Monday, July 25, 2011

How I Learned to Stop Worrying and Love Default

If the debt-ceiling negotiations (summarized here) drag on to the point where there is real doubt about the full repayment of Treasury securities, would that be a disaster? Or could it actually raise output and employment?

Nick Rowe has a very smart argument for the latter, on straightforward Keynesian grounds:
Take the standard ISLM model... Draw the LM curve horizontal; either because the central bank conducts monetary policy by setting a rate of interest, or because the economy is stuck in a liquidity trap where money and government bonds are perfect substitutes. Now let's hit it with a shock.
The shock is that all the bond rating agencies downgrade the rating on government bonds. Specifically, the rating agencies say that whereas the previous default risk was zero, there is now a 1% chance per year that the government will renege on 100% of all promises to pay money on its bonds. (Or a 10% chance per year of reneging on 10% of repayments, whatever.) And assume that everybody believes the rating agencies. Further assume that the central bank holds the interest rate on government bonds constant... What happens?
My answer to this question is the same as the standard textbook answer to the question where the shock is an increase in expected inflation by 1%. ... And if you: did believe in the ISLM model; and did believe that the economy was stuck in a liquidity trap; and did believe the economy needed an increase in Aggregate Demand; and did believe that fiscal policy either should or could not be used, for whatever reason; you would say that this increase in expected inflation is a good thing. At first sight, the answer to the question "what is the effect of a 1% increase in perceived risk on government bonds?" is exactly the same as the answer to the question "what is the effect of a 1% increase in expected inflation?". ... Both result in the same 1% fall in the real risk-adjusted rate of interest on private saving and investment and the same rise in AD and real income.
This is one of those points that seems bizarre and counterintuitive when you first hear it and completely obvious once you've thought about it for a moment. Suppose a bond already pays zero interest; how can its yield go lower? One way is for inflation to get higher, so the principal repayments are expected to be worth less. But another is for the default probability to rise, which also reduces the expected value of the principal repayments. At the level of abstraction where a lot of these debates happen, for important purposes the two should be identical. If you believe the zero lower bound story -- that monetary policy would have brought unemployment down to normal levels by now, if it were just possible to reduce the federal funds rate below zero -- then you should support a (temporarily) nonzero default risk on government debt for the exact same reason that you support (temporarily) higher inflation -- it creates the economic equivalent of negative interest rates.

There are lots of caveats in practice. (There are almost always lots of caveats.) And if you're a ZLB skeptic -- if you don't believe even a negative interest rate would be effective in boosting demand -- then default risk won't help much either. But analytically, it's still an important point. Among other things, it helps explain why the threat of default has not moved the price of Treasury securities at all.
I'd assumed, up until now, that it was because asset owners took it for granted that the debt ceiling would, in fact, be raised, or that if not debt payments would be prioritized over everything else.

I still suppose that's true. But here's a more general reason. Another way of thinking of the zero lower bound phenomenon is that the government's commitment to issue zero-interest liabilities in the form of cash and reserves sets a ceiling on the price of its liabilities (remembering that price and yield move inversely.) This price ceiling means there is excess demand. So if you have some exogenous factor that would normally lower the price of Treasuries, it doesn't do so; at the margin, it just reduces the backlog of frustrated buyers at the current price.

Cool.

UPDATE: And here, right next to the Rowe piece in Google Reader, is an FT Alphaville item about how settlement failures (not delivering a security at the date contracted) seem to be becoming increasingly deliberate in secondary markets for Treasuries, as a form of "unconventional financing." It's not an exact analogy, but there's an important parallel: In private financial markets, when an interest rate is stuck at zero, how completely a debt is honored becomes the natural margin on which terms adjust.

11 comments:

  1. And here in your sidebar, right next to your post, TripleCrisis has this post title: "US debt impasse worries the world".

    So what about all the concern about "confidence"? That's thinking like an economist of a different sort, I suppose...

    Interesting post.

    ReplyDelete
  2. Arthurian-

    The main argument in the TripleCrisis post is that default would be bad news for poor countries that hold Treasuries as foreign-exchange reserves. Maybe so, maybe not -- but to the extent it would be, so would be higher US inflation. So if you think the US would benefit from higher inflation right now -- as do Keynesians from the mighty Krugman to little me -- then you should think it would benefit from a higher default risk too.

    And honestly, I think if the TripleCrisis writer had thought of this argument -- which is no criticism of him; I don't think anyone had thought of it before Rowe -- he would stop worrying about default too. Even the countries with the biggest holdings of Treasuries need to worry a lot more about their export earnings than about capital gains or losses on their reserves. So faster US growth is worth a little inflation/default risk to them too.

    ReplyDelete
  3. I think this argument is a bit twisted.
    IMHO people with money usually ("capitalists?") buy treasuries as "safe heaven" investment; the government uses the money obtained this way to fund works, employ people, pay social programs etc.; this government spending then creates "inflation in wages", that is usually measured by core inflation.
    But, if treasuries are no more safe, people with money will search for other forms of "safe heaven" invesment, such as gold, Picasso paintings and so on. Those goods are usually non reproductable stuff, and as a consequence increased demand in those things doesn't translate in higer wages to anyone.
    On the whole, I think that the argument for higer inflation is that nominal wages should rise with respect to nominal financial assets (debt). Things that are equivalent or similar to inflation aren't useful unless this increase of wager with respect to debt doesn't happen.

    ReplyDelete
  4. I love the title of this post, and as someone that struggled over n days to transmit the value of the ISLM model as a heuristic (I hope I'm using that term correctly) to two batches of intermediate macro students, it warms my heart to see you use it for this application.

    -Nina

    ReplyDelete
  5. So, losses of paper wealth have no impact on people's decisions? So, people just acquire, accumulate wealth for no reason? Only a dyed in the wool mainstreamer would find this not absurd.

    In an economy where the private sector is deleveraging, you need the public sector to run deficits to ensure that the private sector enjoys gains in net wealth and therefore maintains/increases spending. Defaulting on bonds may do wonderful things for inflation expectations, but it is not going to make people rush to the mall. It is going to make them rush to wherever they can preserve their wealth, creating a Latin America circa 1980s scenario.

    A side note, in a non-fiat world, the distinction between money and government bonds is less important than the distinction between money+government bonds and all other assets. I have mentioned this on Nick Rowe's blog, he apparently does not seem to understand.

    ReplyDelete
  6. RL-

    Good points. You're right, the most important effect on inflation is that it reduces private-sector debts relative to private-sector incomes and thereby redistributes wealth from creditors to debtors. Default risk doesn't do that at all. However, that's not the mainstream view, as embodied in ISLM. That is rather about incentives -- wealthholders' choice between holding public and private liabilities at the margin. And if further demand for government securities can't increase government spending (either because of the ZLB, or because of political constraints, or both) then shifting asset demand to any other safe haven has to increase AD. Partly because the elasticity of supply of gold, paintings, etc., while low, is never exactly zero; and more importantly because the capital gains to existing owners of those assets will lead them to increase current consumption. A new asset bubble is certainly preferable to years of ~10% unemployment.

    ReplyDelete
  7. Nina-

    Thanks! I'm teaching intermediate macro myself this fall; need to talk to you about your experience with it at some point soon.

    Anonymous-

    people just acquire, accumulate wealth for no reason? Only a dyed in the wool mainstreamer would find this not absurd.

    What matters -- this is genuine Keynes as well as ISLM -- isn't the decision to acquire wealth as such. (Which, by the way, I do think is undertaken for more or less no reason. My late friend Bob Fitch used to quote Nietzsche: "We must not ask the money-making banker the reason for his restless activity. They roll as the stone rolls, according to the stupidity of mechanics.") What matters is the choice between wealth in the form of public liabilities (i.e. money, including Treasury bonds) and wealth in the form of private liabilities. When people become more uncertain and/or pessimistic about the future, they prefer the safety of public debt to the higher potential yield of private debt; this makes it harder to finance real activity and reduces output and incomes. The goal of monetary policy -- in the postwar Keynesian view embodied in the ISLM framework, which is partial but certainly not all wrong -- is to shift asset demand back toward private liabilities. Normally this can be accomplished simply by lowering yields on public debt, but if that's not possible, then alternative forms of policy become necessary, like QE or higher inflation expectations -- or higher default expectations.

    the distinction between money and government bonds is less important than the distinction between money+government bonds and all other assets

    Maybe I'm confused, but it seems to me that grouping Treasuries and money together as federal liabilities only strengthens Nick's argument. Why do you think it's a problem for it?

    ReplyDelete
  8. Defaulting on bonds may do wonderful things for inflation expectations, but it is not going to make people rush to the mall. It is going to make them rush to wherever they can preserve their wealth, creating a Latin America circa 1980s scenario.

    I think a less dramatic way of saying this would be, It will lead to a lower dollar. Which, leaving aside whether forex markets are that reliable, isn't obviously a bad thing.

    Latin American wealth holders could flee to the US. American wealth holders can flee to ... where?

    ReplyDelete
  9. JW,

    Yes, people acquire wealth for "no reason. The acquisition of wealth is in an end in itself. Besides, it confers relative status. That does not mean it is irrelevant for the economy. The destruction of paper wealth induces those who suffered losses to redouble their efforts to recoup their losses--resulting in a desire to net save. If the government does not run deficits to accommodate that desire to net save, you will get a depressed economy.

    On the government bonds and money issue, I don't think i am with Nick. He assumes that an increase in future inflation expectations is the same as a reduction in current nominal wealth. Needless to say, this would not hold in a world with debt denominated in nominal terms. in short, as long you have balance sheets in nominal terms, abstracting away from them and using the IS-LM analysis is bound to be futile. And you cannot think about Depression economics without debt.

    Yes, US wealth holders would not have the same options as Latin American wealth holders. However, that still does not mean that they are going to go and splurge. Just take a look at what inflation did to the saving rate in the 1970s. Yes, it drove up land prices, but do you think that is even a realistic possibility right now? What such recklessness will do is create tremendous uncertainty and instability.

    ReplyDelete
  10. Hi Anonymous.

    First, not be a jerk, and I really appreciate your comments here, but could you use a handle of some kind? There are potentially a lot of Anonymouses and it leads to unnecessary confusion.

    Much more importantly, I think we are coming from similar premises. And I admit, I find Nick's argument clever enough that I've been drawn into its logic, even tho it's situated in a New Keynesian framework that I would normally object to. That said --

    The destruction of paper wealth induces those who suffered losses to redouble their efforts to recoup their losses--resulting in a desire to net save.

    I don't think this is so obvious. Destruction of paper wealth has at least three, potentially contradictory, effects. First, if people are targeting some ratio of net worth to income, then as you say, it will induce them to save more. But then to the extent it changes their expectations of the return on saving, i.e. if the destruction of wealth is expected to possibly recur, then the tradeoff between saving and current consumption will look less favorable, and the effect will be just the opposite -- lower saving. Finally, if the destruction is (perceived to be) specific to a particular class of assets, then apart from any effect on the aggregate volume of saving, it will result in a shift in portfolio composition, away from the default-prone one. It's this last effect that people are pointing to as potentially expansionary.

    He assumes that an increase in future inflation expectations is the same as a reduction in current nominal wealth. Needless to say, this would not hold in a world with debt denominated in nominal terms.

    It's a reduction in wealth for net-creditor units, and an increase in wealth for net-debtor units. Which, given reasonable assumptions about expenditure functions, should be expansionary.

    Again, I feel like we probably agree on most of the big questions, so I 'm trying to figure out on what exactly we disagree. Let me ask you: Do you think that, in present conditions, an increase in inflation (or expected inflation) would increase real output in the US?

    ReplyDelete
  11. JW,

    First up, I think we have lot in common and I am a fan of your way of thinking. Otherwise I wouldn't be spending my time on your site.

    Second, Nick is pointing to the contradiction in the New keynesian way of thinking. In that i am with him. On other things, not so much.

    On the destruction of paper wealth issue, the effects are different over different horizons. In the short to medium run, people will try to increase their saving and they have still not given up on their past (now inrealizable) targets. In the long-run, they will adjust to the new reality. So, in a sense, going cold turkey might be very painful but would get you out quicker as long as the basic institutions of credit are not completely broken (a la Great Depression). On the other hand, extending the paper wealth adjustment over long periods can create the Japan type of morass.

    If the destruction of wealth is expected to recur, then people will try to find assets which are immune to that (or are perceived to be immune to that). if they cannot, then that economy will become depressed a la Zimbabwe. Recurring destruction of paper wealth (not real wealth) cannot fuel stable economic activity. They will NOT consume more (except perhaps buy real assets, which have an investment value).

    This is a complex topic and needs more elaboration and thinking and your questions deserve far more careful analysis on my part than what I have offered so far.

    On the net-creditor versus net debtor units issue, this is another version of the old mainstream argument against the debt-deflation hypothesis of Fisher. In this particular case, if the government defaults on its bonds, bond holders feel poorer. Clearly they are going to spend less. Unless the government is able to compensate by spending more (I don;t see how it can if it is shut out of the bond markets--remember that the default is happening because the government can't directly monetize fiscal deficits), i dont see how debtor units are going to spend more. you need to bring in some expectations about future taxes and the future path of government deficits. I could argue that the default could severely impair the ability of the government to run deficits for a long time and thereby impair automatic stabilizers, increase the expectation of severe recessions and depress current demand.

    Bottom line: I cannot see how injecting more uncertainty by taking an asset hitherto perceived as "money good" and making it risky helps at all.

    Increase in current inflation is a complete different kettle than "expected" inflation. Rise in current inflation means that resources are being more fully utilized on the margin and most importantly wages are rising. That would reduce existing debt service burdens as incomes rise and also increase the capacity to take on more debt. Do you think a bank will look favorably at anyone's loan request if they pointed to expected future inflation as the reason why they expect to service the debt easily even though their current income is stagnant?

    On expected inflation: the only item sensitive to this is durable goods and structures (including capital equipment). If you expect the price of capital goods to rise by a few percentage points more, would that lead to a rush to buy more equipment even if current capacity is far from fully utilized and current asset prices are declining (housing and commercial real estate)?

    Sorry for posting as anon!!+

    ReplyDelete