Author: John McHale

The debate over the private debt relief has developed a lot of momentum, with the both the Central Bank and the Department of Finance now seemingly increasing the pressure on banks to provide debt relief.   While there certainly is a case for targeted debt relief, it seems to me that there are a number of confusions in the debate.   I think the following points are worthy of some discussion. 

1.       The distinction between accounting loss recognition and debt relief

This is a point that has previously been emphasised by Greg Connor (see quote here).   The processes of loss recognition for accounting purposes and debt forgiveness are very different.    In part motivated by the Japanese experience, it was recognised early in the crisis that it is critical for banks to recognise losses and then to be properly recapitalised.   Failure to do so can lead to the “zombie banks” phenomenon, whereby effectively undercapitalised banks are unwilling to lend.    This can further be associated with “zombie borrowers”, whereby banks engage in “evergreening of accounts – essentially lending more money to prevent default – so as to avoid having to recognise the losses.  

Through the PCAR process, Irish banks have been forced to recognise losses and to be propertly recapitalised, although concerns about mortgage losses have left lingering doubts about whether the process has gone far enough.   But this process is consistent with banks doing everything possible to maximum the recovery of loans – even loans that they have written down on their books.   Minimising the need for yet further capital injections requires that banks only forgive debts if it actually increases the expected recovery.

2.       The “debt Laffer curve”

I have mentioned the “debt Laffer curve” before, but I think my exposition just caused confusion.   But I still think it is a very useful device for thinking about the case for debt forgiveness.   The basic diagram is here.    The horizontal axis measures the present value of payments to the bank assuming full repayment on a given debt obligation.   The vertical axis measures the expected present value of the repayment.   The “debt Laffer curve” shows the relationship between the present value of the debt obligation and the expected repayment.   The basic case for (mutually advantageous) debt relief comes from the possibility of the curve beginning to slope down beyond a certain point.   This means that debt forgiveness could actually raise the expected value of repayment.   This could happen, for example, if lowering the debt burden means the borrower has stronger incentives to raise their income or to avoid default (where repossession would b e costly for the bank).    Such cases are certainly conceivable, but it is a fairly demanding hurdle.  (I would be very interested in commenters’ views on this.)  

There is, of course, the extra complication of the much discussed moral hazard/strategic default.   If the bank provides relief for people in certain conditions, then there is the potential for a bad “pooling equilibrium” where people have the incentive to be in those conditions.   Colm McCarthy has famously put this in a pithy way:

Since you cannot get blood from a stone, it is desirable to streamline the personal insolvency arrangements so as to recognise this reality.  But no incentives should be created which encourage those who can pay to disguise themselves as stones.

Now it may well be that the government wants the bank to provide debt relief even where the curve is upward sloping but relatively flat.   The relief would have the added advantage that it could help stimulate household spending.   But if this is the policy, then it is better to admit up front that the policy means bigger losses for the bank and, where the bank is State owned, ultimately for the State.   There is a danger that State-owned banks will be forced to follow a range of political objectives, storing up longer term fiscal problems and making harder to gauge the ultimate performance of the banks.   If the government wants broader debt relief, then it is better to provide appropriate subsidies and then let the bank get on with – and be accountable for – maximising value. 

3.        Insolvency rules

One way to make it more likely that the bank has an incentive to forgive debt to prevent outright default is to strengthen the “threat point” of the borrower through more debtor-friendly bankruptcy laws.   Making bankruptcy or other insolvency rules more borrower friendly is likely to reduce the value of the banks.   But I think the focus here should be on the overall design of the regime.   The current regime is archaic and brutal and needs to be reformed.   If moving to a modern regime results in losses, then additional losses for the banks is a bullet to be bitten.

However, much of the discussion in pitched in terms of a trade off between debt and creditor rights – and creditors understandably get little sympathy at the moment.   But it is important not to forget the “instrumental role” of creditor rights in ensuring there is an incentive to provide credit in the first place.   Everyone’s creditworthiness is tied to there being reasonable protection of creditor rights. 

Colm McCarthy expresses understandable frustration with the pace of developments in meeting the commitments made on June 29^th(see here).   Central to Colm’s criticism is what he sees as a fundamental inconsistency between the government’s claims of success in its crisis-resolution policies and calls for some form of official relief on banking-related debt.

Since the resort to an EU/IMF bailout in November 2010, the Government has pursued a strategy with two central components. The first is that Ireland’s debt is sustainable, since the economy is recovering and budgetary adjustment will be delivered on schedule. Ireland will re-enter the bond market and exit the programme at the end of 2013. The second is the pursuit of relief from a portion of the bank-related debt, on the grounds that it was improperly imposed.

Last week’s events should highlight once again the inconsistency of this strategy. If things are going fine, why is there any need for debt relief? The best case for debt relief (Greece was relieved of €100bn) is inability to pay.

The insistence that things are fine, that budget adjustments are on schedule, three-month Treasury bills can be sold and Ireland will exit the rescue programme next year, is a serviceable domestic political message. But it is also an open invitation to our European ‘partners’ to offer no assistance whatsoever outside the terms already agreed.

A better negotiating platform, and one with at least equal plausibility, is the following: that the debt is not sustainable and will reach 150 per cent of national income; the economy is flat and will remain so; the politics of further retrenchment are getting too difficult and debt relief is inevitable. The Government should quit behaving like the marketing arm of a debt-management agency.

Although I always hesitate to disagree with Colm, I don’t see the government’s strategy as fundamentally inconsistent.  Since peaking at 15.81 percent on July 15, 2011, the yield on the benchmark 2020 bond has followed a strong downward trend to close at 4.53 percent on Friday (Bloomberg).  Assuming a recovery rate of 50 percent in the event of a private-sector default, that the yield on the equivalent German bond represents the risk-free rate and risk-neutral investors, the implied probability of default peaked at 83.8 percent in July 2011 before more than halving to 39.7 percent on today.   

Of course, a default probability of close to 40 percent is still very high.   I think the main reason that the perception of default risk still remains so high relates to the uncertainty surrounding growth prospects.   A poor outcome on growth could make the necessary fiscal adjustments to meet the conditions for official support without a private-debt restructuring politically – and possibly even economically – impossible.   Adverse growth shocks will also do more damage to the ability to meet deficit- and debt-reduction targets the higher is the starting debt ratio.  

Recognising the common interest in a successful return of Ireland to creditworthiness, there is a case for making adjustments to official policies that reinforces the improvements made so far.   One such improvement would be to lengthen the period for paying down ELA and make continued access to that funding more reliable.  By rewarding countries that meet their commitments rather than the opposite, such actions should also help to reduce official-lender concerns about moral hazard. 

I worry that emphasising unsustainability under current conditions would suggest a weaker commitment to meet the conditions required for official support.   Any resulting weakening of perceived creditworthiness could itself undermine growth by raising the spectre of Greek-style chaos.   I believe it is better to emphasise that Ireland fully intends—and expects – to do what is necessary to avoid default, but there are certain factors that it simply can’t control.    Recognising this, there is indeed a common interest in adjusting official support policies to further support a “well-performing adjustment programme.”

Eddie Hobbs’s widely discussed WSJ article is here.   Seamus Coffey provides a detailed response on the debt points here.

Wolfgang Münchau has an interesting article in today’s FT (see here).   Not surprisingly, he highly critical of official policy towards Greece and pessimistic about the country’s prospects without a change of course.   Central to the argument is that the fiscal adjustment being demanded of Greece is self-defeating: attempts to lower the primary deficit through fiscal adjustments slow the economy so much that the primary deficit actually rises. 

While I agree with Wolfgang’s bottom line, I would put the problem somewhat differently.   Even allowing for two-way feedback between the deficit and GDP, standard analysis shows that, all else equal, changes to the structural primary balance do lower the actual primary balance.   Moreover, this result holds for any size of deficit multiplier.  (See, for example, Equation 4 from Annex B here.) 

But of course all else is not equal.   A big part of what is holding back a Greek recovery is fear of a catastrophe, in which Greece loses official backing and is forced to default and potentially exit the euro.   The tougher the conditions applied to Greece the greater is the fear that it will simply not be able to do what is required for continued support.   This fear acts as a huge drag on the economy, leading spending to contract and limiting any improvement in the deficit. 

Perhaps ironically, a similar phenomenon was nicely captured by Olivier Blanchard in his comment on the original expansionary fiscal contraction paper by Giavazzi and Pagano (see here pp. 111-116).  After considering a version of the basic mechanism discussed by Giavazzi and Pagano, Blanchard notes:

This formalization focuses on the effects of consolidation on the expected level of output; there is another, probably equally important implication of consolidation that this formalization does not capture-the effect of consolidation on uncertainty. Consolidation may be associated, at least after a while, with a substantial decrease in uncertainty, leading to a decrease in precautionary savings, to a decrease in the option value of waiting by consumers to buy durables and by firms to take investment decisions. 

While the mechanisms are similar, I think the situation is quite different for Greece today, with the actions of official creditors key to the restoration of confidence.   Rather than more decisive fiscal adjustment being the route to restore confidence, a better route would seem to be less demanding conditionality for access to official loans – conditions that stand a better chance of being politically acceptable.  This would reduce the pervasive fear that it will all go pear-shaped if the government fails to meet the conditions.   The positive effect of reducing this fear would be to help stop the contraction in Greek economy, and should actually allow a faster improvement in the deficit, ultimately reducing risk to official creditors.   Both sides could gain from a programme with better odds of success. 

Update

It is useful to revisit the arguments above in light of the excellent analysis in the IMF reports linked to by Philip.   The findings in Box 1.1 of the WEO have received most attention this morning.   The key message is that fiscal multipliers appear to have been underestimated.   Given the robustness of the results to a variety of controls, this analysis is broadly convincing.   The arguments and findings in Box 1.3 are also very interesting, showing the adverse affects of uncertainty on growth.  This underlines the cost of the failure to develop crisis-resolution policies that remove the risk of a serious escalations of the crisis in the most vulnerable countries. 

Dan O’Brien provides a useful analysis of the senior debt issue here.