A new direction in global financial regulation

January 26, 2009

John Kemp Great Debate— John Kemp is a Reuters columnist.¬† The views expressed are his own —

UK Prime Minister Gordon Brown’s call today for a new G20 charter of principles on financial regulation¬† reflects an emerging consensus among policymakers that, once the immediate crisis has passed, the regulatory framework must be fundamentally redesigned.

In particular, policymakers are concerned with how to correct the basic moral hazard problem in which bankers have an incentive to extend too much credit, while private firms and households have an incentive to take on too much debt.

A consensus is emerging that the volume of credit expansion needs to be restrained and managed as a separate policy objective. This marks a sharp break with past practice — in which central banks attempted to control the cost of credit by manipulating short-term interest rates, but have increasingly left its quantity to decisions by individual banks and borrowers.

There is also something of an emerging agreement that if credit control is a separate economic objective alongside “internal balance” (output-inflation) and “external balance” (trade and capital flows) then a new instrument needs to be developed to achieve this target.

With three targets (internal balance, external balance and financial balance) Tinbergen’s Rule says there need to be three independent policy instruments — fiscal policy, monetary policy, and a distinct credit policy.

In his recent speech to the CBI Annual Dinner in the East Midlands last week, Bank of England Governor Mervyn King alluded to the need to develop a new policy instrument to achieve credit-policy objectives. He stated a strong preference it should not be interest rates. King argued rates should continue to be used to target output and inflation.

The clear implication is the “new policy instrument” referred to by King will have to be some form of direct quantitative control, so as not to interfere with interest-rate strategy, and allow the authorities to manipulate the volume of credit for any given level of interest rates.

In 1945-1975, banking crises were few, but credit was expensive and unavailable to many households and businesses. The banking system was tightly controlled and the quantity of credit was rationed through a variety of direct mechanisms (reserve requirements, intensive bank examinations, margin requirements and a host of other direct lending controls).

Most of these were dismantled during the 1980s and 1990s. They could be resurrected but this would face stiff opposition from within the industry. It would also face hostility from within the economics establishment (broadly in favour of market solutions) and among politicians (worried about the impact of reduced access to credit for many households, and voters, in the lower half of the income distribution).

Fed Vice-Chairman Don Kohn and Prof Lawrence Summers (now head of the Obama administration’s National Economic Council) both poured scorn on what they saw as a rose-tinted view of the heavy regulatory past at the Fed’s annual Jackson Hole symposium in 2005. Both men are presumably chastened by the subsequent meltdown. But their personal opposition to intensive quantitative controls is probably still intact, and shared by many policymakers at the top of the new administration, in Congress, and among the wider regulatory community.

So the search is on for a compromise. The idea is to create a new instrument or instruments that would work with the grain of the market, rather than cut against it, and enable regulators to exercise some control over the quantity of credit being extended while preserving flexibility for banks to innovate.

If the old pre-1980 quantitative controls are seen as “first generation” methods, the hunt is on for more sophisticated market-friendly “second generation” methods that promote stability while protecting growth.

The first design issue for these new quantitative controls is whether to impose them directly or indirectly.

First-generation quantitative controls were formulated within the central bank and consisted of a series of prescriptive lending ratios.

The trend in recent years has been towards a more indirect approach, in which the central bank and other regulators set out general principles and a flexible framework; banks are then free to manage their business and risk-taking within this. One key question is how far second-generation controls will build on the modern principles-based indirect approach, or revert to a more prescriptive command-and-control one.

The second design issue is how to make quantitative controls “active” rather than “passive”. First-generation controls were largely specified in passive terms: fixed capital and lending ratios that were invariant over the cycle. But there is an emerging consensus second-generation controls should be more active and capable of varying over the cycle, limiting credit growth during the expansion phase, but also mitigating the collapse of credit during a contraction.

Contra-cyclical bank regulation policies are especially popular at the moment, because the industry is in the contraction phase, and contra-cyclicality implies a loosening of policy. The real challenge is to create a contra-cyclical approach that is sufficiently robust it can compel the banks to increase their capital cushions during an upturn.

One option is to impose reserve requirements or risk-weightings which rise above the long-term mean during expansions and are allowed to fall below it during the contraction phase. But that raises thorny questions of who measures the cycle and how. Dating and measuring business and credit cycles, and identifying turning points are notoriously difficult in real time.

To take a recent example: the start of the most recent expansion is controversial, with many commentators now arguing the Fed missed the beginning of the upturn and failed to raise interest rates in a timely manner. If the Fed, or another regulator, had been responsible for adjusting reserve requirements or risk-weightings, as well as interest rates, would the adjustments have been any more successful?

If relying on regulators’ discretion to identify turning points in the credit cycle is problematic, is there a way to make contra-cyclical controls endogenous to the lending system?

The aim would be to make reserve requirements, risk-weightings or other instruments depend on the volume of credit extended in the immediate past period(s). Credit controls would be progressively tightened the longer and faster credit expands, and progressively loosened the longer and further credit falls.

The problem with endogenous credit control policies (like endogenous interest rate policies) is that they do not work well around cyclical turning points.  In the summer of 2007, an endogenous contra-cyclical policy would probably still be tightening conditions in response to the explosive credit growth in 2004-H12007 rather than loosening them to forestall the calamitous collapse of credit that occurred later in the year.

In practice, credit is hard to define, measure and restrict. Conventional bank lending is only one element of an increasingly complex and diverse credit-creating system.

Finance companies, commodity brokers, special investment vehicles, and even hedge funds, all of which are increasingly active in wholesale money markets, may be engaging in credit-creating processes.

The question of what types of credit to control is analogous to the debate during the 1980s about what measure of the money supply to target. Moreover, Goodhart’s Law suggests any statistical or economic relationship between the chosen target measure and the wider economy will tend to break down once pressure is applied for control purposes.

In fact, as soon as the authorities decide on which forms of credit are subject to regulation and control, there is an immediate incentive to create other forms of credit in other institutions that are not subject to control and therefore more profitable. This was precisely the reason for the huge growth in the “shadow banking system” during the 1990s and 2000s.

To have any chance of being effective, the new credit policy will need to cover the whole range of institutions which create credit, not just commercial banks, and need to be applied on a fairly international basis, to prevent this sort of institutional and jurisdictional arbitrage.

No comments so far

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/