Central banks’ function to maintain financial stability: an uncompleted task CHARLES A . E . GOODHART
One of the major problems of central banking is that the pursuit of these two core purposes (CPs) can often conflict, not least because the central bank (CB) currently appears to have only one single main instrument: its command over the short-term interest rate. Indeed, a central purpose of the first two great books on central banking, Henry Thornton’s (1802), An Inquiry into the Nature and Effects of the Paper Credit of Great Britain, and Walter Bagehot’s (1873) Lombard Street, was to outline ways to resolve such a conflict, especially when an (external) drain of currency threatened maintenance of the Gold Standard at the same time as an internal drain led to a liquidity panic and contagious bank failures. Under such circumstances, however, with rising risk aversion, the CB would find that it had two instruments: its ability to expand its own balance sheet, e.g., by LoLR lending, at the same time as keeping interest rates high (to deter gold outflows and unnecessary (speculative) borrowing). The greater problem, then and now, was how to avoid excessive commercial bank expansion during good times. With widespread confidence, the commercial banks do not want, or need, to borrow from the CB. A potential restraint is via shrinking the CB’s own balance sheet, open market sales, thereby raising interest rates. But increasing interest rates during good times (gold reserves rising and high; inflation targets met), i.e., “leaning into the wind,” is then against the “rules of the game,” and such minor adjustments to interest rates as may be consistent with such underlying rules are unlikely to have much effect in dampening down the upswing of a powerful asset price boom-and-bust cycle. Although the terminology has altered, this basic problem has not really changed since the start of central banking in the nineteenth century. An addi-
tional analytical twist was given by Hy Minsky who realized that the better the CB succeeded with CP1, the more it was likely to imperil CP2. The reason is that the greater stability engendered by a successful CP1 record is likely to reduce risk premia, and thereby asset price volatility, and so support additional leverage and asset price expansion. The three main examples of financial instability that have occurred in industrialized countries in the last century (USA 1929-1933, Japan 1999-2005, subprime 2007/2009) have all taken place following periods of stellar CP1 performance. We still have not resolved this conundrum. It shows up in several guises. For example, there is a tension between trying to get banks to behave cautiously and conservatively in the upswing of a financial cycle, and being prepared as a CB to lend against whatever the banks have to offer as collateral during a crisis. Again, the more a CB manages to constrain bank expansion during euphoric upswings (e.g., by various forms of capital and liquidity requirements), the greater the disintermediation to less controlled channels. How far does such disintermediation matter, and what parts of the financial system should a CB be trying to protect; in other words which intermediaries are “systemic”; do we have any clear, ex ante, definition of “systemic,” or do we decide, ex post, on a case-by-case basis? Perhaps these problems are insoluble; certainly they have not been solved. Indeed, recent developments, notably the adoption of more risk-sensitive Basel II capital adequacy ratios and the move towards “fair value” or “mark-to-market” accounting, have arguably tilted the regulatory system towards even greater procyclicality. A possible reason for this could be that the regulators have focused unduly on trying to enhance the risk management of the individual bank and insufficiently on the risk management of the financial system as a whole. The two issues, individual and systemic risk performance, are sometimes consistent, but often not so. For example, following some financial crisis, the safest line for an individual bank will be to cut lending and to hoard liquidity, but if all banks try to do so, especially simultaneously, the result could be devastating. The bottom line is that central banks have failed to make much, if any, progress with CP2, just at the time when their success with CP1 has been lauded. This is witnessed not only by the events of 2007/2009, but also by the whole string of financial crises (a sequence of “turmoils”) in recent decades. Now there are even suggestions that central banks should have greater (even statutory) responsibility for achieving financial stability (e.g., the Paulson report). But where are the (regulatory) instruments that would enable central banks to constrain excess leverage and “irrational euphoria” in the upswing? Public warnings (e.g., in financial stability reviews) are feeble, bendy reeds. All that central banks have to offer are mechanisms for picking up the pieces after the crash, and the more comprehensively they do so (the Greenspan/Bernanke put), the more the commercial banks will enthusiastically join in the next upswing. Besides such public warnings, which the industry typically notices and then ignores, the only counter-cyclical instruments recently employed have been the Spanish pre-provisioning measures, and the use of time-varying loan to value (LTV) ratios in a few small countries (e.g., Estonia and Hong Kong). But the
Spanish measures have subsequently been prevented by the latest accounting requirements, the IFRS of the IASB; and the recent fluctuations in actual LTVs have been strongly pro-cyclical, with 100+ LTVs in the housing bubble being rapidly withdrawn in the housing bust. Indeed, any attempt to introduce counter-cyclical variations in LTVs, or in capital/liquidity requirements will always run into a number of generic criticisms:
• It will disturb the level playing field, and thereby cause disintermediation to less regulated entities (in other segments of the industry, or in other countries). It will thus both be unfair and ineffective.