Buiter On “Unnecessary, Undesirable and Unlikely” Sovereign Defaults, And The IMF's Triumph Of Dogma Over Evidence And Logic
A month ago we discussed a recent paper by the IMF's Carlo Cottarelli titled "Default in Today's Advanced Economies: Unnecessary, Undesirable, and Unlikely"which we noted had "conclusion that are at best questionable, and at worst, completely worthless." Today, Citi's Willem Buiter does a much more eloquent job if dissecting the paper yet at the end of the day, reaches precisely the same conclusion: "A recent IMF study argues that sovereign default in today’s advanced economies (AEs) is “unnecessary, undesirable and unlikely”. We do not agree with this unqualified blanket assertion...the study represents the triumph of dogma over evidence and logic." And since sovereign default will be the primary topic for years and years to come, as more and more countries go tits up, here are Buiter's key disagreements with the IMF.
Buiter asks rhetorically if there are any completely safe sovereigns, and answers his own question:
The short answer is: no. We believe the common practice among financial analysts of treating the term structure of interest rates on the sovereign debt of most AEs countries as being free of default risk to be no longer appropriate, if it ever was. ‘Rates analysis’ has to be done simultaneously with ‘credit analysis’. Japan lost its Moody’s Aaa rating in May 2009 and is currently rated Aa2. Moody’s downgraded Italy in May 2002 and it is currently rated Aa2 (see Figure 19). Even the fiscally best-positioned G7 countries, Germany and Canada, face major fiscal challenges. Germany would not be able to join the Euro Area (EA) today, because it would fail to meet the deficit criterion (no more than 3% of GDP) and the debt criterion (no more than 60% of GDP). Indeed, the aggregate EA fails both criteria by wide margins, and of the 16 member states, only Luxembourg and Finland qualify on both criteria (see Figure 18). The reasons for the absence of completely risk-free sovereigns, despite the massive growth in sovereign tax capacity, are threefold.
First, demand for public spending has grown even faster than the capacity to tax, and has outstripped the capacity to raise revenues in a way that: (a) is politically acceptable and (b) does not materially damage incentives to work, to educate oneself and to take risk as a saver, investor or entrepreneur. Second, in a number of countries with high public debt burdens and large primary structural deficits, political polarisation has increased to the point that even if everyone recognises the unsustainability of the fiscal programme and the need for and desirability of early fiscal tightening, agreement on such a programme may be postponed as each interest group tries to minimize its share of the total burden of adjustment. Third, as part of the widespread erosion of social capital (in the sense of trust between government and citizen, and between citizens), tax administration has become less effective in many advanced economies, with rapid growth of the shadow economy and in tax avoidance and tax evasion (see e.g. evidence in Schneider and Buehn (2002) and Figure 20).
Next, Buiter touches on a topic that was discussed earlier by the CFR: how many credits will trade through the spread of the sovereign, or in other words, how and why do companies trade with less risk than their host nation.
The sovereign has the monopoly of (legitimate) taxation. It follows that, in a world with full information, effective tax administration and no legal or other constraints on the capacity of the sovereign to tax any entity in its jurisdiction, any sovereign that wants to have a credit rating at least as high as that of any entity in its jurisdiction can achieve this objective by appropriating the resources of that entity through taxation. But the capacity of the sovereign to tax faces at least two obstacles. The first would be the rule of law. There are likely to be constitutional or other legal or conventional obstacles to taxing private entities merely because they have a better credit rating than the sovereign. The second is weak tax administration.
We suspect that the capacity to tax has been eroded in a number of advanced economies to the point that we may well see a higher incidence of private entities from the advanced economies ‘trading through the sovereign’, starting with internationally operating companies and multinationals domiciled in one of the fiscally challenged EA member states. Foreign subsidiaries of western multinationals operating in emerging markets have frequently traded through their ‘local’ sovereigns. We would not be surprised, for instance, to see leading Spanish banks trading through the Spanish sovereign and we have already observed such instances for Italian or Greek companies (see Figure 21).
Next, the Citi strategist deconstructs the IMF statement alongs it core components lines, asking is it truly "Unnecessary, undesirable and unlikely?"
We assume that default is necessary if the sovereign does not have the ability to pay, even if it gave its best effort to the endeavour, without this involving inflicting cost on the nation and its inhabitants that a fair-minded, independent observer would consider impossible or intolerable. Ability to pay is a fuzzy concept, which gets tangled up with willingness to pay in any real-world application. Even the least well-off countries among the advanced industrialised nations, like Portugal and Greece, are ‘rich’ as regards their ability to restore fiscal sustainability by raising taxes or cutting public spending without forcing standards of living down to levels that would threaten health, housing, education and a reasonable social safety net – provided the polity can agree on, and the domestic policy institutions can implement, a reasonable and fair distribution of the fiscal burden.
We would argue that the undesirability of a sovereign default is not obvious, at least for the most highly indebted EA sovereigns, especially once their primary budget balances are positive or close to it. This would include Greece, Italy and even Ireland (if we consider off-balance sheet liabilities). Any breach of contract damages the rule of law, but we believe there are circumstances where default may be the lesser evil and there may indeed at times be positive externalities associated with a reminder that sovereign debt is not categorically safe.
The outcome of a cost-benefit analysis of sovereign default depends on what the alternatives are. For Greece, the only options are fiscal pain or default. This is because, first, it is unlikely that there will be significant cross-border fiscal support for fiscally challenged sovereigns within the EA, second, growthpromoting reforms are neither easier nor more urgent when sovereign insolvency threatens, and, third, as a member of a currency union, Greece does not have national monetary policy as a discretionary source of finance.
Sovereign default redistributes resources from the creditors to the taxpayers and the beneficiaries of public spending that would be cut in the absence of a default. These competing claims carry different weights in different times and circumstances. With Greece likely to have a general government gross debt not much below 150% of annual GDP by mid-2013 (if the debt is not restructured before that time), an annual interest bill of between 6% and 7% of GDP would represent a significant permanent fiscal burden.
The main penalty for default is temporary exclusion from the international, and sometimes domestic, capital markets. Default will be more attractive, other things being equal, the larger the initial stock of debt, the smaller the current primary deficit and the lower the likelihood that the government would wish to access capital markets in the future. If defaults are partial, then rational defaults will likely only occur when a sovereign runs a primary surplus large enough to service the debt it does not default on.
Default can have adverse effects even when capital market access is regained, as the cost of new borrowing for a country that has previously defaulted will be higher. Evidence on the duration of the exclusion is mixed. Uruguay defaulted in May 2003 and regained access to the international capital markets in October of the same year. The NPV haircut for the creditors was relatively small (between 13% and 20%). Other recent defaulters, such as Russia, Argentina and Ecuador, also regained market access quickly. Provided the restructuring is managed in an orderly and not too confrontational manner, the period of exclusion from capital markets thus tends to be measured in months rather than years.
There can also be systemic externalities from a sovereign default, such as subtle ‘rule of law’ externalities that may occur when the sovereign, the ultimate enforcer of contracts in its jurisdiction, engages in a breach of contract itself. More easily identified are possible contagion externalities – an increase in the likelihood of default by other parties, sovereign or private, as a result of a sovereign default. But the cost of not defaulting and continuing to carry a high public debt burden can also be significant. Thus, the March 2010 Fiscal Monitor of the IMF presents econometric evidence suggesting that, on average, a 10 percentage point increase in the initial debt-to-GDP ratio is associated with a slowdown in annual real per capita GDP growth of around 0.2 percentage points per year, although the impact is smaller (around 0.15) in AEs.
Default can be likely even if it is not necessary, as the likelihood of default also depends on the sovereign’s willingness to pay. We interpret willingness to pay broadly, to include the ability of the political system to deliver a sustainable fiscal burden sharing solution that makes the creditors whole. Our analysis of default risk is not based on what market spreads may or may not tell us, but on fundamentals: the initial debt and deficit (including off-balance sheet contingent) liabilities) and the capacity of the country to generate future primary surpluses. In our view, markets may well overestimate the default probability of the sovereign for a number of advanced countries, but at least in the case of Greece, the market probabilities are likely to underestimate the sovereign default risk, at least at horizons longer than 3 years.
Default, if it occurs, occurs in countries with extremely high government debt and deficits, the countries whose polity and society at large are the most polarised, with the weakest political institutions and leadership, the highest interest rates and the lowest growth rates. Fortunately, such countries are outliers. But Greece very much looks like such an outlier. According to Cottarelli et al, its cyclically adjusted primary balance is -10% of GDP, while the debt-to-GDP ratio stabilising primary balance is 5.5% of GDP. Greece therefore has to engage in a permanent fiscal tightening of (at least) 15.5% of GDP. The size of the tightening is enormous, and possibly without precedent. What is more, this increase in the primary surplus has to be maintained permanently for sovereign solvency to be maintained.
It is probably worth pointing out that 70% of Greek general government debt is held abroad. If national governments value their national constituencies more than they value foreign holders of their debt, the degree of foreign ownership of the debt will have an impact on the likelihood of default. In the case of default, creditors pay; taxpayers and beneficiaries of public spending shoulder the burden if the fiscal adjustment route is chosen.
It is politically difficult for a national government to impose possibly heavy burdens on its taxpayers and on the beneficiaries of its public spending programmes if the answer to the question: “why do we have to bear this burden” is: “we impose this burden on you to make the creditors whole”. Clearly, fiscal adjustment is more likely to be supported by the polity if the creditors share in the pain and take a haircut.
At this point Buiter takes a detour to mock the Greek structural reform process which has been praised so highly by the IMF, which we won't comment on as it is so blatantly obvious that Greece is making up virtually all its economic data, that even a cavemen should get it by now, even one whose ultimate shareholder is the demented Omaha duo. A far more interesting discussion focuses on Ireland and we will leave it to you to read up on why the "Irish Government will want to prepare and position itself optimally for the eventuality that it might have to choose between imposing a haircut on the holders of (some of) the unsecured bank debt it guarantees and imposing a haircut on the holders of Irish sovereign debt."
Buiter's conclusion so much more elegant than the IMF's it requires no commentary.
Sovereign default depends on many factors – initial conditions, the country’s external environment, economic, social and political institutions and its policies. In the case of Greece, this set of circumstances suggests that default (by which we mean any restructuring with maturity lengthening or an NPV haircut) is not just a distinct possibility, but a high probability event. The size of the required adjustment needed just to stabilise the burden of public debt, the eventual size of the burden of interest payments and the lack of a strong social and political consensus on domestic burden sharing that might make the extreme fiscal austerity manageable, make it unlikely, in our view, that the Greek sovereign will only impose fiscal adjustment costs on its citizens. Instead, we expect that the country’s creditors will be invited to share the burden.
Greece is certainly an outlier. We emphatically reject the notion that an average of advanced AEs (or the median AE) is likely to default. But when it comes to default, outliers matter. Ireland may yet be forced by the size of the explicit and contingent liabilities it has taken on (through its guarantee of most of the unsecured debt of its banking sector) to choose between saving the unsecured creditors of (some of) its banks and saving the holders of Irish sovereign debt. Fundamental structural reforms that would boost the sovereign revenue base are few and far between in Portugal, Spain and Italy. Although financial markets may, in our view, overestimate the probability of default for the peripheral European countries bar Greece, they would certainly do well to recognise that sovereigns can default anywhere, and that one or more are likely to do so during the decade to come, even in the EU.
Nothing is absolutely certain in human affairs. It is possible that Greece succeeds in the economic, political and social transformations that would permit it to get out of its current predicament without a sovereign default. But in our view, the blanket statement that sovereign default in today’s AEs is unnecessary, undesirable, and unlikely represents the triumph of dogma over evidence and logic.
Is Sovereign Default “Unnecessary, Undesirable and Unlikely” For All Advanced Economies? (pdf)