A monetary policy–based explanation of swap spreads in China

• Published date: 01 November 2023
• Author: Longzhen Fan,Xin Hou,Qian Sun
• Date: 01 November 2023
• DOI: http://doi.org/10.1002/fut.22451

Received: 8 February 2023

|

Accepted: 26 June 2023

DOI: 10.1002/fut.22451

RESEARCH ARTICLE

A monetary policy–based explanation of swap spreads

in China

Longzhen Fan

1

|Xin Hou

2

|Qian Sun

1

1

Department of Finance, School of

Management, Fudan University,

Shanghai, China

2

Fixed Income Department, Guangfa

Securities, Guangzhou, China

Correspondence

Qian Sun, Department of Finance, School

of Management, Fudan University,

Shanghai, China.

Email: sunqian@fudan.edu.cn

Funding information

National Natural Science Foundation of

China (NSFC), Grant/Award Number:

71972044

Abstract

The dynamics of swap spreads in China cannot be explained by commonly

recognized factors documented in the literature. A unique feature of China's

financial system is that commercial banksare not only long‐term loan providers

but also dominant bond investors. Thus, the loan prime rate (LPR) is an

opportunity cost for commercial banks' bond holdings. However, the LPR is

largely determined by the central bank and often deviates significantly from

equilibrium. In contrast, the swap rate is largely determined by the market. Our

illustrative model and empirical evidence show that the LPR and funding

availability affect bond yields more than swap rates, while volatility in the

money market interest rate affects swap rates more than bond yields. Therefore,

the swap spread is largely driven by the monetary policy. We also show that

swap spreads can be a predictor of excess returns on bonds but not swaps.

KEYWORDS

excess return, loan prime rate, money supply, repos, swap spreads

JEL CLASSIFICATION

G12, E44

1|INTRODUCTION

Interest rate swaps are popular instruments for managing interest rate risk. The swap spread is conventionally

measured as the difference between the swap rate (the fixed rate of a swap) and the Treasury bond yield with the same

maturity. Understanding the determinants of swap spreads is useful for investors to more accurately hedge, arbitrage,

and speculate in swap and bond markets. We examine the swap spread in China, which is one of the major fixed‐

income securities markets.

We define the swap spread in China as the swap rate (the fixed leg of a swap) minus the China Development Bank

(CDB) bond yield of the same maturity. We use the CDB bond rather than the Treasury bond (T‐bond) issued by the

Chinese government, namely, the Ministry of Finance (MOF), for two reasons. First, the CDB is a policy bank fully

owned by the Chinese government. CDB bonds are guaranteed by the MOF and can be a good proxy for T‐bonds, but

CDB bonds are much more liquid than T‐bonds in the market. Second, CDB bond yields are taxable as the return from

swap trading, while T‐bond yields are not.

1

J Futures Markets. 2023;43:1645–1667. wileyonlinelibrary.com/journal/fut © 2023 Wiley Periodicals LLC.

|

1645

1

T‐bond yields in the United States are taxable.

Our study is motivated by three observations. First, the Chinese interest rate swap market has been growing very

fast since its debut in June 2006. In 2021, the swap trading amount totaled RMB 19.8 trillion, which is 8.63% of the total

cash bond trading (RMB 229.1 trillion). What are the determinants of swap spread in China? Do swap spreads contain

useful information for predicting future excess returns on bonds and swaps? These questions have yet to be answered.

Second, previous studies focus on LIBOR‐swap spreads in the United States and find that the credit risk of LIBOR,

the convenience yield of T‐bonds (due to their safety and liquidity), and some demand‐specific factors contribute to

swap spreads (Duffie & Singleton, 1997; Feldhütter & Lando, 2008). However, swap spreads are negative most of the

time across all durations from 1 to 5 years (see Figure 1a) in China; therefore, the convenience yield of T‐bonds or CDB

bonds can be ruled out, as it contributes to positive swap spreads but not negative ones. In addition, swaps in China

usually take the 7‐day repo rate as the floating leg. Unlike LIBOR, the repo rate is risk‐free; thus, the credit risk of the

floating rate should not be the major determinant of the swap spread in China. New explanations are necessary.

Third, negative swap spreads have been observed in the United States since 2008. Specifically, the 30‐year swap

spread has been negative since 2008, while the 10‐year swap spread has been negative, on average, since 2015. The

negative swap spread indicates the existence of arbitrage opportunities. If it persists for a long time, there must be some

limitations to arbitrage. Klingler and Sundaresan (2019) find that the excess demand from underfunded pension plans

in the United States explains the negative swap spread of the 30‐year maturity. Boyarchenko et al. (2018) find that the

regulatory change around 2015 makes arbitrage unprofitable, even with a quite large negative swap spread. Jermann

(2020) argues that the high cost of bond holding and the limit to arbitrage result in a negative swap spread. These

explanations are largely US‐specific.

Based on China's unique institutional setup, we propose a monetary policy–based explanation for swap spreads.

The People's Bank of China (PBoC) set the money growth rate as an intermediate policy target until 2017. Although the

money supply has not been explicitly mentioned as a policy target since then, it is still implicit.

2

The repo (money

market) rate is mainly determined by the market, but the loan prime rate (LPR), which is the benchmark lending rate

quoted by major banks, is still not. The LPR closely follows the midterm monetary policy rate set by the PBoC,

specifically, the 1‐year base loan lending rate at an earlier time and the 1‐year medium‐term lending facility (MLF) rate

more recently. Thus, the LPR can be viewed as a quasi‐official rate for the midterm lending rate. In addition, China's

commercial banks are not only the largest loan provider but also the dominant player in the bond market, holding

about 60% of outstanding bonds. As bonds and loans are close substitutes for commercial banks, the LPR should have a

significant impact on bond yields. In contrast, swap rates are mainly determined by the expectation of the future money

market rate (the repo rate). Therefore, the LPR is often out of line with the money market rate. This inconsistency is

likely a cause of the difference between swap and bond rates or swap spreads in China.

However, when the PBoC tightens (loosens) the money supply, the funds available for commercial banks to

invest in bonds are squeezed (increased), causing bond yields to increase (decrease). In contrast, a swap contract is

anear‐zero investment, and the fixed rate of a swap contract mainly depends on the expected values of the floating

rate. Therefore, a shock to funding availability from the money supply change should affect swap rates less than

bond yields.

When the LPR is high, and the money supply is tight, bond yields are likely to be higher than swap rates and lead to

a negative swap spread. However, when the fund supply is abundant, the swap spread can also be positive. As shown in

Figure 1a, the three major periods with positive swap spreads coincide with easy monetary policy regimes: (1)

immediately after the global financial crisis in 2010, (2) around 2016 (following the stock market crash in mid‐2015 and

the economic slowdown), and (3) during the Covid‐19 pandemic in 2020.

We build a simple illustrative model to show that swap spreads in China are affected by the expectation for the

future money market rate, the expected volatility of the future money market rate, the quasi‐official rate in the

credit market (the 1‐year LPR), and the money supply. We derive hypotheses from the model to explain swap

spreads and test them using relevant data from January 2010 to December 2021. The test results support our

hypotheses.

We further infer from the model that swap rates incorporate expectations for future market rates better than bond

yields, because bond yields are more restricted by the LPR and more sensitive to the funding situation. The swap spread

is more likely to reflect the deviation of the bond yield from the average of the current and expected values of the short‐

term rate. Thus, we further hypothesize that the swap spread should contain more relevant information for future

2

See the next section for a discussion of China's monetary policy.

1646

|

FAN ET AL.