Stigma, schmigma

Stigma is routinely trotted out by assorted monetary authorities as a reason for not revealing the identities of banks or other entities that borrow from the central bank at the discount window or at any of the myriad other lending facilities created since the onset of the crisis.  Most recently, the Chairman of the Fed, Ben Bernanke, has used the fear of stigma as an argument for not providing Bloomberg News with information about the identities of the banks that have accessed the Fed’s ever-expanding family of liquidity and lending facilities (see my previous post on this).

The argument is that to reveal that a bank has chosen to use (or been compelled to use) the lending facilities of the central bank, would cause this bank to be perceived as less creditworthy than before (and than it would have been if its use of central bank facilities had not been revealed).  As a result, the bank becomes a less attractive counterparty in private financial transactions.  It becomes more costly and more difficult, perhaps impossible, for that bank to fund itself privately.  Using the loan facilities of the central bank could therefore, paradoxically, lead to the funding situation of the bank being worse than it would have been had it not borrowed from the central bank.

Before I deconstruct the logical structure of this argument, let me make two general points.

First, empirical support for the stigma hypothesis is sadly lacking.  When pushed, central banks provide anecdotal evidence, such as the case of Barclays bank, which borrowed around £1.6 bn overnight in August 2007 at the Bank of England’s standing (collateralised) lending facility following a failure in the computer system that processes trades.  This created quite a kerfuffle – the financial crisis was just starting – and Barclays had to issue a statement that it was “flush with liquidity” before things settled down.

I am underwhelmed by this kind of evidence. First, there is the interest cost of the overnight loan. The standing lending facility (which no longer exists in its August 2007 form today) lent at a rate 100 basis points above Bank Rate.  Bank Rate is the rate at which Barclays could have borrowed from the Bank of England, had it had its house/system in order.  That translates into an effective interest rate charge on the £1.6 bn of around 0.003 percent, say £50,000.00.  No banker gets out of bed for £50,000.00, but even so.  In any case, the financial penalties attached to the use of central bank facilities have been steadily reduced.  The use of the overnight facilities now tend to attract only a 25 basis points penalty.

Second, the failure of the bank’s system was embarrassing and possibly worrying.  How many customers and creditors of a bank finding itself in a Barclays-type situation would have thought: if this bank cannot even keep its core systems going, and if there is, apparently, not enough redundancy and backup built into the operations of the bank to stop such silly mishaps from happening, what else might be wrong with this bank?  I would consider these to be legitimate questions prompted by valid concerns.  Of course, these concerns can be expressed and these questions can be asked only if the identity of the bank using the central bank’s facilities is known to the public. Clearly, the clumsy bank would much prefer not to have its  errors and weaknesses made public, but systemic stability is likely to be enhanced by hanging the offending banks out to dry.

Even if there were a reasonable argument for keeping confidential, at least for a short while, the identity of a suckling bank hanging at the public teat, that secrecy argument does not apply to other aspects of the transaction, including the following: how much was borrowed; on what terms; against what collateral; how was the collateral valued, if there were no readily available market prices available; what haircut was applied to these valuations?

The logic of the stigma argument

Now let’s consider the logic of the stigma argument.

  • If the identity of a bank borrowing from a central bank (or Treasury) facility is made public, it either causes stigma or it does not.  If it does not there is no problem.
  • If the identity of a bank borrowing from a central bank (or Treasury) facility is made public and stigma results, there are again two possibilities.
    • The first is that the stigma is deserved.  The borrowing bank’s reputation for creditworthiness and liquidity prior to the revelation of its borrowing from the central bank, was too favourable.  Access to private sources of funds becomes more costly and more difficult as a result, and may even become impossible.  There is no problem here.  If the bank has enough collateral eligible for funding itself at the central bank’s or Treasury’s facilities, it will be able to survive until it sorts itself out.  If it does not have enough eligible collateral, it should not be in business and deserves to fail.  A properly designed special resolution regime for banks and other systemically important financial and non-financial institutions, with prompt corrective action and wide powers, can take over the running of all or part of the bank if it is deemed systemically important.  If all else fails, part or all of a systemically important bank can be nationalised.  Revealing the identity of the bank that is borrowing from the central bank or from the Treasury does not create a problem here.  It fact, it is likely to improve the efficiency of the financial system and the process of financial intermediation.
    • The final possibility is that the identity of a bank borrowing from a central bank (or Treasury) facility is made public and stigma results, but the stigma is undeserved.   The bank gets hammered unfairly in its access to private funding.  This is bad.  How bad is it?  In all likelihood not very bad.  If the stigma is undeserved, the bank will have ample collateral to fund itself at the central bank or the Treasury facilities, for as long as it takes to convince the markets that that their view of its creditworthiness and liquidity is too pessimistic.  If it does not have enough eligible collateral to see it through, say, a six-month spell of unwarranted private market disfavour, then the stigma was in fact deserved.  That bank ought not to be in business. Central banks now even lend against collateral that is not ‘good’ collateral, in the normal sense of the word, that is, central banks lend against collateral that, although it is liquid in normal times and with orderly markets, is likely to be illiquid when it is foisted on the central bank.

Asymmetric information and stigma

All this begs the question: what causes this stigma, if it is indeed unwarranted?  What causes the markets to take use of the central bank’s liquidity and lending facilities to be a signal that a bank is in worse trouble than previously thought?  If it is a completely arbitrary and irrational distortion in risk perceptions there is little that can be done to remedy the problem.  If however, the underlying reason is asymmetric or private information, there is good news.

Presumably, the bank knows more about its own creditworthiness, its liquidity, the quality of its management, systems and processes, than does the world outside the bank.  If, based on the information that is commonly made public by banks, potential private lenders to the banks cannot discriminate between poor and good credit risks, then the good risks could be crowded out of the markets by the bad risks – a form of adverse selection.

But the solution to this dilemma is simple.  Provide more verifiable information about the quality of your assets and about the variety and depth of your funding arrangements.  Instead, many banks even today continue to exploit established ways and to seek new ways to hide unfavourable information from the markets and the regulators. The latest example of this is the eagerness with which many banks have availed themselves of the recent relaxation in the fair value/mark-to-market auditing and reporting requirements.  After brow-beating the SEC and the Financial Accounting Standards Board and bulldozing the International Accounting Standards Board into diluting the application of the new fair value/mark-to-market principles, scores of banks on both sides of the Atlantic have wasted no time in hiding more of their dodgy assets from the preying eyes of the outside world.

Providing accurate, third-party verifiable information that would eliminate the information asymmetry that may underpin the stigma story is not particularly difficult.  We are not guessing at things that can only be revealed through individual introspection, but about financial instruments and contracts.  There is no deep, ‘technological’ reason for informational asymmetry here.  It’s a choice made by those in charge of an organisation.

So, if unwarranted stigma is indeed due to asymmetric information about the creditworthiness and liqudity of banks, there is a rather simple solution: more and better information about both sides of banks’ balance sheets and about their off-balance sheet exposures.

If the information asymmetry cannot be sufficiently mitigated, and if my earlier proposal for letting the unfairly stigmatised banks fund themselves fully at the central bank and the Treasury for as long as the stigma sticks, is not acceptable, there is a further solution: use the coercive powers of the state to mandate borrowing by all banks.

In this case the regulator (or some other appropriate authority) mandates that all banks make use, regularly and routinely, of the lending facilities of the central bank, regardless of whether the banks need or want it.  The willing borrowers would borrow their fill (subject to them having enough eligible collateral). The amount borrowed by the unwilling banks would be set by the central bank, which could randomize it between reasonable bounds, based on the amounts borrowed by the willing borrowers.  It would be a small tax on all banks, but if it were to resolve the stigma problem (if there is one), it could be worth it.

There is, in principle, another type of solution; that is to try and achieve a ‘separating equilibrium’ in which the good banks (currently illiquid but with a low probability of insolvency – solvent, for short) can and do borrow from the central bank but reveal themselves as good banks through some action, or some signal. To achieve such as separating equilibrium, good banks must be able to send a signal to the market that is costly to them (if it were not costly, it would not be a credible signal – it would be cheap talk) but less costly than to the bad banks (currently illiquid banks with a high probability of insolvency – insolvent, for short).

Borrowing from the central bank without the identity of the borrower being secret, cannot itself, however, be viewed as such a ‘separating signal’.  For both the good bank and the bad bank, borrowing from the central bank when that borrowing becomes public information, is costly; there is financial penalty and there is the public embarassment of being caught short of liquidity.  Would the opportunity cost of borrowing from the central bank be lower to the good bank than to the bad bank? Both would reveal they are illiquid, but the opportunity cost of not borrowing from the central bank would appear to higher to the good bank (which is illiquid but solvent if the central bank provides liquidity) than to the bad bank (which is illiquid and insolvent even if the central bank provides liquidity).

This, however, confuses social costs and returns and private costs and returns. As long as the bad bank’s managers get some benefit from extending the life of their insolvent institution through a loan from the central bank (or as long as there is some probability that the bad bank would regain solvency if if could only get through the immediate liquidity crunch), and as long as there is limited liability for managers and shareholders, a bad bank would be likely to mimick the behaviour of the good bank and borrow from the central bank.

Other than eliminating the information asymmetry by revealing the private information in a verifiable manner, I can think of no obvious signal that would break the back of the adverse selection problem.  Mandating borrowing or living with the probably rather mild consequences of the stigma would both seem to be far superior to keeping the identities of the borrowing banks secret, even for a short time.


In my view, the true reasons for the unwillingness of the central banks to make public the identities of the banks using their liquidity or lending facilities have nothing to do with stigma.  For the banks, commercial confidentiality is an overriding concern.  They see the revelation of the identities of banks borrowing from the central bank as the thin end of the wedge towards more onerous reporting and audit obligations. Even if shareholders might be interested, management and captive boards would not be, as it would dilute their discretion to manage the bank for their own purposes.

There are wider political externalities associated with accepting ‘stigma’ as an argument for hiding relevant information about the use of public resources.  It would create a dangerous precedent as regards accountability for the use of public resources in other areas than liquidity support by the central bank.  I am sure many other beneficiaries of state’s financial largesse would prefer to have their names kept out of the papers.  They should not be granted this wish.  Accountability for the use of public funds is well worth a bit of stigma.

For the central banks, the refusal to reveal the identities of the borrowers is partly just the manifestation in this particular setting of a long-standing central bank obsession with secrecy and confidentiality.  This goes back to the period of central bankers as performers in quasi-religious mysteries, with central banks as their temples.  Significant remnants of this ethic can still be found on the European continent and in the US – less so in the UK.

Many central banks are also far too close to the banks they deal with – they have been the objects of cognitive regulatory capture or other forms of regulatory capture.  As a result they tend to act as advocates or lobbyists for the banking sector rather than as supervisors, regulators and sources of scarce public funds that have to be properly accounted for.

In addition, revealing the identities of the borrowing banks is likely to be seen by the central banks as part of a political drive towards greater accountability by the central banks for their use of public resources – as asset managers or indeed as portfolio managers.  Central banks rightly fear that the pursuit of their traditional objectives – price stability (or price stability and full employment) and financial stability – could be impaired by too close a scrutiny of their performance as managers of ever larger and ever more risky portfolios of public and private securities.  Well, welcome to the 21st century world of central banking.  This is all there is.  You break it, you own it, even if you broke it in a worthy cause.

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