Central banks in a time of crisis: a preliminary scorecard

Joint post by Willem H. Buiter and Anne C. Sibert. Virtually the same post appears on Martin Wolf's Economists' Forum, August 20th, 2007.

Martin Wolf, Chief Economics Commentator of the Financial Times is correct in stating, in his FT column of August 15, 2007, Fear makes a welcome return, that in a crisis the central bank must save not specific institutions, but the market itself. It is, however, necessary to be precise about what it means to save the market, about which markets may need saving and about how the central bank should go about saving the market in such a way as to minimise undesirable side effects.

In a number of recent blog contributions, we have sketched the role of a modern central bank as ‘market maker of last resort’ (MMLR). This MMLR is the analogue, in a world where intermediation is increasingly through financial markets, to Bagehot’s lender of last resort (LOLR) in a world where most intermediation took place through banks (see e.g. Willem H. Buiter and Anne C. Sibert “The Central Bank as the Market Maker of last Resort: From lender of last resort to market maker of last resort"; Willem H. Buiter “Central banks as market makers of last resort, again”; Willem H. Buiter and Anne C. Sibert “A missed opportunity for the Fed”).

The market maker of last resort function can be fulfilled in two ways. First, the central bank can make outright purchases and sales of a wider range of securities than they currently do. Second, central banks can accept a wider range of securities as collateral in repos, and in collateralised loans and advances at the discount window than they currently do. Following Bagehot’s rule, the MMLR should buy these securities outright or accept them as collateral only on terms that would imply a stiff financial penalty to the owner. The central bank of course already applies a liquidity ‘haircut’ even to liquid instruments offered as collateral in repos or at the discount window. Because the MMLR would have to establish a buying price ‘in the dark’, that is, unaided by recent relevant market prices, and would inevitably take on much more credit risk than central banks have become accustomed to, the ‘haircuts’ should be severe – a financial version of ‘short back and sides’.

Making a market for a particular type of illiquid financial asset, say a collateralised debt obligation (CDO), may require knowledge that central banks currently do not have. Central banks can acquire the necessary experience in these markets in the same way that they have gained experience in the market for domestic government securities and in the market for foreign exchange: by being regular market participants. Central banks should regularly, either on their own account or on behalf of customers, conduct a small amount of business in these markets in normal times, simply to get a sense of these markets.

Such market participation may require staff with a different expertise than central bank staff currently possess. Certainly the solution is not to hire the financial engineers and quants, who are so good at exploiting ‘arbitrage’ opportunities and extracting the large returns to risk bearing in ordinary times, but whose lack of consideration for and/or understanding of economic fundamentals significantly contributed to the current market disarray. Instead, central banks should hire economists with solid training in macroeconomics, financial economics, micro-market structure and behavioural economics and then encourage them to become interested in and knowledgeable about financial markets that may become illiquid.

From the perspective of saving or supporting markets, acting as market maker of last resort where appropriate and necessary, the recent actions of the Fed, the ECB and the Bank of England have all left something to be desired.

The ECB simply drowned the markets in high-grade liquidity, adding well over $200bn worth of liquidity against high-grade collateral. As this did nothing directly to assist the markets for illiquid and low-grade securities, the ECB’s action is an example of too much of the wrong stuff and little if any of the right stuff: it lays the foundations for the next credit boom without doing much to alleviate the current credit bust.

The Fed cut the (primary) discount rate from 100bps above the Federal Funds target rate to 50bps above the Federal Funds target rate. This was a mistake and a missed opportunity. The problem was not that eligible financial institutions were unable to pay the original, higher, Fed Funds rate and survive. It was that these financial institutions are holding a lot of assets which have suddenly become illiquid and cannot be sold at any price. Lowering the Fed Funds rate just subsidised any institution with liquid eligible collateral. The Fed should instead have effectively created a market by widening the set of eligible collateral, charging an appropriate "haircut" or penalty interest rate, and expanding the set of eligible borrowers at the discount window to include any financial entity that is willing to accept appropriate prudential supervision and regulation.

The Bank of England has been economically sensible this past week, but it has violated two of its own sterling money market management objectives. As can be verified from its website, and in The Framework for the Bank of England's Operations in the Sterling Money Markets, it violated Objective 1 (to keep the overnight interest rate in line with the Bank’s official rate) and, as a consequence of violating Objective 1, the also violated Objective 3 (to provide a simple, straightforward and transparent operational framework).

Unlike the Fed and the ECB, the Bank of England did not keep the overnight market interest rates close to the Bank’s official rate of 5.75 percent. Instead it allowed the overnight interbank lending rate (SONIA), to average 6.19 percent on August 13, and to at times rise above 6.50 percent. Absent the Bank’s stated money market objectives this would have made perfect sense, as being illiquid is not a commercial bank condition that ought to be subsidized by the central bank. But, it is not in accordance with the Bank’s stated objective of keeping the overnight rate in line with the official rate. And, in violating this objective, it also violates the objective of providing a simple, straightforward and transparent operational framework.

Our solution to this problem is to omit Objective 1 or replace it with something that makes it clear that in disorderly markets, banks should expect to have to borrow from the central bank at the rate applicable to the standing collateralised lending facility rather than at the official (Repo) rate.

So, on balance, the Fed and the ECB are addressing the credit crunch with a larger dose of liquidity-provision-as-usual under orderly market conditions. In addition, the Fed has provided an unnecessary subsidy to discount window users. There is a real risk that either or both may be pushed into cutting monetary policy rates not because they fear developments in inflation (and in the case of the Fed) employment that would, according to their mandates, require them to use more expansionary monetary policy, but as a way of addressing a credit crunch and liquidity crisis. The Bank of England has not made an error comparable to those of the Fed and the ECB, but has created confusion about its sterling money market policy. It should clarify the policy.

On the whole, these central banks have not exactly covered themselves with glory. But there may be future opportunities, perhaps even during the next phase of the current crisis, for them to redeem their reputations.

© Willem H. Buiter and Anne C. Sibert 2007