Slaughtering sacred cows: it’s the turn of the unsecured creditors now

Why are the unsecured creditors of banks and quasi-banks like AIG deemed too precious to take a hit or a haircut since Lehman Brothers went down?  From the point of view of fairness they ought to have their heads on the block.  It was they who funded the excessive leverage and risk-taking of banks and shadow banks.  From the point of view of minimizing moral hazard – incentives for future excessive risk taking – it is essential that they pay the price for their past bad lending and investment decisions.  We are playing a repeated game.  Reputation matters.

Three arguments for saving the unworthy hides of the unsecured creditors are commonly presented:

    • Unless the unsecured creditors are made whole, there will be a systemic financial collapse, with dramatic adverse consequences for the real economy.
      • If the unsecured creditors are forced to take a hit, no-one will ever lend to banks again or buy their debt.
        • The ultimate ‘beneficial owners’ of these securities – notably pensioners drawing their pensions from pension funds heavily invested in unsecured bank debt and owners of insurance policies with insurance companies holding unsecured bank debt – would suffer a large decline in financial wealth and disposable income that would cause them to cut back sharply on consumption.  The resulting decline in aggregate demand would deepen and prolong the recession.

          I believe all three arguments to be hogwash.


          As regards the first argument – financial Armageddon – it may have escaped people’s notice that, with the exception of a few struggling survivors, the large border-crossing banks in the north-Atlantic region would be insolvent but for past, present and anticipated future government financial support.  Insolvent financial institutions on opaque tax-payer underwritten life support should instead be put into administration; if they are systemically important, the desired results can be achieved through the special resolution regime (SRR) with prompt corrective action (PCA) that most countries in the region now have in place.  A standard and desirable feature of such (proto-) insolvency procedures is the mandatory conversion of unsecured debt into equity and/or the write-down of (part of) the unsecured debt.

          The financial market cardiac arrest that followed Lehman’s demise was not evidence that unsecured creditors should be spared.  Lehman’s insolvency and liquidation was the correct outcome, but the process was badly mishandled.

          First, following the decision of the Fed and US Treasury to arrange for and underwrite the take-over of Bear Stearns by JPMorgan Chase, the market had been lulled into believing (I certainly did), that the US government had underwritten the entire balance sheet of the internationally visible US banking sector.  Creating that belief, allowing people to act on it for six months. and then to say  ‘well, actually, we did not mean that’ would obviously rattle the nerves of many.   Letting the unsecured creditors of other banks and non-bank financial institutions take a hit would not create the same surprise today, judging from the CDS rates for most banks and the default risk premia on bank bonds.

          Second, there was no SRR for investment banks at the time of the Bear Stearns crisis.  Nor was there at the time of Lehman’s collapse, six months later.  The only insolvency option was to put these institutions in Chapter 11 or Chapter 7.  That problem no longer exists, because the category of free-standing investment bank to which Bearn Stearns and Lehman belonged no longer has any members.  The last two surviving members, Goldman Sachs and Morgan Stanley, became bank holding companies, supervised by the Fed and with the SRR and PCA regime appropriate to such institutions, administered by the FDIC.

          Third, even absent a SRR for investment banks, it is extraordinary that the SEC, the Fed and the US Treasury did not between them come up with a way to ring-fence the ‘financial infrastructure services’ provided by Lehman through its role as a custodial counterparty in tri-partite repos.  Again, with all major banks now subject to an SRR with PCA, this issue ought not to arise again.  I hope that for systemically important non-bank financial institutions, including insurance companies like AIG, there now also exists an SRR, which allows the government to take over the management of the company, with full powers to sell assets, spin off business units and ring-fence subsets of the assets and liabilities.

          The Fed’s and US Treasury’s multi-stage bail-out of AIG provided a massive windfall to a particular class of unsecured creditors, owners of CDS written by AIG.  About $58 bn was paid out by government-bailed AIG to banks headquartered outside the US.  While not all of that money necessarily represents a loss to AIG or to the US tax payer, it is surprising, to say the least, to have official US concern for the well-being of unsecured creditors of US institutions extend to foreign creditors.  An admirable manifestation of internationalism.

          The global financial cardiac arrest that occurred in the second half of September 2008 (and which now has passed), was in my view due more to the sudden dawning of the realisation that most of the leading border-crossing banks headquartered in the north Atlantic region were insolvent (but for the tax payer’s past, present and future favours), that occurred when AIG had to be rescued.  The insurance that banks had bought against default on their bond investments, first from the monolines and then from AIG and similar shadow-banking sellers of shadow-insurance products, turned out to be pretty much worthless.

          The fear that followed the markets’ realisation that the banking sector faced an insolvency rather than a liquidity problem did not abate when Treasury Secretary Paulson presented his first TARP proposal to the Congress.  This consisted of three A4 sheets, which said: “I want $750 bn.  I want it now.  I will use these funds for good works, but I cannot tell you what these will be.  Don’t ask any questions.  And you cannot sue me.” This political blunder convinced many that the Treasury Secretary was out of his depth.  When the initial request was turned into a 300 page regular pork-laden Congressional document, the market breathed a sigh of relief.  But when Congress then turned down this proposal at the first time of asking, the markets realised that the Congress was childish and irresponsible.

          Cardiac arrest seems a reasonable response to this cumulation of bad and worse news.  A repeat does not seem likely.  Major negative surprises about the health of key financial institutions are unlikely, because health perceptions are already so pessimistic.  And compared to the Bush administration, the new administration is unlikely to be as accident-prone in its interactions with the Congress.

          Unsecured creditors’ strike

          As regards the second argument – if banks default on their unsecured debt, no-one will ever lend to them again – it ought to be unnecessary to point to the logical flaws and to the empirical evidence to the contrary.  Unfortunately, I hear the argument sufficiently often, I feel it incumbent on me to debunk it here.

          First, when a borrower defaults on his outstanding debt, his ability to take on new debt and to service it improves.  The only reason not to rush in and offer him your shillings immediately is the possibility that the past default provides evidence that increases the perceived likelihood of a future default.  If the default was a pure ‘bad luck’ default, it would not do so.  If the default is evidence of (unexpected) bad management by the borrower, new credit is less likely to be forthcoming.  If the default is perceived as ‘discretionary’, that is, a case of ‘can pay but won’t pay’, credit is likely to be cut off.

          Second, even sovereign debt defaulters have not suffered dramatically for their transgressions. Serial sovereign defaulters like Argentina tend to be back in the market after as little as three years.  The Russian and Ukranian defaults of 1998 did not banish these countries to the Valley of the Financial Lepers for very long.  Carmen Reinhard and Kenneth Rogoff have written two wonderful historical studies of sovereign defaults on external debt and on internal debt.  I recommend both papers highly.

          Even more relevant is the fact that, if the authorities adopt the good bank solution proposed by (among others) Paul Romer, Joseph Stiglitz, George Soros, Robert Hall and Susan Woodward and myself, it would be the legacy banks, stripped of their banking licenses and not engaging in any new lending or new investments, that would default on the unsecured debt.  The new good bank (with just the (insured) deposits and the good assets of the original bank on its balance sheet in the version of the good bank model proposed by Hall and Woodward), could, until emotions and moods have settled, have its new unsecured debt guaranteed by the government.  Rational would-be new unsecured creditors of the new good bank would in any case not be deterred by the bad experience of the old unsecured creditors of the bad bank.  So there should be no problem attracting new unsecured creditors willing to lend to the banks.

          Finally, a modicum of discouragement of new unsecured creditors is desirable.  We don’t want to return to the reckless unsecured lending to banks of the past – lending that was highly instrumental in bringing us the crisis.  Greater lender caution and prudence and a higher interest rate on unsecured lending to banks will be a positive and desirable feature of the landscape banks will have to operate in during the years to come.

          Pensioners won’t spend

          Default on unsecured debt, whether it takes the form of a write-down or write-off or (partial) conversion into equity will, through the pension funds and insurance companies holding these instruments, affect the consumption decisions of pensioners who find themselves with seriously diminished pensions and insurance policy holders whose policies have diminished in value.

          I agree that this is what will happen.  But what happens if instead the government bails out the unsecured creditors and makes the pensioners and insurance policy holders whole?  The losses are still there and the government has to pay for them.

          Consider the case where the government funds the secured creditor bail-out by raising taxes immediately.  Unless there are important distributional asymmetries in the incidence of the tax increase under the bail-out scenario and the incidence of the losses suffered by the pensioners and insurance policy holders under the no bail-out scenario, the negative effects on demand should be about the same.  If the unsecured creditor bail-out is financed by cutting public spending on goods and services immediately, the negative effect on demand is likely to be greater under the bail-out scenario than under the no-bail out scenario.

          If the government borrows to fund the unsecured creditor bail-out, the bail out will be less contractionary than the no-bail out scenario, unless there is Ricardian equivalence (debt neutrality) – something I consider empirically untenable – or there is financial crowding out through an adverse asset market response to larger deficits.  This could happen if the government has limited ‘fiscal spare capacity’ – its ability to commit itself credibly to future tax increases or public spending cuts is limited.  In that case the government could still achieve a more expansionary effect from the bail-out scenario than from the no-bail-out scenario, if it were to monetise the debt incurred as a result of the bail-out.  Of course, this monetisation could only be temporary if inflation is to be avoided when the economy returns to full employment.

          More importantly, the government could prevent a decline in aggregate demand under the no-bail out scenario by a conventional Keynesian demand stimulus (tax cut or public spending increase), financed either by borrowing or by printing money.  Even a balanced-budget increase in public expenditure on goods and services would be expansionary given a strict Keynesian multiplier.

          So the third argument, that not bailing out the unsecured creditors would lead to a contraction of aggregate demand, may well be true but does not represent an argument for bailing out the unsecured creditors.  The superior aggregate demand effects from bailing out the unsecured creditors by borrowing or printing money, compared to a policy of not bailing out the unsecured creditors exists only if the policy of not bailing out the unsecured creditors is accompanied by the government not doing anything on the fiscal side.  The same demand-supporting effect achieved with the deficit-financed bail-out, can be achieved equally well by not bailing out the unsecured creditors and implementing a deficit-financed expansionary fiscal measure.

          I conclude there are no valid arguments against forcing the unsecured creditors of bank and other financial institutions to sink or swim on their own, without any financial support from the state.  Fairness considerations and moral hazard certainly support putting the unsecured creditors at risk.  They made their bed; now they should lie in it

          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