traditional retail and commercial banks (specifically, via a non-operating holding company structure - NOHC - with ring-fencing) to deal with the second. A similar proposal on separation has recently been proposed by the UK Vickers Report. To reduce periodic solvency concerns and liquidity crises like those currently pressuring banks in Europe in 2011, a consistent policy of separation also needs to be introduced on the continent.
In the past two decades financial deregulation and the more widespread use of securitisation, OTC derivatives and repo financing have facilitated a move to what finance specialists refer to as "complete markets" in all bank products. Bank business models have changed to exploit opportunities for fees and for regulatory and tax arbitrage. This move away from "traditional banking" to a form of "capital markets banking" was associated with an explosion of leverage and the greater mixing of mark-to-market products (usually associated with securities firms) with retail and traditional commercial banking assets and liabilities. Standalone IBs were subsidised by their favourable treatment under Basel I and II in their dealings with other banks. Holding companies that owned IBs and universal banks were all direct beneficiaries of the boom in new instruments through their securities dealing, prime broking and OTC derivatives businesses as regulations became even more lax. In the USA the removal of Glass-Steagall opened the way for contagion between IBs and traditional banking in this new world.
In Europe, it is often argued that since Glass-Steagall did not apply and there had been no great difficulties until recent years, then there should be no problem with the universal banking model as such. This is exactly the sort of argumentation that does not recognise the nature of the secular changes and the new environment for banking. In the days of incomplete markets, the universal bank model was much less dangerous and Glass-Steagall-like rules were significantly less necessary than is the case now. Internal contagion between fair value through profit or loss (mark-to-market) versus amortised cost accounting (traditional) products is now more material, and interconnectedness risk through derivative counterparties has risen to new levels.
It may surprise some policymakers to know that in 1998 the notional value of derivatives globally was about two and a half times world GDP, and on the eve of the crisis it had risen to a staggering twelve times world GDP. It is about ten times now. Bankers like to claim that innovation supports growth and derivatives are essential for risk control. Yet this very same period is the worst period of economic growth in the post-War period, and it encompasses the worst financial crisis since the Great Depression. It may also surprise some to know that when the US government chose to settle the AIG derivative exposures to avoid a global meltdown, the amounts involved for some large European banks with respect to one single counterparty were in the vicinity of 30-40% of their equity capital. Nowhere does one see in any bank publication just before the AIG crisis risk exposure reports approaching anything remotely like the amounts that were actually paid. Capital market banks never have much ex-ante risk produced by their models, but they certainly can have massive ex-post losses. It is precisely the fear of contagion and counterparty risk, and the funding problems to which these give rise, that are affecting bank credit default swap spreads in Europe right now.
The logic for separation is simple enough. A retail bank with a leverage ratio will be safe and boring - boring is "good" when it comes to the deposits of unsophisticated investors and "bad" when it comes to the profitability and bonus targets of bankers. With full separation, deposit insurance would apply only to the retail bank and could not act as an implicit subsidy to the IB, which would otherwise benefit from the too-big-to-fail (TBTF) guarantee. TBTF cross-subsidisation is one of the factors that led to the under-pricing of risk in the first place. The boring traditional bank could go on lending to households and to small and medium-sized enterprises (SMEs) regardless of what the IB in the high-risk end of the market was doing. The CEO of the separated IB, on the other hand, would have to earn his profits without the subsidy of TBTF and the socialisation of losses resulting from any bad decisions he or she made. The IB would become a securities firm that would be less levered and less risky than was the case for Lehman Brothers. It would be less likely to fail and, if it did, it could be more easily resolved without the disruptions seen in 2008 and 2009.
The securities firm under the new regime would see the cost of carrying out high-risk activities rise and their volume fall. Cross-subsidisation from the TBTF implicit guarantee would be removed, a leverage ratio would be in place for the financial group and there would be no favoured "bank-like" treatment for the separated securities firm under Basel. In short, the securities firm would be reliant only on its own (separated) capital provided by sophisticated investors: counterparties would demand tougher initial and variation margins and high quality collateral, and investors would expect to be compensated for the higher risks they would be undertaking. Far from being a criticism of the separation proposal which banks often like to imply - "It would raise the cost of doing business" - that is precisely what the policy objective for higher-risk capital markets banking should aim to do.
A full Glass-Steagall style breakup of the banks is unnecessary to achieve the aim of reducing risk. Under the OECD's NOHC structure, the parent would be "non-operating", raising capital on the stock exchange and investing it transparently in its operating subsidiaries - including the ring-fenced retail bank. The non-operating parent would have no legal basis to shift capital between affiliates in a crisis without consulting regulators, but technology platforms and back-office functions could still be shared, permitting synergies and economies of scale and scope. Such a transparent structure would make it easier for regulators and market players to see potential weaknesses and to resolve the IB if need be without interrupting the activities of the retail bank. The proposal would provide no impetus for deleveraging where it really matters, i.e. the lending to households and SMEs; quite the contrary. But it would certainly help to unwind much of the socially less useful mountain of OTC derivatives related to regulatory and tax arbitrage and the high-risk proprietary trading of securities firms linked to banks.
Adrian Blundell-Wignall, Deputy Director, Financial & Enterprise Affairs, OECD, and Special Advisor to the Secretary General for Financial Markets