Even central bankers should be held accountable: they play with public money after all

Bloomberg News filed a federal lawsuit on November 7, 2008, to force disclosure by the Federal Reserve, under the US Freedom of Information Act, about the lending by the Federal Reserve system to private banks. Bloomberg wants to know the identities of the borrowing banks, how much each one borrowed, and the assets the Fed has accepted as collateral for these loans by the Fed.

The request is not prima facie unreasonable.  Under the 11 facilities cited in the lawsuit (which don’t include the $700 bn of the TARP, which is a Treasury programme), the Fed has extended well over $2 trillion worth of credit.  Initially, most of this was secured.  With the growing volume of Fed purchases of commercial paper, and given the range of options for outright purchases of private securities provided by the $800 facility announced on November 26 ($200 bn for consumer credit and $600 bn for purchases by the Fed of mortgage-backed securities and of debt issued by mortgage lenders), the Fed is now also a major unsecured creditor.

The Fed’s exposure to credit risk is likely to escalate rapidly as the Fed engages in large-scale quantitative easing, taking onto its balance sheet, either as collateral or through outright purchases, ever larger amounts of every poorer quality private securities.  A trillion here, a trillion there -  even in Washington DC, you are talking real money for which accountability to the Congress, the US tax payer and the wider public is essential.

Our financial leaders certainly talk the accountability and transparency talk.

Consider Treasury Secretary Paulson’s words at the September 23, 2008 hearing of the  Senate Banking Committee about the need for transparency in the purchase of distressed assets (the original intent of the TARP as a fund for price discovery for toxic assets was still alive then): “We need oversight,… We need protection. We need transparency. I want it. We all want it.’‘ On September 24, 20098, Bernanke also sang the Transparency Hymn in relation to the TARP: “Transparency is a big issue,”.

Transparency is such a big issue, apparently, that it cannot be squeezed into the Fed’s procedures for managing its lending facilities. Following an initial request for this information by Bloomberg in May 2008, the Fed stonewalled, never gave a formal answer, but hinted that it could not provide the information because of a commercial confidentiality exemption clause in the Freedom of Information Act.

At a hearing of the House Financial Services Committee on November 18, 2008, Chairman Ben Bernanke was asked about this request for information.  The hearing can be seen in its entirety here. The transcript will be available at the same URL in due course.

Chairman Bernanke’s aswer consisted of variations on ‘nyet’.

“Some have asked us to reveal the names of the banks that are borrowing, how much they are borrowing, what collateral they are posting,” …  We think that’s counterproductive.”

The arguments for ‘nyet’


Chairman Bernanke, in the House Financial Services Committee hearing, played the familiar ‘air on a stigma string’ as one argument for not releasing the information.

“First, the success of this depends on banks being willing to come and borrow when they need short-term cash,” and “There is a concern that if the name is put in the newspaper that such-and-such bank came to the Fed to borrow overnight for a perfectly good reason, that others might begin to worry is this bank creditworthy and that might create a stigma, a problem, and might cause banks to be unwilling to borrow, and that would be counterproductive.”

Trust us, we know what we are doing; your money is safe

Chairman Bernanke also said that “We take collateral, we haircut it, it is a short-term loan, it is very safe, we have never lost a penny in these various lending programs,”.

An evaluation of the merits of Chairman Bernanke’s ‘nyet’

Chairman Bernanke’s arguments for not releasing the requested information are 10 percent correct, 90 percent self-serving Fed-accountability-avoiding twaddle.

Let me start by noting that, even if it were true that revealing the requested information would violate commercial confidentiality, that such a violation would create stigma for the borrowing banks and that such stigma would result in material damage to the stability of the financial system, it would not automatically follow that the information in question should be kept secret.  There are things that are even more important than commercial confidentiality, bank stigma and financial stability.  Accountability for the use of public money could be one of these things.  At the very least there would be a clash of competing public interests.  This conflict should not be resolved through a unilateral decision by just one of the interested parties, the Fed.


It is correct that the immediate revelation of the identy of a borrowing bank could be so market-sensitive, because of stigmatisation effects, that confidentiality as regards the identity of the borrower makes sense for a limited period.  But for a limited period only.  After six months, nobody cares.  After a year, nobody remembers.  Once the loan has been repaid, the stigma issue is no longer relevant.

The key point is that, for democratic accountability for the use of public money to exist, there has to be certainty that at some point there will be full revelation of the identity of each borrowing institution, how much it borrowed, on what terms, and against what collateral.  While there can be a finite (but short) delay in divulging the identity of the borrower, all other information – the amounts borrowed (collectively and by individual anonymous borrowers) should be in the public domain immediately.  Even in the most paranoid of worlds, there is no reasonable argument, other than an unwillingness to be held accountable for possible mistakes, for not releasing, instantaneously, the terms on which the borrowing occurred and the nature and valuation of each specific item of collateral offered .

It should be obvious why it is essential that all information be in the public domain, that is required to assess the Fed’s valuation of the collateral at the time the loan was made.  This is the information required to assess the magnitude of the ex-ante subsidy the Fed provided to the borrower – the quasi-fiscal subsidy, if any, provided by the Fed, based on the information available at the time the loan was made.  This information consists not just of the interest rate on the loan, the fees and the haircut applied to the collateral.  The haircut (discount) on the collateral, which central banks are often happy to provide, are applied to the price or valuation of the collateral, not to its notional or face value.

With illiquid collateral, there often is no market price on which to base a valuation to which the haircut can be applied.  In that case, central banks are supposed to do their own pricing of the illiquid securities offered as collateral. The ECB refers to such a ‘model based’ price as the ‘theoretical price’.  I have repeatedly requested members of the ECB’s Governing Council to see to it that the ECB puts in the public domain (a) the models or methods used by the ECB to price the illiquid securities offered as collateral and (b) to provide the detailed, item-by-item actual valuations (prices) applied to the collateral by the ECB (to which the haircuts, which are in the public domain, then are applied).  I am still waiting to see either the models or the valuations.

The Fed has an even worse track record as regards the valuation of illiquid collateral than the ECB.  In the prehistoric days when there still were independent investment banks in the USA, these primary broker-dealers were allowed to borrow from the Fed (through the Primary Dealer Credit Facility and the Term Securities Lending Facility) not directly, but via their clearing bank (e.g. JP Morgan Chase acted as agent for Bear Stearns in these transactions).  The Fed accepted the valuation made by the clearing bank acting as agent of the primary dealer, of the collateral offered by that same primary dealer.  How could an arrangement so full of potential conflict of interest have been agreed to by the Fed? It is designed to enable to primary dealer and its clearer to collude to pass off pigs ear securities to the Fed as silk purse collateral.

I don’t know whether these kinds of insane valuation arrangements (which apparently have a long history in the Fed’s dealings with the primary dealers) have survived the demise of the stand-alone investment bank as a business model in the USA.  But until we get from the Fed both the models or methods used by it to price the illiquid securities offered as collateral and the detailed, item-by-item actual valuations (prices) applied to the collateral (valuations to which the public-domain haircuts are subsequently applied), I will not accept any assurances that the collateral accepted by the Fed is valued properly, and that the tax payer is therefore protected.

The same lack of openness and transparency that characterises the ECB and the Fed in their valuation of illiquid collateral is also found at the Bank of England.  At least the Bank of England have told us that, like the ECB, they themselves value any illiquid collateral they accept at their various facilities (including the SLS).  But they too have not put in the public domain the methods and models used to make these valuations, or the actual, item-by-item valuations of the collateral they accept.

As long as the models/methods used by the central banks to calculate the theoretical prices are not in the public domain, and as long as we are not provided with the detailed actual valuations/prices of each security accepted as collateral, I will not believe a word I am told. This form of Publikumsbeschimpfung by the central banks is simply not acceptable in a democratic society.  It is not their money they are playing with.  It is our money.

Trust us, your money is safe

The only correct rejoinder to that argument is: no, we don’t trust you; it is not obvious that you know what you are doing; we need and we want more information to convince us that the public’s money  (our money) is, if not safe, at least put at risk in a wise and responsible manner, in the pursuit of systemic stability and other public goods.

Since the crisis began in August 2007, and even more spectacularly since Lehman’s filing for bankruptcy protection on September 15, 2008, the Fed has taken steadily growing amounts of private credit risk onto its balance sheet.  In the case of secured bank loans, the Fed faces the risk that both the borrower (the bank) will become insolvent and that the securities offered as collateral for the loan will turn out to be worth less than expected, or even to be worthless.  Many if not most surving internationally active US banks are ‘solvent’ today only because of a range of Federal government and Federal Reserve support programmes, ranging from the capital injections financed from TARP money to the massive purchases and guarantees of agency mortgages, to toxic asset guarantees (Citi Group),deposit and other debt guarantees by the FDIC etc.

Most of the internationally active US banks are dead banks walking, supported and held upright by a cast of Federal puppeteers with mixed track records.  Even with the state support they are receiving or are expected to have access to should the need arise, the creditworthiness of these banks, as reflected in their credit default swap (CDS) spreads and their spreads over US Treasuries (which themselves now have rather larger CDS spreads than they used to have before the crisis) is poor indeed.

That leaves the Fed, and behind the Fed the US tax payer or the beneficiary of existing US public spending exposed to the credit risk on the collateral backing the loans.

Before September 2007, the Fed accepted as collateral really just securities issued or guaranteed by the US Federal government and certain US government agencies. Since then, the range of eligible collateral, like the range of eligible counterparties, has been widened considerably.  Even in 2007 and early 2008, The Fed accepted mortgage-backed securities issued by or guaranteed by Fannie Mae and Freddie Mac.  These entities subsequently turned out to be insolvent and were rescued by the Fed and the US Treasury. When Chairman Bernanke tells the US Congress: “We take collateral, we haircut it, it is a short-term loan, it is very safe, we have never lost a penny in these various lending programs,” is he really saying:  I know it is safe, because if the entity issuing or guaranteeing the collateral goes broke, the Fed and/or the US Treasury will bail it out?

By continuing to accept instruments issued or guaranteed by Fannie and Freddie as collateral, the Fed clearly does not accept assets that were safe from the perspective of the US tax payer, because what safety there is comes from that tax payer’s own guarantee.  In addition, the Fed now accepts as collateral a wide range of private asset-backed securities and other private securities.  Even if these are triple A-rated, they cannot be considered safe.  Many of these assets are illiquid.  This illiquidity is at least in part a reflection of the default risk the market attaches to them.  How can any instrument backed by US mortgages be considered safe when house prices are still falling and most mortgages are non-recourse?  How can any instrument backed by US credit car receivables or car loans be considered safe?  By making statements that are so blatantly untrue, the Chairman of the Fed risks losing the trust of the markets and of the public.  Instead of instilling confidence, such lack of trust may lead markets and consumers to fear the worst, and to cut back on spending and financing even more than is warranted by the poor fundamentals.

With the Federal Funds target rate at 100 basis points, the Fed is almost out of conventional monetary policy ammunition.  Quantitative easing – the massive expansion of the Fed’s balance sheet, through the acquisition of private and public securities, financed through increases in the monetary base (mainly bank reserves with the Federal Reserve) – is the main remaining monetary policy tool.  The Fed will be acquiring, either through outright purchases, or as collateral in its lending operations, a steadily wider range of private securities.  These securities will be subject to steadily higher degrees of credit risk.  This deterioration in the average credit risk attached to the Fed’s assets reflects both the deterioration of the credit quality of existing Fed holdings of private securities as the economic slump worsens, and the purchase or acceptance as collateral of additional classes and categories of securities with ever lower degrees of creditworthiness than the current average. The end to this quantitative easing is not in sight.  I can see the Fed purchasing index funds of US equities before this is over.

Regardless of whether the Fed takes on additional credit risk through the collateral it accepts at any of its current or future facilities, through outright purchases of risky securities or through unsecured lending to banks, corporations, partnerships or individuals, it is essential for the political health of the nation that there be full accountability for the use of public funds in these operations.  That means that sooner or later (and preferably sooner) we have to know exactly who borrowed how much and on what terms.  It means that the public has to know, without any delay, not just the valuations put by the Fed on illiquid securities it accepts as collateral, but also the pricing and other terms on which the Fed purchases securities outright.

I consider the Fed’s stonewalling of the Bloomberg News request for information – the refusal to provide any information because a small component of the requested information was deemed to be commercially confidential – to be outrageous and unacceptable.  The same holds for the refusal of the ECB and the Bank of England to put in the public domain their models, methods and myths for pricing illiquid assets.

As regards the specific request of Bloomberg News, a short delay in making public the identities of the individual borrowers (sellers of securities to the Fed) may under certain conditions be justified.  All other information (what collateral was offered, what securities were purchased, valuations, terms etc.) must be in the public domain immediately.  Central banks have no immunity from accountability for the use of public resources.  Congress, the Courts, the media, the tax payer and the public at large should reject Chairman Bernanke’s ‘nyet’ to a legitimate request for information.

The central bankers’ ‘Paradise Lost’: the ‘Treasuries only’ model

Central bankers in most of the overdeveloped world grew up with the model of ‘Treasuries only” central banking.  In this model of central banking, the central bank conducts open market operations only through treasury bonds and bills, and only sovereign or sovereign-guaranteed securities are accepted as collateral at the discount window or in repos.  The balance sheet of the central bank effectively contained just sovereign debt instruments and some foreign exchange reserves.  Credit risk was not something central banks had to worry about.

Times have changed.  Central banks increasingly look like old-style investment banks, hedge funds or the riskier among the universal banks.  The asset side of the balance sheet now contains both a growing share of illiquid securities and a growing share of private securities with varying but rising degrees of credit risk.  Unlike investment banks and hedge funds, central banks by and large don’t have to worry about the funding side of their balance sheets.  Domestic currency liabilities can, if push comes to shove, always be inflated away.  Or the Treasury can transfer domestic resources to the central bank to repair a hole in its balance sheet caused by the implosion of the central bank’s assets.

As long as their short-term foreign currency liabilities (actual or contingent, explicit or implicit) are not too large relative to their liquid foreign currency assets (plus any foreign currency credit lines, swaps and overdraft facilities they can arrange), central banks are in pretty good funding shape.  Where the potential short-term foreign currency liabilities are large relative to the liquid foreign currency assets, the domestic and international solvency of the whole state sector has to back up or bail out the central bank.   When the potential foreign exchange solvency gap of the central bank exceeds the unused fiscal capacity of the government (the ability of the state to effect a successful internal and external resource transfer of the right magnitude), we have to rely on international bailouts.  When that is not forthcoming, we get Iceland.

It is a long way from Iceland, via Switzerland, Denmark, Sweden, Luxembourg, Belgium, the Netherlands and the UK, to the US.  But even central bankers in the US have to recognise that, since there are no safe assets any longer, they will have to make investment decisions that may turn out badly.  As they take on their balance sheets ever more risky securities and cash flows (or become exposed to credit risk through various off-balance sheet vehicles), the likelihood rises that the central bank could make a large loss and may require recapitalisation by the Treasury.  Even if the central bank suffers a large loss, it may have done the right thing ex-ante.  It may even knowing make decisions that will result in a large loss – and do so in the public interest and for the common good.  But we cannot make that determination unless we know who borrowed what on what terms and with what kind of collateral or who sold what and at what price.

For a central bank to do its systemically important job in this crisis, it will have to act as lender of last resort, market maker of last resort and purchaser of securities of last resort for a steadily wider and riskier range of securities.  Its performance will not just be judged on its achievements as regards price stability, employment and growth and financial stability, but also on its performance as an investor or portfolio manager.  That’s a tough set of tasks, and the last of these, asset manager on behalf of the tax payer, is not something any of the current heads of the leading central banks signed up for.


It is in principle possible for the central bank to play by the rules of the ‘Treasuries only’ model, even if the state as a whole (the consolidated central bank and general government) has to leverage up by purchasing or accepting as collateral for lending, a ever-wider range of private securities.  This would involve the central bank swapping with the Treasury, in exchange for T-Bills and T-Bonds, any private securities the central bank purchases or accepts as collateral.  Alternatively, the Treasury could own and operate, separately from the central bank, the facilities that lend to the private sector against non-sovereign or non-sovereign-guaranteed collateral, or that purchase private securities outright. The Treasury could provide T-Bills or T-Bonds in exchange for these private securities, which could then be taken to the central bank to obtain central bank liquidity.  In the UK, for instance, the Treasury could (and probably should) own and operate the Special Liquidity Scheme (SLS).

But this return of the central bank to the gentle embrace of the ‘Treasuries only’ model of central banking, would not change the reality of quantitative easing by the state and the risk-transfer to the state this involves.  The central bank’s balance sheet would expand by the same amount, but this time with public debt rather than private securities on the asset side.  The Treasury would be holding a growing portfolio of increasingly risky private securities, with the expanding stock of Treasury debt on the the liability side of its balance sheet.  Presumably, the Treasury would still call on the central bank to act as its agent in the acquisition of securities and in the valuation of illiquid collateral.  Or perhaps this task could be better performed by (former) investment bankers, working either in the Department of the Treasury or under contract to the Treasury, if the Treasury were to sub-contract these asset valuation and portfolio management activities out to a private entity. In that case the central bank would be financially safer, but also less relevant, as it would be involved only in the passive last stage of quantitative easing: the monetisation of Treasury securities that had earlier been swapped against risky private securities by some agency other than the central bank.

If central bankers cannot take the heat of running the asset side of an investment bank, they should get out of the kitchen.

Maverecon: Willem Buiter

Willem Buiter's blog ran until December 2009. This blog is no longer active but it remains open as an archive.

Professor of European Political Economy, London School of Economics and Political Science; former chief economist of the EBRD, former external member of the MPC; adviser to international organisations, governments, central banks and private financial institutions.

Willem Buiter's website