The Old Lady’s Untruthful Accounts: Another Fine Mess
Julian D. A. Wiseman
Abstract: two recent documents from the Bank of England further illustrate the dysfunctional complexity of the implementation of monetary policy.
Publication history: only at www.jdawiseman.com/papers/finmkts/untruthful_accounts.html. Usual disclaimer and copyright terms apply.
Contents: Introduction; The Speech of Paul Fisher; The Statement on the Emergency Liquidity Assistance.
Introduction
The Bank of England has just published two documents that are revealing about the inadequacies of the BoE’s interaction with financial markets. First, a speech given on 19th November 2009 by Paul Fisher, the Bank of England’s Executive Director Markets: The Bank of England’s Balance Sheet: Monetary Policy and Liquidity Provision during the Financial Crisis. This speech that appears to be an explanation and justification of the Bank’s “Sterling Monetary Framework”, this being “the means by which the Bank implements monetary policy, and offers liquidity insurance to the banking system” (¶3, all such paragraph labels being within that speech). The second document, which has attracted much more press attention, dated 24th November 2009, is the Statement on the Emergency Liquidity Assistance (ELA) extended to two institutions, RBS and HBOS, in the Autumn of 2008. This reveals that, at the peak of the credit crisis, the BoE lent £61.6bn to the two banks.
Paul Fisher’s speech contains several misdescriptions. These are only partly the fault of Mr Fisher, who now has the tough job of justifying decisions pre-dating his tenure. Alas the Bank’s implementation of monetary policy has long been needlessly complicated and dysfunctional, and as the required fundamental change in reasoning would sweep away the previous labours of his superiors, is likely to remain so for some years. The second document reveals some of the messy consequences of this complexity.
The Speech of Paul Fisher
¶4: “The SMF [Sterling Monetary Framework] has to control the amount of money in the economy consistent with monetary policy objectives.”
This lies somewhere between misleading and wrong. The net amount of money provided by the BoE to the banking sector is, roughly, the size of the note issue, less the quantity of money lent by the BoE to its own customers (primarily the UK government and foreign central banks). Thus the size of the note issue is determined by growth in the economy, by inflation, by some seasonal factors, and by the behaviour of its (usually well behaved) customers. Thus the net amount of money to be added is not, and cannot be, determined by the SMF. If the BoE were to supply more than this sum, and not drain the money by borrowing it back from the markets, then the excess would sit in some banks’ accounts at the BoE—effectively borrowed back overnight. The job of the SMF should be to implement monetary policy, that is to set traded short-term non-credit interest rates. If done well, this would be done so robustly that emergency changes and secret actions would not be needed, even in a crisis.
¶8: “In normal circumstances, we would implement the desired monetary policy stance by setting interest rates.”
The BoE should do so, but never has. Instead the BoE has supplied fixed quantities to the banking system, and allowed access to extra funds at a rate described by the Old Lady as “penal” (¶13). The Bank has attempted to de-stigmatise loans by making them secret. It would be more robust to make a price, without fixating on quantity. Rephrased, choose one: price or quantity.
¶10: “commercial bank reserve accounts … act like current accounts for commercial banks.”
They should, but don’t. A current account can be overdrawn (up to some limit), or it can contain a positive balance. Instead reserve accounts cannot be overdrawn; money must be borrowed via a different means. The availability of such borrowing is intermittent, or at a penal rate. Therefore, in times of market stress, commercial banks wishing to avoid stigma should hoard some money. This hoarding of precautionary balances exacerbates that market stress. Indeed, the BoE could and should do is allow commercial banks a current account. Such an account would allow collateralised overdrafts (up to some large limit), costing the policy rate. And a positive balance (again, up to some large limit) would be remunerated, at slightly below the policy rate (typically 10bp below). So with the policy rate at 0.5%, the BoE would be making a market in overnight money at 0.4%/0.5%. That is, a large amount could be deposited at 0.4%; and a virtually unlimited amount could be borrowed at 0.5%. Commercial banks could have access to effectively unlimited sums of GBP merely by parking the collateral at the BoE.
To emphasise, the BoE does not offer current accounts to banks. Commercial have to borrow via the BoE’s “standing facilities and open market operations” (¶12), the money being paid into their reserve accounts, which may not be overdrawn. In a crisis the difference is huge: with a true single-account model, eligible collateral is liquidity: collateral is left in the payment system, allowing payments out, which—if not returned by the end of the day—automatically and unfussily become overnight secured overdrafts charged at the policy rate. But with the BoE’s current model, eligible collateral is only the right to queue for liquidity at certain times, or to access it at that which is in practice a penal rate.
¶14: “The Bank’s Open Market Operations (or OMOs) are the means through which we provide sufficient money to the system so that banks can collectively meet their targets for reserves accounts.”
¶15: “OMOs take the form of lending money against high-quality collateral … for a fixed term. These loans could, in principle, be exclusively done via short-term operations.”
¶16: “However, effective implementation of monetary policy does not require the Bank to roll over its entire stock of loans each week. Such ‘churn’ would be inefficient. … In order to avoid too high a churn in its assets, part of the notes issue can be backed by instruments with a longer maturity.”
Because banks must actively borrow via OMOs, there is indeed a transaction cost to such activity, and so some of the BoE’s loans to the market can indeed be long term. But if the BoE were instead providing overdraft facilities, these loans would be redundant and marginally unhelpful. If a commercial bank can borrow overnight, every night, merely by finishing the day overdrawn (of course, against good collateral), then there would be no advantage to committing to borrowing for a longer period. If the BoE were to make a market, it could cease the needless complexity of this additional step.
¶21: “During the Autumn of 2007, it was clear that the lack of liquidity in markets was preventing banks from funding themselves through normal means. Initially, we provided additional 3 month operations, offering 4 auctions of £10 billion each, against wider collateral and at a penalty rate. These were not taken up by the banks. By December 2007, bank funding markets had become even more difficult. So, there was a co-ordinated approach by the major central banks who all announced additional measures. In the UK, we offered further expanded 3 month repos. This time without a penalty, against a much wider range of collateral than accepted previously.”
¶33: “That meant that there was more central bank money in the system than was needed to meet the demand for banknotes and reserve accounts. … So our liquidity insurance operations were creating a challenge for monetary policy control. We needed a new instrument to drain money from the system. That new instrument was the one-week Bank of England bill. … At their peak, over £100 billion of bills were issued on 8 January 2009”
So the Bank started lending money to the market for three months, and then borrowed it back at one week. Why? The lending was to allow commercial banks to know that they had access to funds with which to make payments. But, since the money wasn’t needed to make payments, the BoE’s borrowing took it away again. It would have been simpler, more understandable, and more robust, just to provide commercial banks with an overdraft limit. A bank with collateral can then be sure of being able to make payments—even though most of that collateral would typically go unused.
The Statement on the Emergency Liquidity Assistance
Next we turn to the announcement dated 24th November 2009, the Statement on the Emergency Liquidity Assistance (ELA) extended to two institutions, RBS and HBOS, in the Autumn of 2008.
From 1 October 2008 the Bank provided ELA to HBOS and from 7 October also provided ELA to RBS. The RBS facility was repaid by 16 December 2008, and the HBOS facility by 16 January 2009.
Use of the facilities peaked at £36.6bn for RBS (on 17 October) and at £25.4bn for HBOS (on 13 November). Total use of ELA across both banks peaked at £61.6bn on 17 October. At this point the two banks provided the Bank with collateral (residential mortgages, personal and commercial loans and UK government issued debt) with a total value in excess of £100bn. The banks were charged fees for the use of the facilities.
The borrowing of £61.6bn was collateralised >1.6×, that being a substantial over-collateralisation, especially as the collateral was presumably mostly £-denominated. Also note that the peak of the total lending (£61.6bn) was very close to the sum of the peaks of the separate loans (£36.6bn + £25.4bn = £62.0bn ± rounding). With the dates provided, that suggests that HBOS borrowed almost the same amount on 17th October as on 13th November: presumably HBOS borrowed the maximum and held it as a precautionary balance.
So why has the BoE made this so difficult? If the two banks had had large overdraft facilities, the precautionary element of this borrowing would be redundant. The two banks would have parked their eligible collateral at the BoE, giving them plenty of immediate access to liquidity. Instead of which, as a result of this complication, “the Bank decided to use its powers to limit the extent of disclosure in its financial statements in the 2009 Annual Report”: in other words, henceforth the published accounts of the BoE cannot be assumed to be true or fair.
As the central bank of a solidly triple-A country, that untruthfulness might be tolerable. But if the BoE were to become the central bank of a country less than triple-triple, that could really matter. Did the top management of the BoE really decide that the ever-changing and dysfunctional complication of the current regime of implementing monetary policy was really worth that loss of reputation? Which would be worse: they did consider it; or they didn’t?
Governor, please, stop this flailing around and instead implement monetary policy—a price of short term money—simply by making a price in short term money, in the form of a narrow passive corridor.
— Julian D. A. Wiseman
Paris, 27th November 2009
www.jdawiseman.com |
Also see the letter to Paul Tucker of November 2008 entitled The Implementation of Monetary Policy: The Next Attempt; and the essay of 16th March 2009 entitled On Quantitative Easing.
