Official Demand for U.S. Debt: Implications for U.S. Real Rates

• Published date: 01 March 2020
• Author: IRYNA KAMINSKA,GABRIELE ZINNA
• Date: 01 March 2020
• DOI: http://doi.org/10.1111/jmcb.12660

DOI: 10.1111/jmcb.12660

IRYNA KAMINSKA

GABRIELE ZINNA

Official Demand for U.S. Debt: Implications for

U.S. Real Rates

Weestimate a structural term-structure model of U.S. real rates, where arbi-

trageurs accommodate demand pressures exerted by domestic and foreign

official investors. Official demand affects rates by altering the aggregate

price of duration risk, and thereby bond risk premiums. Although foreign

central banks’ demand contributed to reduce long-term real rates mainly in

the years prior to the global-financial crisis, the Federal Reserve’s demand

lowered rates during the quantitative easing period. Overall, the two-factor

model, augmented to account for changing liquidity conditions, offers a

good representation of real rates during the 2001–16 period; however, we

flag some caveats and possible extensions.

JEL codes: F31, G10

Keywords: term structure of real rates, quantitative easing, global

imbalances, Bayesian econometrics.

SINCE THE EARLY 2000S,PURCHASES of U.S. Treasurybonds by

official investors have reached unprecedented levels, becoming an increasingly im-

portant tool in central banks’ policies. Foreign officials, that is, foreign central banks,

were the first to significantly increase their holdings of U.S. Treasury bonds, as part

of their reserve accumulation policies (ECB 2006). Then, during the recent financial

The authors are indebted to Pok-Sang Lam (the editor), two anonymous referees, Lorenzo Braccini,

Ray Brooks, Ales Bulir, Francesco Columba, Stefania D’Amico, Mike Joyce, Andrew Meldrum, Monika

Piazzesi, Ricardo Reis, Martin Schneider, Dimitri Vayanos,Jason Weiss, and Bank of England and IMF

seminars’ participants, and XIII Annual Conference on Financial Stability and Banking (Banco do Brasil)

participants for valuable comments. Weare particularly grateful to Stefania D’Amico and Carolin Pflueger

for providing some of the liquidity data, and to David Andolfatto and Andrew Spewak for sharing their

Treasury holding estimates. Ross Irons provided excellent editorial assistance. This research was started

when Iryna Kaminska was working at the International Monetary Fund (IMF), and Gabriele Zinna was

working at the Bank of England and visiting scholar at the IMF. The usual disclaimer applies, so that the

views expressed in the study are those of the author(s) and do not necessary represent those of the Bank

of England, Bank of Italy, or IMF.

IRYNA KAMINSKA is at Bank of England (E-mail: iryna.kaminska@bankofengland.co.uk). GABRIELE

ZINNA is at Bank of Italy (E-mail: gabriele.zinna@bancaditalia.it).

Received August 1, 2014; and accepted in revised form February 28, 2019.

Journal of Money, Credit and Banking, Vol.52, Nos. 2–3 (March–April 2020)

C

2019 Bank of England. Journal of Money, Credit and Banking C

2019 The Ohio State

University

324 :MONEY,CREDIT AND BANKING

crisis, as policy interest rates approached the zero lower bound (ZLB), the Federal

Reserve (Fed) launched the policy of quantitative easing (QE), and as a result the Fed

also became an important active investorin the Treasury market. The objective of this

unconventional monetary policy (UMP) was to stimulate the economy by reducing

long-term interest rates through a series of asset purchase programs. Specifically, the

Fed’s large-scale asset purchases (LSAPs) mainly focused on longer term securities,

including government bonds and mortgage-backed securities (MBSs).

Two features in particular distinguish foreign and domestic policymakers from a

typical U.S. Treasury bond investor. First, official investorsstand out for their massive

holdings of Treasury securities. Second, their demand displays relatively low price

elasticity, in that, it is only slightly sensitive to risk-return considerations (see, e.g.,

Krishnamurthy and Vissing-Jorgensen 2012). Of further interest is the rather high

duration of the Treasury securities held by official investors. These facts, taken

together, lie behind the emerging consensus among market participants, academics,

and policymakers that such inelastic, large-scale official purchases of U.S. Treasury

bonds have qualitatively contributed to lowering U.S.l ong-term interest rates.

A natural question though is how big is the impact of officialdemand on yields? We

answer this question by estimating a structural arbitrage-free model on U.S. real rates.

Specifically, we organize the analysis around the “preferred-habitat” model proposed

by Vayanos and Vila (2009), VV hereafter, in which equilibrium interest rates are

determined by the interaction of two different types of investors: those who trade

bonds at different maturities for return considerations (arbitrageurs) and those who

buy longer term bonds mainly for reasons other than returns (like official investors).

We estimate this structural model on the term structure of monthly real rates,

derived from U.S. Treasury Inflation-Protected Securities (TIPS), over the 2001–16

period. The term structure is determined by two factors: the short-term real interest

rate and excess supply. We specify the excess-supply factor as a function of a number

of observable variables, that is, the Fed and foreign official holdings of Treasuries

(demand) and the amount of Treasury securities outstanding (supply), each of which

is expressed relative to the amount of Treasury securities held by investors with

a price-elastic demand for bonds (arbitrageurs). We also account for the changing

liquidity conditions in the TIPS market; we do this by allowing for an additional

factor, which, although it does not enter directly the bond pricing in the VV model,

can still affect the dynamics of real rates. In this way, model tractability is preserved.

We find that the model fits the term structure of U.S. real rates well, across

maturities and over time. Demand pressures have a dominant impact on longer term

rates. In contrast, changes in the short-rate factor mainly affect shorter term real rates,

producing almost no impact on very long-term rates. Wealso find that modeling TIPS

liquidity is particularly important to achieve a good model performance during the

crisis period. This is the time when TIPS liquidity deteriorated as a result of the

short-lived but intense tensions materializing in the repo market (Campbell, Shiller,

and Viceira 2009). It is also apparent that the inclusion of the liquidity factor helps

estimate the short-rate factor more precisely. Indeed, similar to the short-rate factor,

the liquidity component displays a downward-sloping term structure, and hence is

IRYNAKAMINSKA AND GABRIELE ZINNA :325

important for obtaining an accurate decomposition of the real rates into the expected

and term-premium components.

Turning to the analysis of the excess-supply factor extracted from the VV model,

we find that our main hypothesis is confirmed, that is, the excess-supply factor

moves coherently with measures of demand and supply. Specifically, it increases

with supply, whereas it decreases with official demand. In absolute terms, the excess-

supply factor loads more on the Fed’s holdings than on foreign official holdings and

supply, respectively. According to the VV model, this evidence would be coherent

with a higher duration of the Fed’s holdings, and/or with a lower price sensitivity of

Fed’s demand. However, to examine the economic relevance of these results, we turn

to the price impact analysis.

Although the price impact analysis rests on some model assumptions, it helps shed

light on whether the estimated parameter magnitudes are reasonable in an economic

sense. We find that the recent purchases of U.S. government debt securities by the

Fed and foreign officials have indeed affected the level and dynamics of U.S. real

rates. By 2016, foreign purchases of U.S. Treasuries exerted a cumulative negative

impact of around 90 basis points on long-term U.S. real rates, which drops to less

than 50 basis points at shorter maturities. Furthermore, most of their price impact

materializes in the period prior to the crisis, when supply is increasing, but at a much

lower pace than foreign official demand. Thus, the net price impact of foreign official

demand is also sizable.

As to the Fed price impact during the QE programs, our analysis shows that Fed

purchases also contributed to significantly lower real rates. With a focus on LSAP

programs, we find that the price impact of the second Large-Scale Asset Purchase

Program (LSAP2) is larger than that of the other two LSAP programs; this finding

also holds when supply is accounted for, that is, when considering the Fed price

impact net of supply. Moreover, the analysis suggests that the Fed proved effective in

lowering real rates also during the Maturity Extension Program (MEP), although by

a lower extent relative to LSAP2. However, at the time of the MEP, supply pressures

also temporarily reverted, contributing to the reduction in real rates. Overall, the net

price impact of official purchases varies with the supply of bonds, but also with the

risk-bearing capacity of the arbitrageurs, consistently with the VV model.

Our study relates to a number of earlier studies on reserve accumulation by foreign

central banks and its impact on U.S. interest rates (e.g., Warnock and Warnock

2009; Krishnamurthy and Vissing-Jorgensen 2012; Sierra 2014), and to those studies

trying to quantify the impact of Fed asset purchases on U.S. interest rates (e.g.,

Gagnon et al. 2011; Krishnamurthy and Vissing-Jorgensen 2011; Swanson 2011;

D’Amico et al. 2012; Hamilton and Wu 2012; Meaning and Zhu 2012; D’Amico

and King 2013). Although these studies agree on the qualitative (i.e., downward)

effect of official intervention on U.S. Treasury yields, the evidenceis not unanimous

in quantitative terms, as the estimates can vary significantly. In this paper, we bring

together these otherwise separate strands of the literature, by jointly analyzing the

impact of foreign and domestic official demand pressures on the U.S. Treasury bond

markets.