Sunday, November 8, 2015

The 13 Trillion Dollar Question

On Tuesday Nov 10 there will be a conference in Chicago on "The $13 Trillion Question: Managing the U.S. Government’s Debt" hosted by the Initiative on Global Markets at Chicago Booth, and the Hutchins Center on Fiscal and Monetary Policy at Brookings. (The Brookings announcement here.)

Robin Greenwood will present "The Optimal Maturity of Government Debt and Debt Management Conflicts between the U.S. Treasury and the Federal Reserve" arguing that the Fed and Treasury are working to cross-purposes -- the Fed buys what the Treasury sells -- and that the government  should go after low rates on long term bonds rather than the budget insurance of issuing long term bonds.

(The government faces the same decision a homeowner does: borrow at near-zero floating rates,  but maybe rates shoot up and so do your payments, or borrow long at 2% rates, and pay more if rates don't go up. Robin and Larry favor the former. I'm more risk averse. Maybe living in California has sensitized me  that just because you haven't seen an earthquake recently doesn't mean you shouldn't buy earthquake insurance. But it's a good argument to have qualitatively -- what's the risk, and what's the reward.)

I will present "A new structure for Federal Debt," arguing for an overhaul of which instruments the Treasury issues, to make them more useful for financial markets and financial stability as well as for government borrowing and risk management. (Earlier blog post about this paper here.)

There will be extensive discussion and broader issues, and (the big draw) a panel of Seth  Carpenter, Charles Evans, and Sara Sprung, moderated by David Wessel.

The conference is by invitation, but you can still sign up here until they run out of room, or email Jennifer (dot) Williams at chicagobooth (dot) edu. It will also be viewable by live webcast, link here, starting 1:30 central.

Update: Video of the event here.


Session I - The Optimal Maturity of Government Debt and Debt Management Conflicts between the U.S. Treasury and the Federal Reserve


Robin Greenwood, George Gund Professor of Finance and Banking, Harvard Business School
Samuel G. Hanson, Assistant Professor of Business Administration, Harvard Business School


Guido Lorenzoni, Breen Family Professor, Northwestern University


Austan Goolsbee, Robert P. Gwinn Professor of Economics, University of Chicago Booth School of Business

Session II - A New Structure for U.S. Federal Debt


John H. Cochrane, Senior Fellow, Hoover Institution and Distinguished Senior Fellow, University of Chicago Booth School of Business


James J. McAndrews, Executive Vice President, Federal Reserve Bank of New York


Anil K Kashyap, Edward Eagle Brown Professor of Economics and Finance, University of Chicago Booth School of Business

Session III - Panel Discussion


Seth B. Carpenter, Assistant Secretary for Financial Markets, Department of the Treasury
Charles Evans, President and Chief Executive Officer, Federal Reserve Bank of Chicago
Sara Sprung, Managing Director, Neuberger Berman


David Wessel, Director, The Hutchins Center on Fiscal and Monetary Policy, Brookings Institution


  1. From your article:

    Goals -
    1. The first, traditional, goal of debt management is to fund deficits at lowest long-run cost to the taxpayer.

    No, that is not the goal of debt management. A government does not sell debt to fund deficits (it can literally print money to cover budget shortfalls). A government sells debt to commit itself to a level of spending into the future.
    2. A second goal is to provide liquid and otherwise useful securities that the market desires, securities that enhance financial and macroeconomic stability, and securities that the Government has a natural advantage in producing.

    That macroeconomic stability that you reference comes from the insurance aspect that government debt provides. Government debt that pays no rate of return offers nothing in terms of financial or macroeconomic stability. This comes into direct conflict with objective #1 - the lowest long run cost to the taxpayer is zero.

    3. A third goal is to manage the risks of interest rate increases and other adverse events to the U. S. budget and to the economy.

    If the U. S. government is concerned about interest rate increases adversely affecting the U. S. budget, then the U. S. government should get out the borrowing business altogether - that means printing money, running balanced budgets / surpluses, or selling equity instead of debt.

    4. Macro-economic stabilization is a new fourth goal.

    United States Macro-economic stabilization is the fourth new goal. Neither the Federal Reserve or the U. S. Treasury department are under direction to provide for global macroeconomic stabilization.

    Here are your recommended securities:


    1. Fixed-value, floating-rate debt
    This debt has a fixed value of $1.00, and pays a floating overnight interest rate.

    2. Nominal perpetuities; fixed-coupon debt
    This debt pays a coupon of $1.00 per bond, forever.

    3. Indexed perpetuities
    This debt pays a coupon of $1 times the current consumer price index (CPI).

    4. Tax-free debt
    Treasury debt should be free of all income, estate, capital gains, and other taxes.

    5. Variable-coupon debt
    Long-term debt should allow the government to temporarily lower coupons without triggering legal default.

    I fail to see how any of your suggested securities satisfy any of your suggested goals.

  2. I will not be able to attend the conference but my question to Seth Carpenter and Charles Evans would be this:

    To preserve the independence of the central bank and to maintain the separation of powers between the legislative and executive branches of government, should the U. S. Treasury sell equity like (risk bearing) securities that DO NOT commit the Federal Government to a level of expenditures in the form of interest or other payments?

  3. “The government faces the same decision a homeowner does..”. Sounds like someone can’t distinguish macro from micro.

    The reality is that the state faces no problem whatever from rising interest rates. If rates quadruple, all the state has to do is tell those who want their debt rolled over with a view to getting four times as much interest to shove off. Get lost. Scram.

    I.e. the state just prints dollars, and pays off creditors as their Treasuries mature. Of course that will have a SLIGHT stimulatory or inflationary effect. But that’s easily solved by increased taxes or reduced public spending. I said “slight” because the evidence from QE seems to be that printing money and buying government debt does not have a HUGE effect.

  4. Coincidentally, $13trillion is the amount the Fed is supposed to have loaned to banks during the crisis. What I want to know in relation to that second $13trillion question is what was the average rate of interest the Fed charged for that largesse?

    I very much doubt it was anything near what Walter Bagehot would have regarded as a “penalty rate”. Far as I can see it was about zero. But information on this is hard to come buy, and you can see why: the revolving door brigade and the socialism for the rich brigade don’t want anyone to know how large and comfortable their featherbed is.

  5. I read through that, and I didn't see anything about Biblical tithing. How do you expect to be taken seriously amongst today's leading conservative thinkers if you don't try to justify any of this based on the divinely inspired thoughts of late Iron Age desert dwelling goat herders???

  6. There seems to be a conservative or right wing meme evolving that the federal government should run deficits, in order to print up debt that the private sector desires.

    Just in time for the next GOP president: "Deficits do not matter."

    I also expect to see some new ideas to emerge that justify an easier monetary policy, if we do get a new GOP president.

    Shoot for zero percent interest rates to fight inflation!


  7. "The first, traditional, goal of debt management is to fund deficits at lowest long-run cost to the taxpayer."

    Unless we ran persistent budget surpluses (unlikely) higher interest on the debt will just increase the debt and cost the taxpayer nothing. That is, the debt can be perpetually rolled over or, given the legal authority to do so, just be eliminated in exchange for bank deposits (similar to QE - this is a neutral "duration but not asset altering" transaction as Dr. Cochrane has explained.) No taxpayers (or grandchildren) involved!
    So I submit that the taxpayer need not worry about the "cost" of higher interest rates on Treasury securities. Indeed, as Ralph pointed out, it is the private sector which is the beneficiary of this higher interest income.

    1. Charles,

      "That is, the debt can be perpetually rolled over..."

      I sincerely doubt that the U. S. government or any government could sell bonds if it's annual interest expense exceeded it's annual tax revenue. That seems to be your implication, that bondholders don't care where the money comes from that is used to repay them. Ponzi finance schemes tend to go up in flames when they are revealed.

      "or given the legal authority to do so, just be eliminated in exchange for bank deposits"

      Except the central bank does not have that legal authority.

      "So I submit that the taxpayer need not worry about the cost of higher interest rates on Treasury securities."

      I agree with the premise, but disagree with the rationale. The taxpayer and the bondholder can be the same person. And so (IMHO) the existence of government debt is about incentives.

      Should a government sell bonds - irrespective of interest rates, term structure, deficits / surpluses, etc?

      John Cochrane seems to think it should, I disagree.

  8. Typo:
    Page 20 of your essay. 0.95 x 0.95 = 0.9025, not 0.96

  9. Suppose the government finances the bulk of its debts on the long end of the curve, and compare it to the alternative of financing short. What are the risks / costs.

    It pays a higher rate on average than it does if it finances the bulk of its debt on the short end.

    And, if it finances short? It runs the risk that rates move higher.

    But, if rates move higher, what happens to the fiscal situation of the government.
    Rates move higher because the real rate is moving higher, which is historically a symptom of stronger than expected GDP growth. In which case government revenues are likely rising faster than expenses.

    Or, rates move higher because inflation moves higher. In which case revenues and expenses are both rising.

    On balance, I would say we are better-off finance more debt off the short end, but not all debt off the short end.

    1. Douglas,

      When doing an analysis of government financing, it is important to remember that a government can sell zero coupon securities. When you or I borrow money, we are required to being paying interest and principle back in a relatively short amount of time. A government can sell bonds where the interest and principle may not be paid back for a significant period of time (30 to 50 years depending on the government).

      And so, are we "better off" if the government sells 1 year zero coupon bonds at a 1% interest rate or 1 billion year zero coupon bonds at a 10% interest rate?

      "On balance, I would say we are better-off finance more debt off the short end, but not all debt off the short end."

      From an incentives perspective, I believe that the U. S. economy would be better served if the U. S. Treasury sold equity claims on future tax revenue in lieu of debt.

    2. It doesn't matter if the debt is coupon or zero coupon. Any coupon bearing debt can be stripped into zero-coupon cash flows. Simpler to just think about it all as future cash flow.

    3. Douglas,

      "It doesn't matter if the debt is coupon or zero coupon. Any coupon bearing debt can be stripped into zero-coupon cash flows. Simpler to just think about it all as future cash flow."

      From the tax payer's standpoint it matters a heck of a lot. With zero coupon debt, the taxpayer does not make payments on the debt until it reaches maturity.

      If the U. S. government sold billion year bonds, then for the next billion years no principle or interest payments would be made by the taxpayer. Or if a billion years isn't long enough for you, how about a trillion years?

    4. Douglas,

      You said: "On balance, I would say we are better-off financing more debt off the short end, but not all debt off the short end."

      Who is the "we" that your are talking about and how are they better off?

      If the "we" is taxpayers that are alive today, then "we" are better off passing the bill onto to future generations. That means zero coupon debt that comes due a long, long time from now.

      If the "we" is bondholders that are alive then "we" are better off with coupon debt that pays as large a portion of available tax revenue as possible (100%).

      In any insurance program (that's what government debt really is), there is an effort to match the duration of assets and liabilities. A government's assets are it's people that pay taxes. A government's liabilities include the bonds that it makes payments on from those taxes.

      And so, it would make sense for a government to match the duration of it's debt to the average remaining time left that it's population will be paying taxes. A young population obviously can make debt payments for a longer time than an older population.

      And so you might say that "we" are better off if our government pursues a maturity structure for it's debt that matches it's future revenues holding tax policy constant.


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.