Friday, February 3, 2012

Sargent on debt and defaults

Tom Sargent's Wall Street Journal oped is well worth reading closely. It's a very short summary of his Nobel prize speech

As readers of this blog will probably know, I think Europe should stop bailing out bondholders of Greek and other debt. (See the Euro collection and Euro tags to the right.)

"What about Alexander Hamilton?" has always been a nagging doubt.


Hamilton famously brokered the deal by which the Federal Government assumed state debts. By doing so, he created a group of citizens with a strong interest in the success of the Federal Government. But it was a bailout of the states; it was a bailout of their creditors, many who had bought up debts cheaply. It was explicitly a case of greater "fiscal union."  Perhaps this is Europe's Hamilton moment?

As Tom points out, there are some important differences. The states had borrowed money to fight the revolutionary war, not to import Porsches or build cozy crony economies and fat welfare states. U.S. Taxation was low everywhere. Even the new Federal taxes were only tariffs, amounting to 2% of GDP, not 50% and up taxation in the Eurozone.  The Federal debt ("Eurobonds") was backed by directly levied Federal taxes, not by voluntary contributions or even by remittances from member states. And those direct taxes were to be legislated by a directly-elected legislature, not Brussels technocrats.

Tom points to a second episode: the state defaults of the 1830s and 1840s. Here, many states had borrowed a lot to finance infrastructure projects ("canals to nowhere?") that were not generating enough revenue to pay back the debt. 

Reputation, pre-commitment and moral hazard are big in Tom's thinking and his account of the sophisticated thinking of our ancestors. The US chose to pay off its revolutionary war debt, according to Tom, to enhance its reputation and credibility as a serious nation and future borrower. But this decision led to moral hazard: states and their creditors believed the US would always bail them out. The US chose not to bail out the states (really, their creditors) the second time around. It suffered a financial crisis as a result, but put state overborrowing and default off the table for a hundred and fifty years.  (When Tom talks about reputation and pre-commitment, he's not blowing smoke; he understands the equations.)

This second episode strikes Tom as a better antecedent to the Eurozone. Let Greece and the others default precisely to save the euro and European union.  Now is the time to clarify that the no-bailout clause is real, and that the euro will not be inflated. A crisis is the price of not having been clear about that moral hazard up front. But the union project is too important to abandon.

There are lots more  little gems in Tom's paper. The interaction of monetary and fiscal policy is one; the fact that framers spent all their time on fiscal policy, and money was, as in the constitution, considered just part of the definition of weights and units.

This post is a bit interpretive, and I'm sure I put some words in Tom's mouth here and there. He's always scholarly, informative and precise. If so, well, there's no substitute for the original. 

Important Note: The comments section is running off the rails. Please be polite and stay on topic. You're welcome to disagree, or point out flaws in my arguments or facts -- actually I like that, as I learn something. But if you want to spew venom, go back to Krugman and DeLong's blogs. I'm going to turn off comments if this doesn't end asap.

32 comments:

  1. I don't have much to say on this post since I pretty much agree with Sargent, mostly because I don't see any plausible alternative. I've heard some people throw around this idea of "synthetic eurobond," and I'm still trying to think what I make of that.

    Also, your blog is new so you don't have a lot of comments yet, you could just moderate the comments rather than close them altogether.

    ReplyDelete
  2. It is an interesting Thought, What if the Constitution had been rejected and we had continued with the Articles of Confederation?

    I am not so sure that we would not be better off today. The profligate states would have been forced to change, The southern states would never have seceded, and eventually slavery would have ended because it was becoming economically nonviable with mechanization.

    We would have avoided a over powerful central government, but no doubt the American states, due to their basic nature would have been able to come together for mutual defense when threatened.

    ReplyDelete
  3. Excuse my ignorance of international finance (i'm a law student, not an economist), but why exactly did credit markets ever lend to Greece at low rates in the first place? It couldn't have all been bad accounting, could it? Did financial markets just assume that Germany was bluffing about the "no-bailouts" clause?

    ReplyDelete
    Replies
    1. lenders, especially commission paid lenders (Buffett/Munger hire for life and guarantee a handsome income to avoid such agency problems) do not make good loans, they make the least bad loan

      Delete
    2. It helped a lot that many of those lenders were banks, and were allowed to treat Greek government debt as risk-free, requiring no extra capital. They also could use the greek debt as collateral, i.e., borrow from the ECB and lend to Greece.

      The bailout perception was certainly part of it. However, joining a currency union also means that default will be much more costly than the old fashioned way out by devaulation. Making it more costly to welch on a debt makes you more credit worthy in the first place. It sort of ties you to the mast to repay.

      Delete
  4. I like reading what you have to say and hope you don't pay too much attention to the nonsense in the comments section. Maybe it's better to disable it since people are going to be directed here from the Krugman and DeLong blogs anyway.

    ReplyDelete
  5. But can the euro even be saved? I like to think of the issue that Greece is facing is more about low productivity and still being forced to use the euro. To become competitive in the face of low productivity, wages and prices must fall, or else the country's currency must depreciate. But nominal wages generally exhibit downward rigidity -- how often do employers offer to pay workers less this year than last and succeed? Prices tend to be somewhat sticky as well (New Keynesians and RBC people can argue this point). Therefore, the only way to increase exports is to have the currency depreciate. But Greece is stuck in the euro, so that channel can't work. So I don't see how Greece can save itself without exiting the euro. Or am I missing something?

    ReplyDelete
    Replies
    1. Tying such wildly differing nations to the same currency was idiotic to begin with. But do not think they are tied forever to the Euro. If the political situation continues to deteriorate in Greece you might see a party take power that will leave the union.

      Delete
    2. I am not at the level of the people here, but I have read somewhere that this kind of rigidity is just nominal. I mean, besides he resistence to implement a decrease in salaries, the purchasing power can vary in time.
      The idea is that US$ 1.000,00 in 1950 had not the same purchasing power of the exact same amount of money in 2012.
      If internal prices change, the shifts in the demand and supply curves happens besides nominal downward rigities. Once in this scenary the same nominal salaries can buy less and less, we have the same effect of a nominal devaluation of salaries.
      It may sound insane and perhaps I misunderstood completely what is being said in this post, but if I understood right the main idea is that nominal rigidity means the possibility of real devaluation. I mean, when we change how many products one Euro can buy in Greece, didn't we changed the real interst rate?
      A default has the same effect once a country cannot simply refuse to pay its debt. At some point it has to pay it due law enforcement. I have seen cases in Brazil where it took a very long time but it has to be payed due the fact that acquiring a debt means a legal contract that can be charged in justice courts. Which means that if I am not making a huge mess, a default means the transference of a current liability (which has to be payed now) to somewhere in the future. It is a limit unpalatable measure that can be used in limit situations.
      The main risk of such policy is the implementation of an inflation culture ( a good example is the adoption of automatic wage corrections, which can make inflation spiralize and turn into hyper inflation - which is nightmare!).
      Did I get everything wrong?

      Delete
    3. Sorry, I whent to reread what I wrote and had to correct it:

      I am not at the level of the people here, but I have read somewhere that this kind of rigidity is just nominal. I mean, besides he resistence to implement a decrease in salaries, the purchasing power can vary in time.
      The idea is that US$ 1.000,00 in 1950 had not the same purchasing power of the exact same amount of money in 2012.
      If internal prices change, the shifts in the demand and supply curves happens besides nominal downward rigities. Once in this scenary the same nominal salaries can buy less and less, we have the same effect of a nominal devaluation of salaries.
      It may sound insane and perhaps I misunderstood completely what is being said in this post, but if I understood right the main idea is that nominal rigidity means the possibility of real devaluation. I mean, when we change how many products one Euro can buy in Greece, didn't we changed the real exchange rate?
      A default has the same effect once a country cannot simply refuse to pay its debt. At some point it has to pay it due law enforcement. I have seen cases in Brazil where it took a very long time but it has to be payed due the fact that acquiring a debt means a legal contract that can be charged in justice courts. Which means that if I am not making a huge mess, a default means the transference of a current liability (which has to be payed now) to somewhere in the future. It is a limit unpalatable measure that can be used in limit situations.
      The main risk of such policy is the implementation of an inflation culture ( a good example is the adoption of automatic wage corrections, which can make inflation spiralize and turn into hyper inflation - which is nightmare!).
      Did I get everything wrong?

      Delete
  6. You are being overly sensitive to the comments. I do not think I have seen a comment directed at you personally.

    There is a wide range of opinion on the issues of economics, government and regulation that you touch on. It is unrealistic for you to expect that everyone else will approach those issues in the same way that you do. People are going to respond to the aspects of an issue that is important to them and will do so within their own frame of reference and vocabulary.

    If you think that the comments you are getting are venomous then I suggest you need to spend some more time exploring just how wide a range of views is held and strongly held by intelligent, well meaning people.

    ReplyDelete
  7. I think I disagree---I think a nation or state should never, ever default on debts. It is better to print money than to default on debts----that's why sovereign nations need the power of the printing press.

    The problem with Greece is that it cannot print more money, to pay off debts, and that would also lower exchange rate of drachma, boosting exports and tourism. This is a much smoother solution than what is going on now.

    I am not recommending reckless fiscal spending for any nation, whether on social welfare or parasitic military coprolite. But a sovereign nation should control its currency, or else it not sovereign anymore.

    ReplyDelete
    Replies
    1. What is the practical difference between defaulting, and printing money to pay off debt? In either case, the creditors get shafted. Inflation isn't a sustainable way to pay debt, because no one wants to lend money at a real loss after inflation.

      If Greece were to abandon the Euro and attempt to inflate away their debt, the result would be a proportionate decrease in standards of living. Alternatively, if they were to dramatically cut government spending, and allow the market to allocate those resources, the result would be new growth, additional productivity, and ultimately higher standards of living. This is a much smoother solution than either the litany of bailouts, or printing money.

      Also, none of the Eurozone countries "control [their] currenc[ies]," but they've maintained the essential components of sovereignty.

      Delete
    2. A default could cost the creditors everything. Paying off the debt through seigniorage would not necessarily cost the creditors very much. It would still be undesirable from the creditors' perspective, but getting 95% of the their money back is better than a complete default. However, in this case, the debt is denominated in euro. Going off the euro and inflating the drachma would have no effect on the value of the debt, because the seigniorage would devalue the drachma, not the euro. Greece has no ability to devalue the euro.

      Some inflation might help the Greek economy, though. With unemployment close to 20%, it is clear that real wages are too high. Much of the unemployment comes from labor laws that make hiring a very risky decision. However, a little inflation (say 4%) would bring down the real wage enough to remove some of this surplus labor.

      Delete
  8. Just noticed the new graphic, funny!

    ReplyDelete
  9. John, I agree with all what you and Thomas Sargent say.
    But...What do you say to the crowd over at VoxEu.org, like Charles Wyplosz who are shouting from the rooftops that the ECB should print almost an unlimited amount of money and fund the European Bailout Fund? (See his letter to Jens Weidmann at VoxEu.org.)

    I am asking, because Wyplosz and others over there (and others here in the US, some people at MIT for example), they are not stupid, they *DO know the equations as well* as Thomas Sargent and you do. (Wyplosz teaches at the Grad Institute in Geneva.) There are several others like him, who *also know the equations*.
    Thus, I ask again: what do you say to them?

    Here is the thing: their argument is that a bailout is LESS damaging in "social cost" than a full fledged bankruptcy of Greece and others; because the "bailout social cost" can be dealt with "later", but a full fledged bankruptcy would have catastrophic social costs that societies would not be willing to bear, and thus, the bailout is cheaper in terms of social cost.
    So, I ask again and again, what do you say to them? Would you answer with equations of time consistency and moral hazard? They would just laugh. And these are serious people (again, MIT, Geneva, etc, not community college economics).

    ReplyDelete
    Replies
    1. So what is the equation for "bailout social cost" and "dealt with later"?

      and couldn't John just say the same? Isn't any such "equation" based upon a series of subjective calls and measurements based upon thousands, perhaps tens of thousands, of social metrics that no "equation" could ever reliably "measure" unless it is featured on Saturday Night Live?

      I don't dispute the need for such analysis but to quantify it into anything such as "equation" is an over-reliance and dependence on the "serious-people".

      Delete
    2. MAS,
      not sure what you mean.
      There are models of moral hazard that measure the social cost, the welfare loss, of moral hazard behavior. Any micro text gives you examples; David Kreps has a nice one in his textbook.
      With your commment on "social metrics" and so forth, you just threw out basically all of econ research published in all relevant econ journals in the last 55 years. Ok fine. This means that for you, economics has not much interesting to say, and you will never be convinced of the contrary.

      Delete
    3. I am convinced economics has a lot of interesting things to say, I don't think I'd be on this Blog otherwise.

      I am just trying to question the precision of adding up the metrics on column A and compare them to the metrics on column B and we have a winner when discussing something as complicated as bailout v. bankruptcy. I know you aren't trying to simplify the problem to that level but when I hear talk of equations for infinitely complex problems it makes me weary.

      A model is only as good as the predictive inputs the modeler can make. I assume you would agree though that there are lots of subjective judgment calls in those equations as we have sparse examples of previous behavior.

      My further point is couldn't John just take the same model make slightly different judgment calls (equally cited and researched) and come up with the opposite result?

      I would be interested in a essay for a layman to read if you have a link to one. Thanks.

      Delete
    4. MAS,
      couple of comments.
      As for equations and math, well, modern econ uses them, basically since Paul Samuelson published his Foundations of Econ Analysis (there was math before, like Cournot's model on oligopoly, Walras, von Neumann, but the audience was much more restricted). It was Samuelson that formally introduced math to econ.
      [BUT, as a side note...some time ago, there was a discussion in the profession on the usage of math in the major departments. The American Economic Association even got involved and published reports on this; I think this was 15/20 years ago. Some people worried that grad students could solve the most complex dynamic stochastic general equilibrium models, but could not answer the simplest questions like what happens to the price of a haircut when the supply of scissors falls. So, your worry has been present in the profession.]

      Predictive power: sure, economists work on their models all the time; they try to adapt to new situations, introduce new variables, and make them more "real world like". Look at the whole DSGE literature, for example (and not to speak the Game Theory literature).
      Take a look at Stephen Williamson's blog today:
      http://newmonetarism.blogspot.com/2012/02/why-economists-are-right.html, where he discusses an article by David Levine about rational expectations in the Huffington Post.

      As for a layman essay, you have to be more specific what you are looking for. My suggestion is this: browse the Journal of Economic Perspectives (http://www.aeaweb.org/jep/index.php).
      This journal is published by the American Economic Association (our prime association for economists), and all it does is to publish articles for laymen, in non-techical format. Browse this journal, I think you will find it interesting.

      Hope all this helps.

      Delete
    5. Regarding the essay, I meant the moral hazard models and their success if they've been empirically tested?

      Delete
    6. MAS,
      there is a (relatively small) literature that tests econometrically the presence of moral hazard behavior in different situations. (Same thing with Adverse Selection.) It is a fairly recent literature, though.
      If you Google "moral hazard empirical evidence", some papers do come up.

      Delete
    7. Thank you. I took a quick review of "Testing for Asymmetric Information in Insurance Markets" The Journal of Political Economy, Vol. 108, No. 1. (Feb., 2000).

      I didn't have time to digest the entirety of the article but I think the first line of the introduction to the work supports the idea that we lack sufficient information to make judgment calls on "equations" measuring social costs of bankruptcy v. bailouts. To wit:

      "In the last 20 years, contract theory has developed at a rapid pace. But, until recently at least, empirical applications have lagged behind." and "This lag between theory and empirical applications probably has several explanations. One is the scarcity of adequate data sets that offer a large sample of standardized contracts for which performances are recorded.' However, insurance offers a promising field for empirical work on contracts.

      and they are focusing on insurance contracts which are highly standardized and recorded meticulously by insurance company actuaries.

      So if economists haven't even tested their own theories on moral hazard and asymmetric information in the insurance industry until 2000, because of the lack of data, then how could any test for the social cost of bailouts v. bankruptcy carry strong predictive correlation other than the judgment of the modeler.

      But, I'm not saying the theories/equations don't have a use, I'm saying we shouldn't unjustifiably rely on them because there are 1) a large amount of subjective judgments that go into these models, 2) practically infinite amount of data points that could go into such an equation and3) the lack of any empirical testing for such models.

      BTW - thank you for your replies. this is very interesting to me. hopefully I'm not boring the rest of the crowd.

      Delete
    8. MAS,
      The asymmetric info literature is relatively recent. As far as I know, the earliest paper on asymm info is Arrow's 1963 paper on medical insurance, where he posed the problem. It lay dormant, then came Akerlof 1970 with this lemons paper, and then slowly the theory body was developed more and more over the next two decades or so. Only then, since the early 90s people started looking for ways to start testing.
      This is how research evolves. Models are proposed, then rejected or modified, improved, changed, made more "realistic", etc etc, until some generalized view arises that this is the way to go. Is there "subjective judgement" in modeling? Sure. But, as they say in some quarters, "it takes a model to defeat a model".
      Yes, economists do it with models (sounds like the econ blog of the same name...). Models that try to distil the unnecessary stuff, and only keep the "necessary" to make a point. Is it subjective? You bet ya it is. However, that's how it works.
      Lack of empirical testing? That comes, sooner or later. In the case of asymm info, as I mentioned, the literature is new, and thus, there is not too much out there, but the literature is growing.

      Delete
  10. Jason-

    As a creditor, wouldn't you prefer the gradual reduction of bond value through inflation to an uncertain settlement in court or by political body? Right now, no one knows what Greek bondholders will get. 20 cents? 60 cents? You tell me.

    And, if inflation was the solution, you would have the option of cashing in your bonds for drachmas and going to Greece to spend the proceeds.

    A hysterical obsession with gold, currency and inflation is not the foundation of pro-growth monetary policies. As the situation in Japan clearly shows, indeed, in a modern economy a zero inflation rate and subsequent near zero interest rates leads to epic economic quagmires. No way to stimulate, except by QE (monetizing the debt, that is).

    Japan is a horrifying example of how monetary authorities can asphyxiate an economy--don't forget, Allan melter, Ben Bernanke, Milton Friedman and John all told the Bank of Japan to go hard and heavy to QE. If you get more conservative than those fellows, you wear jodhpurs and jackboots.

    Voices have been silenced now in this regard, as it is no longer politically correct to advocate monetizing the debt. Ron Paul is braying about the gold standard. The right-wing has dropped into an abyss when it comes to sensible monetary policy, and the left-wing is calling for more fiscal deficits. Dark times.







    I find the right-wing is afflicted by a peevish fixation on inflation. This is hampering sensible monetary policies.

    ReplyDelete
    Replies
    1. Benjamin - The problem with inflation as a fix for Greek investors is that it creates winners and losers far beyond Greece. The losses from global distortion in capital markets could exceed the loss in the Greek markets that you are trying to paper over.

      Better that Greece tell the European banks that they are not going to pay and that the loses fall on the shareholders and money market investors who should have been paying closer attention to Greek creditworthiness.

      Delete
  11. Absalon-

    It is true that Argentina rallied after defaulting on bonds, but I still think it is a moral imperative that nations honor debt payments, at least nominally. I think a nation should control its own currency--it is part of being a sovereign.

    In modern nations, there has always been some inflation, and it never really hurt much. Only deflation, or inflation well into double digits seems to cause any harm.

    If Greece could have run even 8 percent inflation for a several years, it would have boosted its domestic output and deleveraged enough to meet debts.

    Putting an economy into a recession is always the worst choice, and the one they are making.

    ReplyDelete
    Replies
    1. I sort of agree with you, but all this talk about inflation is besides the point. In the first place they are tied to the Euro, in the second place their debts are so high It would take more than single digit inflation to really help. And finally, there is another thing that could be tried which never seems to be tried. Stop spending so much.

      Delete
    2. As a saver who lived through the 70s I disagree.

      What moral imperative could possibly say you have to pay your debts but it is acceptable to pay those debts with counterfeit, worthless paper? Your inflation is just calling for a more limited but still deliberate debasement of the currency.

      The only way to save Greece is to cut the standard of living in Greece (partially through the expedient of enforcing existing tax laws, partially through legislated roll backs of wages and pensions) a forced write down of the national debt and terming out the balance of the debt at low interst rates. The alternative is leaving the Euro followed by significant inflation which will take longer and cause more ultimate pain.

      Delete
  12. "What about Alexander Hamilton?" has always been a nagging doubt.

    Seems the writer has more than a little self doubt.

    ReplyDelete
  13. Absalon:

    In fact, we seem to have global gluts of capital now. Savers are no longer precious.

    I never bought the "inflation is theft" argument, anymore than I buy the "deflation is theft argument."

    If both arguments are true, then we must maintain absolutely zero inflation all through time. While services and goods radically evolve---how to measure the value of digital photographs? We will tie ourselves into knots with such a futile policy--and a policy that suffocates the economy, as shown in Japan.

    When one buys bonds, one takes risks, same as the homebuyer, the stock buyer, or any other investor. We must run monetary policy for the greater good, not for any one investor class. There is no sacred investor class.

    Better to run moderate inflation, for a variety of reasons. I prefer balanced federal budgets, and robust monetary stimulus when necessary.

    ReplyDelete
  14. Both cases are not particularly good, since debts were not in a domestic paper standard (fiat money). Most State debts, when they were nationalized in the 1790s, and when they defaulted in the post-Jacksonian crisis, were tied to the gold standard and the pound. A foreign currency. Europe prints its own money, and they could do QE like the US Fed.

    ReplyDelete

Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.