Early intervention regimes: the balance between rules vs discretion

Speech by Mr Fernando Restoy, Chairman, Financial Stability Institute, Bank for International Settlements, at the FSI-IADI Meeting on early supervisory intervention, resolution and deposit insurance, Basel, Switzerland, 12 September 2017.

On behalf of the FSI, I would like to extend a warm welcome to all speakers and participants for taking the time to attend this important meeting. I would also like to thank IADI, in particular IADI Chairman Tom Hoenig and Secretary General David Walker, for partnering with us on this event.

This meeting takes place at an interesting juncture - almost exactly 10 years after the financial crisis, which is widely acknowledged to have started when BNP Paribas stopped redemptions on three of its money market funds due to its inability to value their subprime mortgage exposures. This was followed shortly thereafter by an old-fashioned bank run at the UK mortgage lender Northern Rock. The failure of Northern Rock turned out to be the first of many high-profile bank failures in a number of countries, with additional collapses averted only by extraordinary government bailouts.

Fortunately, the post-crisis regulatory reforms have led to a more resilient financial system where financial crises are, we hope, less frequent and less costly. Yet, the world economy still faces a number of risks - including the protracted low interest rate environment, excessive indebtedness, potential for regulatory arbitrage and the overcapacity of the banking sector in some jurisdictions - which, collectively, may contribute to new episodes of financial instability. In this context, I cannot emphasise enough the critical role that all the safety net authorities that are gathered around this table can play in fostering a stable financial system.

In my brief remarks this morning, I will touch upon the evolution, nature and use of early intervention frameworks including their role during the financial crisis. I will also provide a few thoughts about the age-old debate on the appropriate balance between rules and discretion as it relates to early intervention frameworks. I will conclude by providing some high-level observations which I hope will stimulate lively debate over the next two days.

What does early intervention mean?

The term "early intervention" may mean different things to different people depending on one's vantage point. For supervisors, early intervention might be about the ability to change management behaviour, through moral suasion or more formal supervisory actions, while a bank's financial condition remains sound. This is broadly consistent with the goals of "risk-based supervision" - the long-standing mantra in the supervisory community - that is, the ability to identify and address weaknesses in risk management and governance before these shortcomings permeate a bank's balance sheet and adversely affect earnings and regulatory capital.

For deposit insurers and resolution authorities, the focus of early supervisory intervention might be focused more narrowly on taking intervention measures to try and avert a bank failure when a bank shows signs of distress; or in their absence, closing a bank when capital is still positive in the hope of minimising losses to the deposit insurer or the taxpayers. The reality is that what constitutes "early intervention" can take many forms and depends on what point in the life cycle of a problem bank we are talking about.

What I would like to highlight, however, is the point at which early intervention measures are taken against banks in distress. For this is likely to be the point that captures the attention of both supervisors and deposit insurers in equal measure. At the same time, this is the exactly the point when the interests of supervisors and deposit insurers start to diverge, at least in some cases.

Nature of early intervention frameworks and challenges for supervisors

In most countries, prudential frameworks generally provide supervisors with a broad range of powers to take (discretionary) action based on the perceived risk profile of the bank and the nature and severity of the identified problems.

These powers provide supervisors with maximum flexibility to make the "punishment fit the crime" and to weigh the classic trade-offs. On the one hand, well designed enforcement actions that address the root cause of identified problems can help to rehabilitate a bank when bank management might still be at the denial stage. On the other hand, acting too early, too explicitly and too forcefully might result in the "medicine killing the patient" and in triggering unintended effects on other institutions.

In some countries, legislation or prudential regulation requires supervisors to impose or, at least formally consider imposing, certain supervisory actions if a financial institution breaches defined thresholds or meets certain conditions. Depending on the jurisdictions involved, those regimes are referred to as prompt corrective action (PCA) or early intervention measures (EIM).

Rules versus discretion in the context of early intervention frameworks

PCA was first introduced in the United States in 1991 after the savings and loan and commercial banking crisis of the late 1980s. PCA prescribes a series of both discretionary and mandatory corrective actions to be taken promptly as capital ratios decline to certain levels, with actions ranging from restricting growth, when a bank is considered undercapitalised, to closing a bank if tangible capital falls below 2% for a certain period of time. The main intent of such frameworks is to limit regulatory forbearance and to reduce costs to the deposit insurance fund.

The PCA concept has been incorporated into European Union legislation through a specific regime, known as early intervention measures, which includes two components. First, it provides supervisors with a broader range of powers to impose corrective measures on banks falling under the EIM regime than those available under normal supervisory actions. Second, there is an obligation for supervisors to formally determine whether to apply early intervention measures for all banks that are assigned a low supervisory rating.

The main difference between the PCA and the EIM regimes is that the former prescribes a compulsory application of remedial measures when specific capital triggers are breached while the latter provides the supervisor with more discretion on whether to impose EIM.

This leads me to the heart of the debate in supervision - what is the right balance between rules and discretion, particularly in the context of taking action against weak banks. In the aftermath of the Financial Crisis, and as more rules are increasingly being hard-wired into regulation, should we pause and ask ourselves if the pendulum has moved too far in favour of black-and-white rules? What is the role of discretionary-based supervision in an increasingly rules-based prudential world? To help us answer these questions, let's take stock of the publicly available evidence, which is admittedly rather limited.

Outcomes - the evidence

In September 2011, the US Office of the Inspector General published a report that assesses, among other items, the outcomes of banks that were subject to PCA requirements during the financial crisis and whether US supervisors implemented the PCA requirements promptly. From January 2006 to March 2010, approximately 489 banks in the US were considered undercapitalised and thus subject to PCA triggers. The report includes two key findings:

• First, 60% of all PCA banks failed (291 of 489); this observation supports the conclusion that, by the time troubled banks became subject to PCA provisions, there might have been few options available to resolve the problems.
• Second, critically undercapitalised institutions (those with tangible capital below 2%) were closed promptly as required under PCA, but overall losses to the deposit insurance fund were significantly higher than those experienced in the 1980s banking crisis, when no PCA framework was in place.

Another personal observation (not from the study) is that systemically important banks (SIBs) facing severe stress have rarely been subject to PCA or formal early intervention actions in either the United States or the European Union. In the United States, the banks subject to mandatory PCA triggers were mainly small institutions with extremely poor - and, in many cases, irretrievable - capital situations. In the European Union, there seems to be little evidence that the formal EIM regime was actually applied to weak institutions with a systemic relevance.

Interpreting the outcomes

So what are we to make of these outcomes and how does this inform the "rules versus discretion" debate in the context of early supervisory intervention?

The limited data points from the Financial Crisis indicate a pragmatic application of the early intervention rules in both the United States or the European Union. In both cases, the evidence suggests that, in practice, corrective actions taken against systemically important weak banks were addressed by supervisors using mainly their normal supervisory powers, which are typically more discretionary and discreet in nature. Public disclosures of formal early intervention actions and the potential adverse effects such disclosures might have on financial stability may also have contributed to the use of more discreet supervisory powers in both the United States and the European Union.

Do these observations mean that supervisors did too little too late? In hindsight, there is little doubt that supervisors in some jurisdictions could have acted earlier and more forcefully to deal with weak banks before, during and immediately after the crisis. Indeed, some observers question whether the lack of formal early intervention actions taken against weak banks that subsequently failed is an indication that excessive forbearance contributed to the failures. But such an analysis is likely to offer somewhat biased results as it ignores weak banks that did not subsequently fail. In other words, sometimes discretionary gradualism - that is, discreet action taken under regular supervisory powers - if implemented appropriately in case-by-case situations, could also lead to bank recoveries.

As policymakers, we always need to weigh the costs of early intervention measures (ie the risk that premature, heavy-handed action may accelerate a bank's demise) against the benefits (ie promptly and forcefully addressing a bank's viability to reduce losses to the taxpayer). While a relatively rigid, rules-based early intervention framework may certainly reduce the scope for forbearance, it also places policymakers in a supervisory straitjacket, which may unduly constrain their ability to adopt the most effective means to maintain financial stability.

Regulatory and supervisory implications

Let me conclude by offering a few thoughts on the implications for the design and role of early intervention frameworks as well as on some key implementation challenges.

• First, supervisors must have the necessary powers to take a range of discretionary measures during the normal course of supervision (ie restrictions on dividends, the ability to require banks to hold more than the minimum regulatory requirements, restrict businesses and operations, limit compensation, remove board members etc).
• Second, effective discretionary-based supervisory actions should be taken well before a bank starts showing stress in its capital position. Various forward-looking, judgment-based supervisory tools that have been introduced post-crisis (such as supervisory review of business models, culture and behaviour, and supervisory stress tests) provide a tangible means of addressing problems at an early stage.
• Third, formal early intervention regimes, such as PCA and EIM, are useful backstops to discretionary-based, forward-looking supervisory tools. In particular, PCA-based triggers may be effective in extreme situations and can be a useful mechanism to limit supervisory forbearance and reduce losses to the deposit insurer. Nevertheless, supervisors need to retain sufficient flexibility in their use of enforcement tools in order to tailor the supervisory response to institution- and context-specific circumstances.
• Last but not least, close cooperation among all relevant players - supervisors, deposit insurers and resolution authorities - is particularly important in dealing with problem banks and indispensable in promoting the safety and soundness of the financial system. While fully respecting the specific responsibilities of the supervisory authority to assess banks' solvency and viability, all relevant safety-net players must be involved in some manner when supervisory action is taken against weak banks, and not only when those banks are subject to formal early intervention regimes or when they are failing or likely to fail.

In this spirit of cooperation and collaboration, I am delighted to see such a nice mix of supervisory authorities and deposit insurers around this table. This provides a unique opportunity for all of us to exchange views and share challenges; and I hope you will also have time, after our discussions close, to impose a "prompt intervention measure" on yourselves and enjoy the sights of Basel.