Financial regulation: are we reaching an efficient outcome?

Armstrong, Angus ; Davis, E. Philip

This issue of the National Institute Economic Review includes articles by six renowned financial economists who each investigate one key aspect of the Global Financial Crisis (GFC) and the regulatory response. The authors, who will also present at the National Institute's Annual Finance Conference at the Bank of England in March, were asked to have in mind the following guidance in preparing their articles:

"Since the Global Financial Crisis a number of regulatory policies have been discussed, proposed and sometimes implemented to address the shortcomings in the regulatory framework. These include capital and liquidity regulation (notably via Basel III), developments in cross-border bank resolution, macro-prudential policies and addressing the issue of too-big-to-fail. It is timely to take a critical overview of these various measures to see whether we are closer to a financial system that is both appropriately stable and efficient in fulfilling its functions to the wider economy. Could better ways to address the underlying problems be conceived? And what are the open questions?"

Anat Admati focusses on capital regulation for banks. Equity holders and creditors have competing interests as the former benefit from the upside of risks while the latter share only the downside of risks. In non-bank firms equity holders have an incentive to increase risk, but this is offset by the rising cost of debt and use of covenants as debt holders seek to protect themselves. In banks this offset is weaker as depositors are insured so they have no incentive to monitor and price risk-taking. Since deposits are unsecured by collateral, banks can use the assets purchased with deposits as collateral for non-deposit debt funding at low cost. The motivation for capital regulation is to protect taxpayers, who insure the depositors from the consequences of these risk-taking incentives.

Admati suggests that the widely held view that 'holding equity is costly' results from a focus on the private costs to bankers and their shareholders of not being able to pass on risk to creditors and taxpayers in this way. This idea that equity is expensive is, she argues, widely believed owing to the political influence of bankers. In fact any such private costs are more than offset by the social benefits of financial stability. There is no fundamental reason why banks should be more highly leveraged than other corporations.

Admati contends that Basel III is a missed opportunity: capital requirements are still too low. She also criticises the use of risk weights (especially zero weights) which offer incentives to manipulate disclosure and maximise risk. Furthermore, she suggests that unreliable non-equity securities should not be counted as capital as bail-ins are unlikely in a crisis. Instead she recommends 20-30 per cent equity ratio with a transition financed by zero dividends. Meanwhile, the tax code should be amended to reduce the incentive of banks and other corporations to take on debt instead of equity.

David Miles contends that the key problem building up to the GFC was not 'light touch regulation' but banks operating under Basel II with very high leverage on risky assets often as a result of low risk weights. The acceptance of this financial structure was again the suggestion that 'equity is costly' to banks. In fact, the full social cost of low equity in terms of financial crises is substantial and the impact of high equity on bank funding is low. The use of debt-based instruments in total loss absorbing capital (TLAC) and the complexity of multiple capital buffers are shortcomings of Basel III. Miles is sceptical of the need for liquidity requirements in addition from adequate capital regulation as enough capital is usually a guarantee of liquidity. But the key point is that regulators are not requiring adequate capital ratios.

We agree that excessive leverage and distorted risk weights were central to the GFC. However, questions remain about how these distortions were permitted. For example, subordinated debt holders have an incentive to monitor risk, the argument that creditors knew that governments were a back-stop lacks direct evidence. Moreover, under Basel II regulators had supervisory discretion and powers to increase transparency to enable market discipline. It is an open question whether regulators can ever limit risk taking by rules within such complex institutions. A fundamental change in corporate structure (for example, removing limited liability) may be necessary to change incentives in opaque institutions.

Gianni De Nicolo's paper on liquidity regulation highlights how a number of important externalities linked to liquidity were brought out by the GFC. These include "fire sale" externalities where illiquid assets have to be sold at below fundamental values, "strategic complementarities" where banks adopt similar strategies and thus increase systemic risk, and "network externalities" where contagion risks arise from failure of banks to internalise liquidity risks arising from concentrated exposures across the system.

While there is a consensus in the literature that these are important market failures, many authors contend that capital regulations are sufficient and that liquidity regulations impose extra social costs on the economy by restricting maturity transformation. This debate is particularly important in the context of two new Basel regulations. First, there is the Liquidity Coverage Ratio, which requires banks to hold reserves of liquid assets to meet short-term (30 days and under) liabilities. Second, there is the Net Stable Funding Ratio which is the ratio of the available amount of stable funding (customer deposits, long-term wholesale funding, and equity) to the required amount of stable funding over a one-year horizon. The latter is especially seen as requiring changes in banks' structural funding while also requiring adaptation by central banks in their operational frameworks, as it is likely to reduce money market volumes and increase the attractiveness of long-term central bank refinancing. One suggestion is that ex ante prompt corrective action elements in liquidity regulation could provide appropriate financial stability protection at lower cost.

Overall, we are sympathetic to the view that well capitalised banks should be able to obtain liquidity readily. But this comes back to the point about how to make it in bankers' own interests to hold enough capital and liquidity, rather than hoping that imposing ever tougher rules will be enough. Also, financial institutions can only be liquid if they operate in liquid markets. This requires an appropriate market infrastructure including rules, reporting requirements and clear legal and accounting frameworks. Some of the most important global markets proved to be at best illiquid and at worst rigged with illegal activity.

James Barth and Clas Wihlborg define 'too big to fail' (TBTF) where a bank is seen to generate unacceptable risk to the banking system and the economy if it were to default and fail to fulfil its obligations. Costs imposed on the economy are firstly that competition between banks is distorted if large banks gain an interest rate subsidy from the expectation of rescue. Second, a few large banks may have a very strong political influence on regulators. And third, there may develop a link between bank risk and sovereign risk, as the cost of bailing out a bank contributes to a nation's fiscal crisis. The problem has been growing historically as large banks continue to grow and dominate financial systems. 'Big' may be defined in various ways, including not only various measures of size but also complexity, whereby empirical work shows that number of subsidiaries and involvement in market based activities contribute to systemic risk.

The importance of complexity as well as size (also interconnectedness, substitutability, cross-jurisdictional activity) is reflected in the definition of Global Systemically Important Banks (G-SIBs) under Basel III and stricter regulatory capital requirements. However, the definition of Systemically Important Financial Institutions (SIFIs) is not internationally consistent, being defined at a national level. Further reforms aimed specifically at TBTF, such as the Dodd-Frank Act, the UK Vickers legislation and EU Liikanen report, address it in one or more of the following ways: restricting bank size directly, separation of different activities by ring fencing, requiring higher capital and providing an orderly wind-down framework.

The authors note that the costs of TBTF regulation in terms of lost economies of scale or scope are rarely allowed for, nor are the risks of activities shifting to the shadow banking sector. Yet they are sceptical whether the reforms underway are strong enough to allow a large bank to be resolved with uninsured creditors sharing the losses. We would add that it is difficult to credit that most of the largest banks on the eve of the GFC are even larger today. Moreover, TBTF was not limited to deposit taking intermediaries. Under the authors' definition, it is perfectly possible that very large insurance or asset management firms may become TBTF. Thorsten Beck highlights how cross-border banking has grown rapidly in recent decades, not only in OECD countries but also in developing countries. Supervisory cooperation is essential because failure of a bank in one country can give rise to substantial externalities in other countries, notably given the ongoing integration of financial systems. Indeed, the failure of large cross-border banks such as Lehmans, Fortis and the Icelandic banks was a salient feature of the GFC. Efficient resolution proved particularly difficult and led to political conflict between the countries concerned. Reasons for difficulties include not only a lack of bank resolution frameworks even at a national level, but also differing legal and regulatory systems that limited scope for cooperation. National governments represent their own taxpayers and the incentive of local supervisors is to focus on national stability concerns.

Three traditional instruments to deal with cross-border banks are consolidated supervision, Memoranda of Understanding (MoUs) and Colleges of Supervisors. All have significant limitations, for example the non-binding nature of MoUs. The GFC shows the need for adequate resolution mechanisms, loss allocation and information flows at cross-border level. Since the crisis, a number of helpful developments have occurred, such as living wills, strengthening of cross-border regulatory cooperation and mandating of some MoUs. In the EU we have seen the introduction of supranational supervision under the Banking Union. But even this may not resolve the issues in cross-border failures, as the safety net has not been moved to a supranational level. In developing policies, Beck also argues that regulators need to become more aware that there is a feedback loop from changes in supervisory architecture to the decisions of cross-border banks.

Finally, Dirk Schoenmaker and Peter Wierts remind us that the authorities stood back and allowed imbalances to develop that led to the GFC. The consensus was that focussing monetary policy on consumer prices and supervision on individual institutions (whose models assumed risk is exogenous) were sufficient for monetary and financial stability. In fact the neglect of asset prices, leverage incentives, fragility to shocks and of the endogenous nature of risk in a downturn was catastrophic.

There is now a new consensus that macroprudential policy (MPP) in the time-series dimension should focus on systemic resilience to financial shocks, while the cross-sectional dimension must address TBTF. Whether MPP should be used to increase financial resilience or constrain financial booms and the balance between MPP and micro-prudential financial policy are unresolved. Furthermore there are important issues in the relation of MPP to monetary policy, not least in the light of the impact of near-zero interest rates on risk-taking and therefore financial exposures. It is unclear that inflation targeting is always consistent with financial stability. Particular issues arise for MPP in a monetary union, where a one-size-fits-all monetary policy requires variation of MPP at a national level.

Core to reporting for MPP should be measures of the financial cycle, with a particular focus on credit and real estate prices. Whereas Basel III mandates a countercyclical capital buffer for banks, this may be inadequate to break a credit boom. The authors recommend a similar buffer for liquidity as well as tying remuneration packages to long-term bank performance, and application of instruments cross border as recommended by the G-20. They also recommend a time-varying leverage ratio across all financial institutions to dampen the credit cycle, with much lower leverage than in Basel III to constrain credit growth.

There has been progress in regulation since the 2007-9 GFC. But many policies were set even before the crisis had finished, let alone understood. Most authors contend that Basel III falls short of what is required in many ways: levels and quality of capital, the form of liquidity regulation, risk weights, inconsistent definitions and the nature of countercyclical buffers. Authors also highlight the political influence of banks as a barrier to reform and the differing interests of countries involved in international banking regulation. Appropriate incentives and information remain central to financial stability. A fundamental question is whether stability can be imposed by regulation or requires changes in the legal structure of opaque firms to align risks with principals' returns.

Angus Armstrong * and E. Philip Davis **

* National Institute of Economic and Social Research, email [email protected]; ** National Institute of Economic and Social Research and Brunei University, email [email protected].