By Michael Pettis · February 25, 2015 · Balance sheet fragility
It is hard to watch the Greek drama unfold without a sense of foreboding. If it is possible for the Greek economy partially to revive in spite of its tremendous debt burden, with a lot of hard work and even more good luck we can posit scenarios that don’t involve a painful social and political breakdown, but I am pretty convinced that the Greek balance sheet itself makes growth all but impossible for many more years.
The history is, to me pretty convincing. Countries with this level of debt and this level of uncertainty associated with the resolution of the debt are never able too grow out of their debt burdens, no matter how determined and how forcefully they implement the “correct” set of orthodox reforms, until the debt is resolved and the costs assigned. Greece and Europe, in other words, have a choice. They can choose to restructure Greek debt explicitly, with substantial real debt forgiveness and with the costs optimally allocated in a way that maximizes value for all stakeholders, or Greece can continue to struggle for many more years as the debt is resolved implicitly, with the costs allocated as the outcome of an uncertain political struggle.
Until one or the other outcome, the country is not a viable creditor and it will not grow. There is no way to get the numbers to work. If Europe policymakers who oppose a rapid resolution of its debt crisis continue to prove as intransigent over the next few months as they have been in the past week, I suspect that they will only be able to pull off one of their goals, which is to embarrass Syriza and get it thrown out of office.
But I suspect that many European policymakers incorrectly think Syriza is as radical as it gets, and once Syriza is discredited, almost any alternative leadership would be better. I disagree. If Syriza is discredited, and the Greek economy continues to stagnate as I expect, the alternative could very easily be Golden Dawn or some other group of radical nationalists determined to blame foreigners for their problems, and Germany will have set itself up for much of the blame. It is ironic, because in my opinion Angela Merkel is not and has never been the bully that she is made out to be, and the main reason Germany seems to be running the show is that no one else has ever dared to disagree with her or to take any position of real leadership. For that reason she and Germany are being seen as far worse than they actually are.
And this is clearly not just about Greece. Everyone understands that Greece has already restructured its debt once before and received partial forgiveness — in fact once coupon reductions are correctly accounted for Greece’s debt ratio is probably much lower than the roughly 180% of GDP the official numbers suggest. Most people also understand that the Greek debate is not just about Greece but also about whether or not several other countries — Spain, Portugal and Italy among them, and perhaps even France — will also have to restructure their debts with partial debt forgiveness.
What few people realize, however, is these countries have effectively already done so once. This happened two and a half years ago at the Global Investment Conference in London when, on July 26, 2012, Mario Draghi, President of the European Central Bank, made the following statement:
When people talk about the fragility of the euro and the increasing fragility of the euro, and perhaps the crisis of the euro, very often non-euro area member states or leaders, underestimate the amount of political capital that is being invested in the euro. And so we view this, and I do not think we are unbiased observers, we think the euro is irreversible. And it’s not an empty word now, because I preceded saying exactly what actions have been made, are being made to make it irreversible.
But there is another message I want to tell you. Within our mandate, the ECB is ready to do whatever it takes to preserve the euro. And believe me, it will be enough.
As soon as Draghi made the statement to do “whatever it takes”, markets recognized that the ECB was in effect guaranteeing the bonds of EU member states whose credibility was in question, and yields immediately dropped. It is important to understand why this was effectively a kind of debt restructuring. With Draghi’s statement, there was an immediate transfer of wealth from the ECB — or, more appropriately from the group of countries behind whom the credibility of the ECB is maintained, and this means Germany above all — to the governments whose creditworthiness was in doubt.
Not only was their debt effectively restructured, in other words, but they were also granted partial debt forgiveness. I will explain this later, but I want to start off by making two points. First, Draghi’s promise was far from an optimal resolution of the European debt crisis, in part because it does not address the key issue of uncertainty, to which I will return. Second, there is an enormous amount of confusion about debt restructuring, especially at the sovereign level.
Many people believe that any sovereign debt restructuring, and its accompanying implicit or explicit debt forgiveness, is already an admission of failure, with moral and nationalist overtones. But as the implicit restructuring that occurred with Draghi’s promise to do “whatever it takes” suggests, debt restructuring is actually a process that involves increasing the value of the obligations and operations of the restructuring entity. It can be done well, it can be done badly, and it can be done disastrously, but it is a financial operation with a clearly defined goal of improving the overall wealth of stakeholders, and while it is reasonable that stakeholders negotiate the ways in which this additional wealth will be allocated, the negotiation should not prevent the restructuring.
In order to understand the differences between optimal and suboptimal sovereign debt restructuring it is necessary that we understand the relationship between the asset and liability sides of a balance sheets and how the interaction between the two create or destroy value. In this blog entry I hope to address the relevant issues. It will be divided roughly into three parts in which I want progressively to describe what an optimal debt restructuring should accomplish.
Why must Europe restructure much of its debt? The purpose of a debt restructuring is to make all parties better off by increasing the value of the associated instruments and improving future growth prospects for all the relevant stakeholders. Once the existing debt structure adversely affects future growth prospects and reduces the current wealth of the relevant stakeholders, it makes sense to consider ways in which the debt can be restructured so as to improve both current value and future growth prospects.
For most economists, debt is the way operations are funded, and the best debt is the cheapest. I am not suggesting that economists are unaware that certain debt structures are riskier than others, but for the most part they ignore the structure of the balance sheet and focus primarily on the way assets are managed. The moment debt levels become high, however, or create institutional distortions, they begin to affect, and usually constrain, value creation. Debt has four very separate and very important functions, and it is important to understand what they are before deciding what an optimal balance sheet looks like.
Once we understand the role and impact of the structure of the balance sheets, it becomes possible to describe what an optimal debt restructuring should accomplish.
Debt can be thought of as a moral obligation when a loan is extended from one individual to another, especially if there is no interest on the loan. But loans to businesses or to sovereign entities are business transactions, and they should be managed as such. The only moral obligation in restructuring sovereign debt, it seems to me, is for policymakers to fulfill their political responsibilities to do what is in the best interests of their citizens and to participate in a responsible way in the global community. The debt restructuring process is, in other words, morally neutral.
The decision to restructure debt
Normally the relevant characteristics of a bond issued by government are very clear. Needless to say most borrowing agreements specify a series of payments to be made in a specified currency on a specified date, but this is not what I mean when I say the relevant characteristics of a bond issued by a government are very clear. The clarity that matters is in the resolution of the debt.
Let me step back for a moment. Debt is always “resolved”, in the sense that either the obligor makes the contracted payments as expected and without difficulty or, in the case where it is unable to make the contracted payments, there will either be an equivalent payment absorbed by the creditor, in the form of losses and a write down, or some other entity will step in and directly or indirectly make the payment. Put differently, debt is always “paid”, if not by the borrower, then by someone else. The resolution of debt can be explicit, transparent and certain, or it can be non-transparent and uncertain, but however it occurs, debt is always resolved in a way that requires implicit or explicit wealth transfers from one or more entities.
This is true of both private debt and sovereign debt, and it is true whether the debt arises directly or in the form of contingent liabilities. However in the case of sovereign debt, all the costs associated with servicing and paying down the debt are ultimately assigned to different groups of stakeholders. These include workers and ordinary households, small and medium-sized companies, large companies and multinationals, wealthy households, or, especially in the case of countries with very large public sectors, local and central governments through asset liquidation and land sales. Stakeholders can even include foreigners, so in the case where a country’s external debt is restructured with partial debt forgiveness, the debt is “resolved” in part by wealth transfers from foreign creditors.
When I say that “normally” the relevant characteristics of a bond issued by a government are very clear, I mean that “normally” there is no uncertainty associated with how these payments will be resolved. The government runs a budget, in which expenditures are funded in a fairly explicit manner by tax collection and by borrowing, or sometimes by asset sales, and there is an assumption that the costs associated with servicing the government’s outstanding bond obligations will be met through normal budge operations. Governments also sometimes resolve debt by hidden means, by monetizing the debt or through financial repression, in which case it is usually middle class savers, most of whose savings are in the form of monetary assets, who end up paying for the debt.
As long as debt levels are not “excessive” and the government is believed to be solvent — which in the context of sovereign debt always refers to the ability of the government to service its contractual obligations at a cost that is politically acceptable — there is very little uncertainty associated with the resolution of debt servicing costs. Payments will be made out of ordinary budget operations and none of the stakeholders worry that the high debt-servicing cost will force unexpected changes in the way in which costs are allocated and the debt resolved.
Once debt levels are excessive, however, questions arise about the country’s solvency and about the steps the government will take to resolve the debt. This is the difference between a fully creditworthy borrower and one that is not – in the former case there is no uncertainty about how the cost of resolving the debt will be allocated and in the latter there is. As I will explain later, this uncertainty forces stakeholders to alter their behavior, and they do so in ways that always automatically either increase balance sheet fragility further, reduce growth, or both.
What makes matter worse is that when debt levels are excessive it is not just how the costs will be allocated that is uncertain. In most cases the amount of the obligation itself becomes uncertain. This is because depending on how economic circumstances evolve, additional contingent liabilities may be forced explicitly or implicitly onto the sovereign credit. When the government loses credibility, it is almost always the case that total debt ends up being even higher than originally contracted, and there are many reasons why this happens. A slowing economy, for example, may increase the number of bankruptcies that the banking system has to absorb, and because in many cases the banking system is already effectively insolvent, and must be explicitly or implicitly bailed out by the government, already high levels of sovereign debt will rise even further. If the country is forced to devalue the currency, to take another example, the value of external debt will soar, or if slower growth causes fiscal revenues to be less than expected and fiscal expenditures greater, debt will rise further.
This creates by the way a problem that is not sufficiently recognized in the debate about the European debt crisis. A debt crisis must be resolved quickly because there is a self-reinforcing component within the process that can be extraordinarily harmful. High levels of sovereign debt create uncertainty about how the costs of resolving the debt will ultimately be assigned. This uncertainty causes growth to slow by adversely changing the behavior of a wide variety of stakeholders in the economy (as I will describe later). As the economy slows, contingent liabilities within the banking system rise, tax revenues decline and fiscal expenditures rise, all of which push up sovereign debt levels even further and increase both the cost of resolving the debt and the uncertainty about how the costs will be assigned. The consequence of this self-reinforcing deterioration in the sovereign balance sheet is, at first, a slow grinding away of the economy until the market reaches some point, after which the process accelerates and debt can spiral out of control.
But long before things spiral out of control, when debt levels are excessive — and debt levels are excessive as soon as there is wide-spread uncertainty about the amount of the costs and the sectors to whom they are going to be assigned — the economy is being damaged and, with it, political, economic and social institutions are being degraded. With very high levels of debt the decision Europeans face is whether to move quickly to reduce uncertainty and assign the losses to sectors best able to absorb the losses, and in ways that maximize future value creation, or to allow these losses to be assigned implicitly, over many years, and in ways that are hard to control.
This is the heart of the matter. The point of a debt restructuring is to eliminate or reduce the uncertainty associated with the resolution of the debt so that the total value of the debt increases and future growth is not constrained. To summarize:
Under “normal” conditions, the obligations associated with debt are explicit and there is very little uncertainty about how the debt will be resolved. The revenues sources needed to service the debt are clearly identified.
When debt levels become “excessive”, that is when the existing revenues sources are no longer sufficient easily to service the debt, uncertainty arises about how the debt will be resolved and even about the amount of the debt to be resolved. This is exacerbated by the highly reflexive relationship between rising uncertainty and rising debt, so that rising uncertainty associated with the resolution of the debt forces adverse stakeholder behavior, which causes the uncertainty associated with the resolution of the debt to rise further.
How do we know when debt levels have become “excessive”? Debt levels are excessive when the uncertainty associated with the resolution of the debt is high enough to change the behavior of stakeholders. To put it in terms guaranteed to infuriate policymakers, a country has too much debt whenever the market believes it has too much debt. Anyone who does not understand why it is as simple as this does not understand the economic impact of debt.
The purpose of a debt restructuring, then, is to reduce or eliminate the uncertainty associated with the resolution of the debt because this uncertainty automatically reduces value and future growth. If done correctly, a debt restructuring increases the wealth of stakeholders and improves future growth prospects.
Mario Draghi and the 2012 European debt restructuring
This is why I argue that by promising to do whatever it takes, Mario Draghi and the ECB effectively engineered a debt restructuring with implicit debt forgiveness. They changed the contractual terms of peripheral European sovereign debt by implicitly allowing bondholders to put their bonds to the ECB whenever the ECB determined that bond prices had fallen enough to damage Europe more than was tolerable. The fact that both the legal obligation of the ECB to purchase the bonds and the effective purchase price are uncertain does not change the underlying dynamics of the restructuring, and the proof is that Draghi’s statement has caused bond yields to drop to levels that imply a miniscule probability of default.
Notice how the process worked. Draghi’s promised immediately reduced a larger part of the uncertainty associated with the resolution of the debt. The collapse in uncertainty reversed the reflexive process in which rising uncertainty caused declining economic expectations, which caused rising uncertainty. The reversal immediately put the affected sovereign borrowers in a virtuous circle in which declining uncertainty led to improving performance which caused uncertainty to decline further, and while this move by Draghi to reverse the downward spiral was probably necessary, it should reinforce recognition of how intensely pro-cyclical is the dynamic process into within these countries are locked. If it reverses, it will reverse in exactly the same way, and precisely because all market participants know this, even a partial reversal could very quickly intensify.
Ignoring legal and political issues, there are at least two important reasons why Draghi’s promise was far from an optimal resolution of the European debt crisis. First, and most obviously, it reduced the enormous uncertainty that existed at the time, and this is very important because by then it was clear several countries were well along the process in which the debt spirals out of control, but there still remains an enormous amount of uncertainty associated with the debt resolution. Not only are the terms of the implicit “restructuring” uncertain, but in nearly every country in Europe debt is growing faster than debt servicing capacity, most economies are struggling, unemployment is high, the political atmosphere is tense and unstable, and as if all of this were not enough, uncertainty in every country in Europe is linked tightly together so that an increase in uncertainty in any one country automatically causes uncertainty to rise in all countries, in a process often referred to as contagion.
Contagion is often presented as if it were “merely” a psychological problem that can be defused by popular faith, but contagion is far more mechanical than this. One of the imbalances within Europe before the crisis, for example, was the imbalance between deficient demand in Germany and excess demand in peripheral Europe, the balance of which determined the level of the euro. This imbalance has become more complex since the crisis, but it should be clear that if any country on one side of the imbalance were to defect, or were to engineer a rapid adjustment (for example if a peripheral European country were to secede from the currency union), the amount of the imbalance would remain the same but it would be absorbed within a smaller group of countries. If Portugal were to leave the euro, for example, Spain would immediately suffer from having to absorb part of the imbalance that Portugal had absorbed, so that a Portuguese defection automatically has a contagion effect on Spain. The contagion also exists through the banking system (problems in one country cause bank portfolios in others to deteriorate, thus increasing contingent liabilities that must be absorbed by the government) and through the value of the Draghi put itself.
This is the second important reason why Draghi’s promise was far from an optimal resolution of the European debt crisis. European debt is locked into an intensely reflexive process in which the value of the Draghi put is a function of the creditworthiness of the ECB, and the creditworthiness of the ECB is a function of the value of its total obligations, including its implicit guarantees. Any deterioration in the creditworthiness of the ECB undermines the value of the Draghi put, and the resulting increase in the uncertainty associated with the resolution of, say, Spanish debt automatically forces up the Spanish debt burden and so further weakens the ECB’s creditworthiness.
We could perhaps ignore this process and hope that it is never set off except for the fact that in every important country in Europe, the debt burden is worsening (debt is growing faster than debt servicing capacity), so that the ECB’s implicit obligations are rising faster than its debt-servicing ability. The recent McKinsey study of global debt has terribly alarming figures for Europe – of the thirteen countries whose debt has risen by more than fifty percentage points in the past seven years, ten of them are members of the European Union, and the top six include Ireland, Greece, Portugal and Spain, all of whose debt has grown by more than 70 percentage points. The ECB’s creditworthiness, in other words, is deteriorating on a daily basis.
Some analysts deny the possibility of credit deterioration in the ECB because of its ability to create unlimited amounts of currency and fully monetize its obligations. Monetizing debt, however, is not the same as resolving it. It simply involves a transfer of wealth from those who are long monetary assets to those who are short — or, to simplify, from Germany to Spain. If the ability of the central bank to force through these wealth transfers were fully credible, as would be the case of the Fed in the US, the PBoC in China, the BoJ in Tokyo, or the BoE in London, these analysts might be right and the question would never arise.
But because sovereignty is not centralized, the ECB does not operate in the same way as these central banks do. Monetizing the debt involves wealth transfers in the same that gold transfers might have done in Europe in the late 19th century. German’s willingness to allow the ECB to guarantee increasing amount of Spanish debt is probably analogous to the willingness of the Bank of France to lend gold to the Bank of England during a British liquidity crisis if it fully expected the crisis to be resolved and the gold to be returned. It would certainly be willing to take some risk because an English financial crisis would have an adverse impact on the French economy, but there is a limit to the amount of risk it is willing to take, and everyone fully understand this.
In fact I see the credibility of the ECB put as one of the biggest “tail risks” in the market. From the early days of the European Union Germany has traditionally picked up the bill for many of the costs of the European Union, including the costs of some fairly petty European grandstanding, and Germany did it again in a major way during reunification, even though we tend to forget the latter because it involved fiscal and sovereign centralization. This history makes it tempting to assume that the determination of the ECB to monetize the sovereign obligations of all member countries, which is based on the willingness of Germany to allow unlimited wealth transfers, is itself unlimited. But this history should also alert us to the likelihood that Germans are unhappy with this arrangement and might at some point refuse to continue writing checks.
In my February 4 blog entry I argued that while German institutions and policymakers are as responsible as those in peripheral Europe for the debt crisis, in fact it was German and peripheral European workers who ultimately bear the cost of the distortions, and it will be German households who will pay to clean up German banks as, one after another, the debts of peripheral European countries are explicitly or implicitly written down. The overwhelming, and overwhelmingly favorable, response I received makes it clear to me that far more Europeans understand this than perhaps their political leaders want to believe. Among other things this suggests that it does not require lack of solidarity with their fellow Europeans to drive ordinary Germans to refuse to pay for the worsening crisis. It could just as easily be their unwillingness to continue to participate in a process in which workers and middle class households in Europe are being forced to pay to maintain policy mistakes that have benefitted mainly wealthy owners of European assets.
With sovereign debt levels rising month after month in Europe, and with several opportunities for adverse shocks any of which could cause a surge in uncertainty, I don’t think we can simply assume infinite German patience, and so the value of the Draghi put must be eroding, however slowly. I am not saying that I think there is currently any question of ECB credibility, but no institution has infinite debt capacity, and none can support a steadily deteriorating debt burden without risking the possibility of an erosion in credibility. Of course if doubt were ever to emerge about the credibility of the Draghi put, conditions would almost certainly deteriorate much faster than anyone expected simply because of the intensity of the self-reinforcing process that caused bond yields to collapse after July 26, 2012. A system that works powerfully in one direction can work just as powerfully in the other.
What are the functions of the liability side of the balance sheet?
It should be clear for the reasons I discuss above that debt levels in Europe are excessive and that there is a great deal of uncertainty about the resolution of sovereign debt. In order to explain why this uncertainty requires that Europe move to restructure its debt, or, to use a perhaps less politically loaded phrase, to reduce the uncertainty associated with debt resolution, it is necessary to understand how debt — and the balance sheet, more generally, by which I mean not just outstanding government debt but the entire financial system and the financial operations of the economy — affects growth. Balance sheets can exacerbate growth in both directions, so that just as financial crises are always balance sheet crises, there is also almost always an important balance sheet component to every growth miracle.
This needs explaining. For most academic economists, “growth” occurs primarily as a function of the way the asset side of the balance sheet is managed. If assets are managed well, their value increases, and the better they are managed the greater the increase in value, so that the growth of the business or the economy is largely a function of the how productively assets are managed. Policymakers who want to increase economic growth, according to this view, would do so by improving the quality of infrastructure, removing legal and regulatory constraints for businesses, better aligning incentives to encourage productive investment and behavior, and so on, in order to improve the productivity with which the country’s assets and resources are utilized.
Finance specialists, however, understand that the optimal management of assets only provides the upper limit of growth for a business. This is because under certain conditions the liability structure of a business can act as a constraint on growth. If its debt level is high enough that either it distorts incentives by distorting the way value is distributed, or it causes stakeholders to respond to uncertainty by undermining value creation, or it exacerbates the impact of adverse external shocks, not only will growth in the value of the business entity be much lower than it could be, but in fact it can easily become negative. Financial specialists refer to the process — in which the rising probability of default forces stakeholders to alter their behavior in ways that undermine growth — as the incurring of “financial distress” costs.
What financial specialists often don’t understand, however, is that the liability structure of a business can also goose growth above its “natural” level. When it reinforces behavior or exacerbates shocks, periods of growth are often periods of very high growth. This becomes obvious from the observation that it is often the highest of flyers during boom periods that are most likely to collapse or get into trouble as soon as a boom period reverses. At least part of their growth during the boom was not sustainable, but was instead the consequence of a distorted balance sheet that exacerbated the positive shocks typical of a “boom”.
The same process occurs within any economic entity, including a national economy. It is not an accident that in nearly every case in history in which countries have excessively high debt levels or have undergone debt crises, policymakers have never been able to keep their promise that, with forbearance from creditors and the implementation of the right reforms, the country can grow its way back into full solvency. Historical precedents are pretty clear on this point. Countries suffering from debt crises never regain growth until debt has been partially forgiven — explicitly or implicitly — and the uncertainty associated with its resolution has either been sharply reduced or eliminated.
Debt matters. For most orthodox economists debt is simply the way operations are funded, and what matters mainly is the cost of the debt, or at best the cashflows associated with debt servicing. But in fact the liability structure of the balance sheet of any economic entity has several functions that significantly affect economic behavior. This is as true of small economic entities, like households and businesses, as of large, like countries or even the global economy. The main functions are:
The liability structure is the sum of the way assets and operations are funded, and debt can be used either to fund investment or to reallocate consumption over time.
The liability structure affects the ways in which operating earnings and economic growth are distributed. For a business entity, the various ways in which the debt servicing agreements are indexed and their levels of seniority determine how operating earnings after taxes are distributed, with the residual going to owners. It is a little more complicated when we are dealing with a national economy, but the principle is the same. What matters for a national economy is that the government is a hugely important player, and the ways in which it collects taxes, both direct and indirect, the investment decisions it makes, and the policies it implements to enhance or undermine certain kinds of economic activity, whether explicit or implicit, collectively act as the primary determinant of how growth in the economy is distributed among different stakeholders. As a consequence all stakeholders will adjust their behavior in response to their perceptions of how government policies will affect them.
The liability structure also determines how exogenous volatility and external shocks are absorbed, and here is that the concept of balance sheet “inversion” can be especially useful. As I explain in my book, The Volatility Machine, an inverted balance sheet is one in which borrowings are structured pro-cyclically so that they reinforce external shocks by automatically causing behavior to change in ways that exacerbate the impact of the shock (unlike a “hedged” balance sheet, which automatically dissipates shocks). External currency debt, inventory financing, short-term or floating-rate debt, asset-based lending, and margin lending in the stock market, can all be highly inverted forms of borrowing because a rising market allows investors to increase their purchases, which pushes prices up further, and a declining market forces investors to sell, which pushes prices down further.
The structure of a country’s banking and financial system can also either exacerbate or dissipate external volatility, in the former case by imbedding pro-cyclical mechanisms into the economy. In China, for example, because its rapid growth was driven primarily by investment, which tends to be heavily commodity intensive, rapid growth pushed up global commodity prices nearly tenfold during the first decade of this century. Of course business entities that ran the largest commodity inventories profited heavily at the expense of their more prudent competitors, and this caused a “Minskyite” shift in the composition of industry towards greater speculative risk-taking by rewarding commodity speculation heavily. This process of course caused the financial system to become increasingly exposed to commodity price risk, so that the faster China grew, the more rapidly Chinese banks were able to expand credit, which itself caused the economy to grow even more rapidly and commodity prices to rise further. Of course the whole process went into reverse once Beijing, recognizing the deep imbalances that were building up within China, moved to slow investment growth, and we are only now seeing what will turn out to be a long, drawn-out process of pro-cyclical adjustment.
In economies with highly inverted balance sheets, shocks can be so self-reinforcing that periods of rapid growth become growth “miracles” but, as happened in nearly every growth miracle case in history, the subsequent periods of decelerating growth can swiftly degenerate into collapse or lost decades of unexpectedly low growth. In Europe’s case, the intensively reflexive relationship between sovereign creditworthiness and the domestic banking system may be he most worrying form of inversion. For the banks in peripheral Europe a substantial share of liquid assets consists of their government bonds, so that any deterioration in sovereign creditworthiness would cause enormous losses for the banks, but because the main source of contingent liabilities for the government is the banking system, losses in the banking system would immediately cause the sovereign creditworthiness to deteriorate. Balance sheet inversion is the single most important reason for sovereign financial crises, as I explain in my book, but policymakers are not nearly as terrified by it as they should be.
Finally, expansion or contraction in the liability side, which occurs most easily in the form of expansion or contraction in credit, can speed up or reduce expected growth in operating earnings or GDP.
How does a badly structured balance sheet constrain growth?
There is an enormous difference in the level of sophistication between corporate debt restructuring and sovereign debt restructuring. In the discussions in Europe, most of which are dominated by discussions about Greece, there have been suggestions by Finance Minister Yanis Varoufakis as well as analysts like Martin Wolf and Wolfgang Münchau, both at the Financial Times, about aligning debt servicing payments with the Greek ability to pay, for example in the form of GDP-linked bonds. But aside these few, much of the focus seems to be on determining what combination of coupon and principle will lead to the maximum annual payment Greece can manage over the next two or three years.
This is a very bad way to decide on how to restructure the sovereign obligations of Greece or that of any of the peripheral European countries that are likely to follow. The debt restructuring should be designed to maximize stakeholder value so that all players are better off after the restructuring than before. Key to a successful restructuring is the elimination, or at least the substantial reduction, of the uncertainty surrounding the resolution of the debt. I would argue that there are at least five components to a good restructuring and these can best be understood by recognizing the ways in which a badly structured balance sheet can constrain growth:
High borrowing costs.
Obviously, the higher the real borrowing costs, the greater the growth constraint, but there is a very important point that is missed in the discussion. It is always easy for a borrower to reduce its borrowing cost by taking on risk. So, for example, short-term borrowing, debt denominated in a major currency rather than the borrower’s currency, or debt that grants optionality to the lender (for example, by allowing the lender to put the bond back to the issuer on a specified date before its final maturity) will always involve a lower interest rate for the borrower.
This is almost never a good strategy for a borrower that needs to restructure its debt. The biggest risk of lending money to higher-grade borrowers tends to be interest rate or currency risk. To lower-grade borrowers, of course, it is the risk of default, and the pricing of its obligations is very sensitive to changes in the perception of default risk. This means that any reduction in borrowing costs that comes about by the borrower’s taking on additional risk is usually less than the direct or indirect cost of the additional risk on the rest of the borrower’s obligations. A country that is forced to restructure its debt should never agree to lower borrowing costs if this means that it has to assume additional default risk.
But this doesn’t mean that restructuring countries cannot benefit from granting optionality to their lenders. If the optionality results in an overall reduction in balance sheet inversion, the borrower benefits in two ways: it reduces the borrowing costs on the specific debt instrument and, by reducing the probability of default, it reduces the direct or indirect cost of the rest of the borrower’s obligations. How could this work in the case of a Greek restructuring?
It could issue a dual currency bond. Borrowers understand that there is some probability that Greece will exit from the euro, either permanently or temporarily. By allowing buyers of this bond to choose at some point in the future between payment in euros at the contracted rate or payment in the new currency at some rate, either specified today or linked today to some index. Investors would only switch payment options under optimal conditions for Greece, when its currency is expected to do well and interest rates fall, allowing Athens both to reduce its borrowing cost and to stabilize its balance sheet. After the 1994 peso crisis. Mexico issued a dual currency bond whose final payment was linked, at the investor’s option, either to dollars or to pesos. The offering was hugely successful and palpably changed expectations about the Mexican outlook.
It could allow investors to index payments to an asset whose performance is positively linked to economic performance — oil, for example, in the case of Venezuela — so that investors receive a higher-then-contracted payment only in the case that Greece’s economic performance is better than expected. Options linked to GDP or to real estate prices or to the Athens stock market (perhaps even to olive oil or to the number of foreign tourists?) accomplish this purpose.
Because Greece imports a number of commodities, it can also share the benefits of any fall in commodity prices by agreeing to pay a higher coupon if the price of a specific commodity falls below a certain price.
One of the great advantage of these structures is also too little appreciated by issuers, investors, and bankers. Even if the expected payment of an instrument with optionality is identical to that of an instrument without optionality, simple arithmetic makes the former more valuable for the investor even as it is more valuable. This is because there is an asymmetry in the discount rate such that higher expected payments are discounted at lower rates and lower discounted payments at higher rates. If the optionality is linked to a liquid asset and detachable, the investor can immediately capture this value directly.
Financial distress costs.
Earlier on I pointed out that all the costs associated with servicing and paying down the debt are ultimately assigned to different groups of stakeholders. These include workers and ordinary households, small and medium-sized companies, large companies and multinationals, wealthy households, or, especially in the case of countries with very large public sectors, local and central governments through asset liquidation and land sales. Stakeholders can even include foreigners. When uncertainty arises about how sovereign debt is likely to be resolved, all of these stakeholders must alter their behavior to protect themselves from bearing a disproportionate share of the costs. Their effect in reducing growth is what is referred to in finance theory as financial distress costs.
The process is both automatic and straightforward. The ways in which a country incurs financial distress costs include:
Workers understand that unemployment is likely to rise and remain high and so they tend to cut back on spending. They also tend to unionize, and their unions become more militant in their relationship with business.
Ordinary households worry about future income or consumption tax increases, and so they too cut back on spending, and because they also worry that their savings will be confiscated to pay the debt, most usually in the form of inflation, financial repression, currency depreciation, or frozen deposits, they often withdraw their savings from the domestic financial system.
The best educated, the young, and people with the most valuable skills emigrate as the excess debt weighs down future growth prospects.
Small and medium-sized companies, fully aware that during crises wealthy businesspeople often bear the brunt of public anger, worry about being expropriated or forced to pay higher taxes, and so they disinvest and become reluctant to hire additional workers even in the exceptional cases when they need them.
Large companies and multinationals also worry about expropriation and taxation and so disinvest or move operations abroad.
Banks worry about deteriorating collateral values and cut back on risky lending.
Exporters worry about currency depreciation or confiscatory rules on foreign currency, and so they drive down inventory and reduce the amount of earnings they repatriate.
Wealthy households move money abroad to avoid higher income or asset taxes.
Foreign and local creditors reduce loan maturities and raise interest rates.
Policymakers respond to the increase in social and political instability by shortening their time horizons.
These changes in behavior in response to rising uncertainty about how debt will be resolved are probably the most damaging impact of a debt crisis because they erode not just economic institutions but also social and political institutions. They can also be intensely self-reinforcing, so that as stakeholders respond to rising uncertainty, their responses collectively increase debt, reduce growth, and exacerbate balance sheet fragility (by shortening debt maturities, for example), all of which only increase uncertainty. One of the saddest aspects of this process, historically, has been that typically the most sophisticated, who are often the wealthiest, are the first to protect themselves, and the least sophisticated are the last, with the result that the distribution of losses is asymmetrical in a way that maximizes the social damage.
High debt levels often create disincentives for business entities to behave in line with the interests of the economy overall. The most obvious example of this is in countries that have currency and balance of payments crises. The urgent need for foreign currency often results in short-term policies that penalize businesses with foreign currency, including businesses that generate foreign currency. The result is that instead of encouraging businesses that increase the available pool of foreign currency, these policies discourage them. Latin America in the 1980s suffered from this misalignment of interests.
The same occurs when debt restructurings involve complex currency or debt repayment regulations, the rationing of credit, or significant distortions in the price of credit. In these cases business managers are more heavily rewarded not for innovation and productivity growth but rather for managing their way through the regulatory process. For example, when Venezuela had multiple foreign exchange rates during the late 1980s, a significant amount of resources was expended by business managers to allow them access to favorable treatment. In China, until about 3-4 years, to take another example, interest rates for bank loans were so heavily repressed that it was widely admitted that companies that had preferential access to bank loans spent more time trying to justify taking on more loans than they did looking for the most productive ways in which to invest the proceeds. With such low interest rates, almost any investment generated returns that exceeded the cost of funding it.
In Europe the determination to remove any possibility of flexibility in the use of the euro may be creating misaligned incentives. If countries like Spain were permitted temporarily to leave the euro, it would be possible to create agreements in which Spanish incentives were very strongly aligned with European interests. Assume for example, that Spain were permitted to reintroduce the peseta with an agreement that it would return to the euro five years later at a rate that implied a 20% depreciation in the currency, and this agreement was supported by debt extensions and the credibility of the ECB. This would immediately provide short-term relief for the Spanish economy while at the same time creating very strong incentives for Madrid to put into place the productivity-enhancing reforms that would make the currency fully viable once it rejoined the euro.
Exacerbating of default probabilities.
Highly inverted balance sheets exacerbate the impact of external shocks. While positive shocks reduce the probability of default and negative shocks increase the probability of default, the impact is not symmetrical and in fact is highly concave. The result is that the probability of default is always higher when balance sheets are more inverted, and so the return that investors require is automatically higher too.
The impact of deleveraging on growth.
Except in the case where all resources, including labor, are fully and efficiently utilized, and frictional costs are low or negligible, increasing leverage usually results in faster growth and deleveraging in slower growth. Highly indebted countries must urgently reduce their default probabilities, and the easiest and most obvious way to do so is to use free cashflow to pay down debt as rapidly as possible.
What must a good debt restructuring do?
Once we understand why distorted balance sheets and excessive debt burdens raise costs, undermine social, political and economic institutions, and constrain growth, and the various ways in which they do so, there are a number of fairly automatic lessons that can be drawn:
Debt restructuring should not be delayed. This is the first and most common mistake made throughout the modern history of financial crises. The costs associated with an excessive debt burden begin to accrue immediately and include the undermining of social, political and economic institutions necessary for economic recovery. The longer the crisis persists, in other words, the greater the long-term cost to the country. Historically in the case of an external debt crisis there has always been strong pressure from the creditor governments to delay recognition of insolvency if this recognition could force the creditor banks into insolvency themselves. In that case the creditor banks need time to recapitalize before they can acknowledge the losses, and it is only after the banks are recapitalized that creditors finally “discover” that the borrower is insolvent. Given the potential size of the European losses, however, it may take many years of unnecessary economic destruction before German and other European banks are finally healthy enough. 
Debt restructuring must address short-term liquidity constraints. This point is generally well understood and is often resolved in the form of ad hoc institutions, for example the Draghi put in the case of Europe.
The debt restructuring should be designed not to reduce the value to creditors but rather to maximize value to stakeholders. The exchange of fixed payment instruments for equity-like or variable-payment instruments in which payments are indexed to economic well-being, for example GDP-linked bonds or coco bonds, can raise the total value of debt instruments significantly while boosting growth.
It must assign the costs to those best able to bear them and whose reduced income will have the least negative impact on future growth. In most cases the resolution of debt is done implicitly over a long time, and usually in the form of financial repression or explicit taxes, so that the costs are effectively assigned to middle class households. This tends to worsen income inequality and weaken consumption. One of the goals of a debt restructuring should be to ensure that the costs are assigned in a way that is least damaging to the economy over the longer term.
It must reduce the impact of external shocks, mainly by indexing payments to revenues and wealth creation, so that the probability of payment difficulties declines.
More generally it must reduce uncertainty about resolving the debt. This is the main reason for a rapid and explicit debt restructuring because it is the uncertainty that undermines the economy.
It must line up incentives so that foreign and domestic stakeholders are rewarded for productive behavior and policymakers are rewarded for long-term policies.
These are not necessarily easy or obvious things to do, but the important point to understand that is that with or without an explicit restructuring, the cost of resolving the debt is going to be borne by different sectors of the economy. The main difference is whether this happens explicitly or implicitly and whether it happens quickly, or it is allowed to drag on for many years, creating conditions of low growth and high unemployment and causing damage to the social, political and economic institutions that will generate growth over the future.
In many countries in Europe there is tremendous uncertainty about how debt is going to be resolved. This uncertainty has an economic cost, and the cost only grows over time. But because most policymakers stubbornly refuse to consider what seems to have become obvious to most Europeans, there is a very good chance that Europe is going to repeat the history of most debt crises. After many years of denying that they are insolvent, and many years of promises that reforms will be implemented that will set off enough growth to resolve the debt, policymakers in countries like Spain will be forced either to change their positions or they will be forced out by voters – simply because economic conditions will have deteriorated so drastically that a restructuring can no longer be put off.
Monetary policy is as much about politics as it is economics. It is about some of the ways in which wealth is created, allocated, and retained. Debt restructuring involves allocating wealth in the most efficient way. It does necessarily not mean, however, defaulting on payments. The only goal of a debt restructuring is to reduce the uncertainty associated with the resolution of the excessive and growing debt burden. There are many ways to do so, and in many cases they require significant debt forgiveness, but pretending that all will be fine if we only grit our teeth and wait longer has almost never turned out to be true.
For now I would argue that the biggest constraint to the EU’s survival is debt. Economists are notoriously inept at understanding how balance sheets function in a dynamic system, and it is precisely for this reason that we haven’t put the resolution of the European debt crisis at the center of the debate. But Europe will not grow, the reforms will not “work”, and unemployment will not drop until the costs of the excessive debt burdens are addressed.
Academics, journalists, and government officials who want to subscribe to my newsletter, which sometimes includes portions of this blog and sometimes (as in this case) does not, should write to me at email@example.com, stating your affiliation, please. Investors who want to buy a subscription should write to me, also at that address.
 In Carmen Reinhart’s “A Distant Mirror of Debt, Default, and Relief” (NBER Working Paper No. 20577, October 2014) the author looks at 42 default and restructuring episodes in the 20th Century and finds that “the post final debt reduction landscape is characterized by higher income levels and growth, lower debt servicing burdens, lower external debt, sovereign credit ratings and capital flows behaved differently in the interwar and modern periods; in the latter case ratings recover markedly.” It is only after the debt forgiveness, they find, that growth picks up.
 In the case of the LDC Debt Crisis of the 1980s, American banks spent most of 1982-88 rebuilding their capital, in part thanks to large hidden transfers from American households spurred by a steep yield curve. It is not a coincidence that two years Citibank announced its reserve position, in May 1987 (within a year of which the weakest of the top ten US banks were also sufficiently reserved) that the first formal debt forgiveness was negotiated, for Mexico, with the issuance of the first Brady bond in 1990. In the German case I expect that part of the debt will be gradually nationalized (as is already happening) while Berlin will take steps to boost bank profitability at the expense of middle class German savers in order than banks can write down the remainder. It is only then that we will see widespread political support of debt restructuring and partial forgiveness, if history is any guide.
© Michael Pettis, 2015, reproduced with permission, http://blog.mpettis.com/2015/02/when-do-we-decide-that-europe-must-restructure-much-of-its-debt/