This chapter examines the potential effect of prospective increases in U.S. federal government debt on long-term bond yields. We present estimates of medium-term demand for U.S. Treasury debt and examine the portfolio adjustments that would be implied by high debt supply. We also investigate the implications of high public deficits for U.S. saving and investment flows using an econometric model. Based on standard empirical estimates of the impact of debt on interest rates, our results suggest that the increase in debt alone could add 50 to 150 basis points to the longer-term borrowing costs of the U.S. federal government.

Abstract

This chapter examines the potential effect of prospective increases in U.S. federal government debt on long-term bond yields. We present estimates of medium-term demand for U.S. Treasury debt and examine the portfolio adjustments that would be implied by high debt supply. We also investigate the implications of high public deficits for U.S. saving and investment flows using an econometric model. Based on standard empirical estimates of the impact of debt on interest rates, our results suggest that the increase in debt alone could add 50 to 150 basis points to the longer-term borrowing costs of the U.S. federal government.

This chapter examines the potential effect of prospective increases in U.S. federal government debt on long-term bond yields. We present estimates of medium-term demand for U.S. Treasury debt and examine the portfolio adjustments that would be implied by high debt supply. We also investigate the implications of high public deficits for U.S. saving and investment flows using an econometric model. Based on standard empirical estimates of the impact of debt on interest rates, our results suggest that the increase in debt alone could add 50 to 150 basis points to the longer-term borrowing costs of the U.S. federal government.

A. Introduction

1. Who will finance the U.S. federal deficit in the medium term? The publicly-held debt of the federal government is expected to increase to 64 percent of GDP in 2010 from 36 percent in 2007. The federal budget deficit is expected to continue rising in the medium term on current policy proposals, which would bring the ratio of federal debt to GDP to levels not seen since the 1950s—on staff’s economic projections, to about 80 percent of GDP by 2015. Although the demand for debt has been brisk recently, the salient question is who will finance the future debt and at what cost once safe haven considerations subside.

B. Post-Crisis Financing Patterns

2. The post-crisis period has seen an increase in domestic holdings of Treasury debt. While domestic residents have historically accounted for the dominant share of holdings, the decade leading up to the crisis saw an increasing share of foreign purchases, in particular from emerging markets. In a reversal of this trend, about half of the net debt issuance in 2008–09 was purchased by U.S. residents, in particular the financial system in 2008, and households, nonprofits, and hedge funds in 2009.

Net Purchases of U.S. Federal Publicly Held Debt in 2008–09

Citation: IMF Staff Country Reports 2010, 248; 10.5089/9781455206759.002.A004

Sources: U.S. Department of the Treasury, Treasury International Capital System; Board of Governors of the Federal Reserve System, Flow of Funds Accounts; and Fund staff estimates.1/ Includes hedge funds and nonprofits.2/ Banks, Mutual Funds, Pension Funds, Insurers.3/ Barbados, the Bahamas, Bermuda, Cayman Islands, Netherlands Antilles, Panama, Hong Kong, Ireland, Luxembourg, Switzerland, and United Kingdom.

3. Going forward, foreign demand is unlikely to keep up with the supply of debt, implying that domestic residents will need to absorb an increasing share of debt issuance. Foreign purchases of U.S. Treasury securities would likely be determined largely by the prospective purchasers’ external surpluses, in particular official reserve accumulation, and motives to stabilize currencies against the U.S. dollar versus other currencies. Given limits on sustained expansions of foreign demand, domestic holdings of Treasury securities would have to increase. Among domestic holders, households and banks currently have the largest amounts of total gross financial assets and therefore the largest capacity to purchase federal debt securities, followed by pension funds and insurance companies. Households could potentially absorb further purchases given the likely increase in their saving rate, although a further portfolio allocation by the U.S. financial sector toward Treasury debt is uncertain against the ongoing deleveraging of banks and improving prospects for returns on private assets.

China: Current Account, International Reserves, and U.S. Treasury Holdings

Citation: IMF Staff Country Reports 2010, 248; 10.5089/9781455206759.002.A004

Sources: U.S. Department of the Treasury, Treasury International Capital System; Haver Analytics; and IMF, World Economic Outlook.

C. Baseline Projections of Demand for Treasury Debt

4. The future supply of debt is likely to exceed overall ex-ante demand by a significant margin. We project hypothetical demand paths on the basis of projections of domestic and foreign financial assets and other macroeconomic quanta (using World Economic Outlook projections of GDP, official reserves, external current and capital accounts) and assuming a return to pre-crisis asset allocation patterns (Table 1). In particular, we assume the following baseline demand paths for foreign residents during 2010–15:

  • China continues to purchase Treasuries at a rate of 36 percent of its official reserve accumulation (close to the average ratio of purchases to reserve accumulation in 2000–07),

  • Oil and other commodity exporters bring their purchases (as shares of current account surpluses or foreign asset purchases) to their 2000–07 averages by 2015,

  • Other countries’ holdings return to 2000–07 averages as shares of GDP.

Table 1.

Projections of Baseline Demand for U.S. Treasury Securities and the Impact of Excess Supply on Long-term Bond Yields

(billions of U.S. dollars unless otherwise indicated)

article image
Sources: U.S. Treasury Department, Treasury International Capital System; The Board of Governors of the Federal Reserve System, Flow of Funds Accounts ; and Fund staff estimates.Notes: GDP, current account balances, reserve accumulation, and capital outflows projections are from the World Economic Outlook. A one percentage point increase in the debt-to-GDP ratio is assumed to increase ten year Treasury bond yields by 2-5 basis points, in line with estimates by Laubach (2008) and Engen and Hubbard (2005).

Argentina, Australia, Brazil, Chile, Colombia, New Zealand, Peru, South Africa, Uruguay.

The Czech Republic, Egypt, Hungary, India, Indonesia, Korea, Malaysia, Mexico, Morocco, Pakistan, Philippines, Poland, Romania, Singapore, Taiwan, Thailand, Turkey, Ukraine.

Barbados, The Bahamas, Bermuda, Cayman Islands, Netherlands Antilles, Panama, Hong Kong, Ireland, Luxembourg, Switzerland, and the United Kingdom.

Canada, Kazakhstan, Norway, Algeria, Iran, Iraq, Kuwait, Libya, Nigeria, Qatar, Saudi Arabia, United Arab Emirates, Venezuela, Russia, and Trinidad and Tobago.

Household financial assets are assumed to increase to 430 percent of personal disposable income by 2015 (based on the projections in the selected issues paper on household savings).

The growth rate of banking sector assets is in line with the baseline scenario in the 2010 U.S. FSAP.

The financial assets of the insurance sector are assumed to grow at the same rate as nominal GDP.

The financial assets of pension and money market funds are assumed grow at the same rate as household financial assets.

Based on an estimated relationship between mortgage issuances and house prices and residential fixed investment.

Under these conservative assumptions, foreign holdings would increase by three percent of GDP over the period 2010–15, while the gap between overall projected debt issuance and foreign holdings would increase by about 23 percent of GDP by 2015. At the same time, exante demand from domestic residents could decline by about 6 percent of GDP under the following assumptions:

  • The Federal Reserve and state and local governments do not increase their nominal holdings (given the plan to gradually reduce the size of the Fed’s balance sheet and ongoing retrenchment by state and local governments),1

  • Pension funds, insurers, and banks reduce their holdings to 2000–07 average shares of financial assets on the normalization of business conditions and risk appetite.

  • Households keep their holdings at their current share of financial assets, which is somewhat higher than 2000–07 averages, but possibly sustainable on increased risk aversion after the wealth losses during the crisis. Household financial assets would rise in line with the staff’s projection of the personal saving rate.

Putting these demand projections together would imply 29 percent of GDP in “excess supply” of Treasury securities.

5. Higher real interest rates would likely be needed to encourage purchases of the future debt issuance in excess of the hypothetical path of demand. Assuming a yield impact of 2–5 basis points for each additional percentage point of GDP in excess debt supply would suggest an increase in yields in the range of 60–150 basis points in the medium term.2 This effect would come on top of the increases due to rising short term interest rates (on the gradual exit of monetary policy from very low policy rates), and the normalization of the term premium excluding the debt effect. The staff’s baseline medium-term projection for the 10-year Treasury bond yield is about 6½ percent. An upside risk to this estimate could stem from a projected increase in closely-substitutable GSE-backed debt of roughly US$3.5 trillion between 2009 and 2015 on the basis of staff’s residential fixed investment forecast.

D. A Model of Saving-Investment Flows

6. To supplement the above analysis of prospective demand based on portfolio choices of the key buyers of Treasury debt, staff has also estimated an econometric model of saving and investment flows using the national accounts data.3 This approach helps weigh the supply of savings by households and corporations against the demand for loanable funds by private and public sectors. The model can be used to test the frequently-discussed hypothesis that sluggish recovery in private demand will ease the upward pressures on long-term government bond yields. The econometric results suggest that while this hypothesis could hold in the near term, private investment will gradually pick up, diminishing the private sector surplus available for financing of the budget deficits. The dynamics of public deficits will then become critical for the path of interest rates and overall saving-investment balances:4

Near-term prospects

7. The general government net borrowing is projected by staff to drop from almost 11 percent of GDP in 2010 to 5½ percent in 2012 on expiring fiscal stimulus, diminishing slack in the economy, and initial steps toward fiscal consolidation planned by the authorities. Under the staff’s baseline assumption of the 10-year T-bond yield at 3½ percent in 2010 and 4¾ percent in 2011, the overall domestic saving-investment balance would have a tendency to remain broadly stable around current levels (-3½ percent of GDP), since strengthening private investment would be funded by reduced public dissaving and high household saving as consumers continue to rebuild their balance sheets.

Medium-term dynamics

8. In the absence of deeper fiscal consolidation in the medium term, however, the improvement in public saving will not last given pressures from population aging, health care spending, and higher debt. On current policies, the general government deficit is projected by staff to increase to 6 percent of GDP in 2015 and continue rising thereafter. Meanwhile, investment rates will return to historical norms. The econometric model suggests that interest rates materially lower than the staff’s medium-term baseline of around 6½ percent would imply a gradual deterioration of the saving-investment gap toward the levels seen during the bubble years, deepening the current account deficit and raising demand for external funding. However, such a scenario with lower interest rates could only materialize if availability of external saving was ample and investor appetite for government bonds relative to higher-yielding instruments such as equities remained strong—unlikely when the recovery is fully underway and the global economy is on a trajectory toward a more balanced growth.

E. Conclusions

9. In sum, analysis both on the basis of investor portfolios and saving-investment balances suggest that, on current policies, the sizeable projected increases in U.S. public debt will likely put upward pressure on government borrowing costs in the medium term. That said, the near-term developments are highly uncertain as U.S. Treasury debt continues to enjoy a safe haven status. Early agreement on longer-term fiscal consolidation plans—including through the dedicated Fiscal Commission—would help to solidify this position and avoid an unnecessary increase in long-term interest rates.

  • Engen, Eric, and R. Glenn Hubbard, 2004, “Federal Government Debt and Interest Rates,” in NBER Macroeconomics Annual 2004, ed. by M. Gertler and K. Rogoff, (Cambridge, Massachusetts: MIT Press).

    • Search Google Scholar
    • Export Citation
  • Laubach, Thomas, 2009, “New Evidence on the Interest Rate Effects of Budget Deficits and Debt,” Journal of the European Economic Association, Vol. 7, No. 4, pp. 858885.

    • Search Google Scholar
    • Export Citation
1

Prepared by Oya Celasun and Martin Sommer.

1

The Fed’s holdings exceed seven percent of 2015 GDP—the historic average size of its balance sheet, implying that it could have an unchanged nominal portfolio of Treasury debt while shrinking its assets.

2

The empirical literature finds that a one percentage point in GDP increase in the public debt to GDP ratio increases long term bond yields by 2–5 basis points (see Laubach (2009) and Engen and Hubbard (2005). In this exercise, we apply this elasticity to a projected measure of excess supply of debt to take into account the sources of demand that would potentially offset the yield impact of increasing supply.

3

The analytical framework will be presented in a forthcoming Working Paper.

4

The sum of private and public saving minus private and public investment.

United States: Selected Issues Paper
Author: International Monetary Fund