McGill

UCLA

Indiana University

THE TERM STRUCTURE OF BOND MARKET LIQUIDITY

Ruslan Goyenko, McGill University

Avanidhar Subrahmanyam, UCLA

Andrey Ukhov, Indiana University

FDIC, Washington, DC 2008

2

McGill

UCLA

Indiana UniversityAggregate Liquidity

In the stock marketSome aggregate measure

Bonds have distinctions:MaturityOn-the-run and Off-the-run status

What is “aggregate liquidity” for bonds?

3

McGill

UCLA

Indiana UniversityON-the-Run vs OFF-the-Run

Treasury market illiquidity literature focus: on-the-run ( Fleming and Remolona 1997 EPR, 1999 JF, Balduzzi, Elton,

and Green 2001 JFQA , Green 2004 JF, Brandt and Kavajecz 2004JF, Chordia, Sarkar and Subrahmanyam 2005 RFS )

Treasury Market Illiquidity Premium – Amihud and Mendelson (1991 JF) – OFF-the-Run evidence Differences between yields and spreads of T-bills and notes with

less than 6 months to maturity – OFF-the-Run securities

Illiquidity of OFF-the-Run issues is not studied

4

McGill

UCLA

Indiana University

Short Run vs Long Run

Shiller and Perron (1985), Shiller (1989) increasing the number of observations by sampling more

frequently while leaving the span in years of data unchanged may not increase the power of tests

Previous Literature uses short time span (7 years most): Fleming and Remolona 1999 JF – August 23, 1993 – August 19,

1994 – intraday data Balduzzi, Elton, and Green 2001 JFQA - July 1, 1991 - September

29, 1995 – intraday data Green 2004 JF - July 1, 1991 - September 29, 1995 – intraday

data Chordia, Sarkar and Subrahmanyam 2005 RFS - June 17, 1991-

December 31, 1998 - daily data Our work - November 1967 to December 2005 –

monthly data

5

McGill

UCLA

Indiana UniversityIlliquidity Differences Across Maturities

Notes and Bills have different “quotation, trading and their quotes are transmitted on different systems. Traders usually specialize in one type of these government securities, and there are differences between the two markets” – Amihud and Mendelson 1991

Flights into or out-of the bond market do not target specific maturity ranges. Beber, Brandt and Kavajecz (2006) - investors price the transaction cost component both when they enter and exit the bond market.

Need to understand the structure of bond market liquidity Our work - three illiquidity ranges – short, medium and

long + on-the-run and off-the-run (six economic variables)

6

McGill

UCLA

Indiana UniversityResearch Questions I

Previous research (Brunnermeier and Pedersen, 2006, and Chordia, Roll, and Subrahmanyam, 2005) -macroeconomic variables and price volatility may impact bond market illiquidity by affecting market-making costs. Do such variables differentially impact on- and off-the-run market making costs, and in turn, illiquidity?

Are bond returns forecastable from illiquidity levels, i.e., is

there evidence of illiquidity premia in the bond market?

How are illiquidity shocks transmitted in the bond market? Are they reflected first in the relatively less active off-the-run issues or the more active on-the-run ones?

7

McGill

UCLA

Indiana UniversityResearch Questions II

Cross-Market Effect: If the illiquidity of certain bonds forecasts those of other bonds by reflecting illiquidity shocks first, then it may forecast returns not just in the own-market but in other markets as well.

How does the predictive power of illiquidity for bond returns vary across maturity and off-the-run status?

8

McGill

UCLA

Indiana University

Summary of Results: time series determinants

For both off-the-run and on-the-run – bond spread (long-short) significantly widens during recessions – consistent with flight-to-quality and flight-to-liquidity

For both off-the-run and on-the-run – Granger-causality goes from short-term to long-term (one direction only) off-the-run short-term Granger causes on-the-run short-term (one

direction only)

On-the-run illiquidity is affected by volatility Off-the-run illiquidity is predicted by

inflation (short- and long-term) monetary policy (all maturities) returns and volatility illiquidity of short-term bonds predicts illiquidity of long-term bonds

9

McGill

UCLA

Indiana UniversityResults II: Illiquidity Premium

VAR analysis indicates – on-the-run (short, medium and long) illiquidity has no effect on bond returns

Short-term off-the-run illiquidity affects returns across all maturities

Medium and long-term off-the-run illiquidity is not priced

10

McGill

UCLA

Indiana UniversityData

Treasury market illiquidity – relative bid-ask spread (CRSP daily Treasury Quotes ) from November 1967 to December 2005

On-the-run – just issued, older securities are off-the-run Short-term illiquidity – Tbills with maturity less or equal to 1 year Medium illiquidity – quotes of 2-to-5 year bonds Long-term illiquidity – quotes of 10-year note

Returns: short-term - the return on 3 month T-bill,

medium and long are returns on 5- and 10-year notes (CRSP Treasury monthly file )

11

McGill

UCLA

Indiana UniversityTreasury Illiquidity

Panel A. The whole sample On-the-run Off-the-run Bond-

short Bond-

medium Bond-long

Bond-short

Bond-medium

Bond-long

Average 0.032 0.106 0.111 0.025 0.108 0.156 St. dev 0.026 0.147 0.076 0.023 0.062 0.105 Median 0.019 0.07 0.099 0.012 0.11 0.142 All numbers are multiplied by 100

Panel B. Recessions (NBER) On-the-run Off-the-run Bond-

short Bond-

medium Bond-long

Bond-short

Bond-medium

Bond-long

Average 0.057 0.124 0.147 0.049 0.149 0.234 St. dev 0.03 0.102 0.082 0.029 0.061 0.105 Median 0.066 0.121 0.131 0.054 0.172 0.263

12

McGill

UCLA

Indiana UniversityOff-the-Run Short-Term

0

0.0002

0.0004

0.0006

0.0008

0.001

0.0012

0.00141968

1970

1971

1973

1975

1976

1978

1980

1981

1983

1985

1986

1988

1990

1991

1993

1995

1996

1998

2000

2001

2003

2005

13

McGill

UCLA

Indiana UniversityOff-the-Run Long-Term

0

0.0005

0.001

0.0015

0.002

0.0025

0.003

0.0035

0.004

196

8

197

0

197

1

197

3

197

5

197

6

197

8

198

0

198

1

198

3

198

5

198

6

198

8

199

0

199

1

199

3

199

5

199

6

199

8

200

0

200

1

200

3

200

5

14

McGill

UCLA

Indiana UniversitySpread (long – short)

On-the-run Off-the-run Bond-long – Bond-short Bond-long – Bond-short Whole

sample Recession

No recession

Whole sample

Recession No

recession Diff. 0.08 0.09 0.078 0.131 0.185 0.121 p-value 0.00 0.00 0.00 0.00 0.00 0.00

Flight-to-quality or flight-to-liquidity

15

McGill

UCLA

Indiana UniversityGranger-causality

On-the-run Off-the-run Bond-

short Bond-

medium Bond-long Bond-

short Bond-

medium Bond-long

Bond-short

11.34 (0.001)

5.28 (0.022)

11.97 (0.001)

8.46 (0.004)

Bond-medium

4.19 (0.041)

0.01 (0.912)

0.07 (0.791)

0.00 (0.960)

Bond-long 0.86 (0.355)

0.05 (0.820)

0.53 (0.467)

6.44 (0.011)

Illiquidity shocks are transmitted from the short end to the long end of term structure

16

McGill

UCLA

Indiana UniversityVAR innovations: Off-the-Run

Volat RET1 RET5 RET10 Bond-short

Bond-Medium

Bond-Long

Volat 1.00 RET1 0.06 1.00 RET5 0.02 0.49 1.00 RET10 0.03 0.38 0.91 1.00 Bond-short

0.17 -0.16 -0.18 -0.11 1.00

Bond-medium

0.02 0.01 0.02 0.07 -0.03 1.00

Bond-long

0.08 0.00 -0.14 -0.05 0.27 0.26 1.00

Consistent with Amihud (2002) for the

stock market

Control variables: Inflation

FED, DEF TERM

17

McGill

UCLA

Indiana UniversityOff-the-Run Bond-Short Illiquidity

Impulse response of bond-short toBond-Long Shock

-0.001

-0.0005

0

0.0005

0.001

0.0015

0.002

0.0025

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Medium Shock

-0.002

-0.0015

-0.001

-0.0005

0

0.0005

0.001

0.0015

0.002

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Short Shock

-0.002

0

0.002

0.004

0.006

0.008

0.01

0.012

1 3 5 7 9 11 13 15 17 19 21 23 25

18

McGill

UCLA

Indiana UniversityOff-the-Run Bond-Long Illiquidity

Impulse response of bond-long to

Illiquidity shocks are transmitted from the illiquidity of the short-end to the illiquidity of the long-

end and not vice versa

Bond-Short Shock

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Medium Shock

-0.002

-0.001

0

0.001

0.002

0.003

0.004

0.005

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Long Shock

-0.002

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

1 3 5 7 9 11 13 15 17 19 21 23 25

19

McGill

UCLA

Indiana UniversityIlliquidity and Monetary Policy

Monetary tightening

Bond-Short to FED

-0.001

0

0.001

0.002

0.003

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Medium to FED

-0.002

-0.001

0

0.001

0.002

0.003

0.004

0.005

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Long to FED

-0.001

0

0.001

0.002

0.003

0.004

0.005

0.006

1 3 5 7 9 11 13 15 17 19 21 23 25

Short-term off-the-run has the immediate and persistent

response to FED

20

McGill

UCLA

Indiana UniversitySummary

On-the-Run – less dynamics: positive shock to FED increases short-term illiquidity; shock to volatility increases illiquidity across all maturities active trading in the on-the-run bonds shields market makers from

increases in inventory and order processing costs due to inflation and tighter monetary policy

Off-the-Run – more dynamics: Inflation and FED increase illiquidity across different maturities Volatility increases illiquidity (consistent with inventory risk (Ho

and Stoll (1983) and O’Hara and Oldfield (1986)) Positive shocks to bond returns across different maturities

decrease off-the-run bond illiquidity (consistent with Chordia, Roll, and Subrahmanyam (2001) )

Illiquidity shocks are transmitted from the short-end to the long-end

21

McGill

UCLA

Indiana UniversityPricing Implications

Response of T-bill returns to off-the-run illiquidityBond-Long Shock

-0.0002

-0.00015

-0.0001

-0.00005

0

0.00005

0.0001

0.00015

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Medium Shock

-0.00015

-0.0001-0.00005

00.00005

0.0001

0.000150.0002

0.00025

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Short Shock

-0.0002

-0.0001

0

0.0001

0.0002

1 3 5 7 9 11 13 15 17 19 21 23 25

Only contemporaneous associations for illiquidity

+

FED, Inflation and DEF have positive affect

22

McGill

UCLA

Indiana UniversityPricing Implications

Response of 5-year bond returns to

Bond-Long Shock

-0.001

-0.0005

0

0.0005

0.001

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Medium Shock

-0.002

-0.0015

-0.001

-0.0005

0

0.0005

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Short Shock

-0.0005

0

0.0005

0.001

0.0015

1 3 5 7 9 11 13 15 17 19 21 23 25

Consistent with Amihud

(2002)

As in Fama and French

(1993) TERM and DEF have

an affect

23

McGill

UCLA

Indiana UniversityPricing Implications

Response of 10-year bond returns to

Bond-Long Shock

-0.0015

-0.001

-0.0005

0

0.0005

0.001

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Medium Shock

-0.003

-0.0025

-0.002

-0.0015

-0.001

-0.0005

0

0.0005

1 3 5 7 9 11 13 15 17 19 21 23 25

Bond-Short Shock

-0.0005

0

0.0005

0.001

0.0015

0.002

0.0025

1 3 5 7 9 11 13 15 17 19 21 23 25

Consistent with Amihud

(2002)

As in Fama and French

(1993) TERM and DEF have

an affect

24

McGill

UCLA

Indiana UniversityConclusion

The source of illiquidity premium (Amihud and Mendelson 1991) in the Treasury market is illiquidity of short-term off-the-run issues

Makes sense because: On-the-run illiquidity is not priced and largely driven by volatility For off-the-run illiquidity: illiquidity shocks are transferred from the

illiquidity of the short-end to the illiquidity of the long-end Short-term illiquidity predicts its own illiquidity and illiquidity of

other maturities Short-term illiquidity also predicts illiquidity premium across other

maturities

Dynamics of off-the-run illiquidity is richer: it is driven by inflation, monetary policy, bond returns and volatility (this information is eventually transmitted into the bond prices)