38
Crawford School of Public Policy CAMA Centre for Applied Macroeconomic Analysis A comment on Wu and Xia (2015), and the case for two-factor Shadow Short Rates CAMA Working Paper 48/2015 December 2015 Leo Krippner Reserve Bank of New Zealand and Centre for Applied Macroeconomic Analysis (CAMA), ANU Abstract Shadow Short Rates (SSRs) estimated from shadow/lower-bound term structure models (SLMs) can be useful for monitoring of the stance of unconventional monetary policy and for quantitative analysis, but only if they are relatively robust. I show from several perspectives that SSRs from three-factor SLMs, which includes Wu and Xia (2015) SSRs, are not robust, and how that arises from the inherent flexibility of three-factor SLMs. Such SSRs should therefore be avoided. However, I also show that estimated SSRs from two-factor SLMs are relatively robust. Hence, two-factor SLM SSRs appear to be good candidates for monitoring and quantitative analysis, but ideally with appropriate robustness checks including alternative monetary policy metrics. | THE AUSTRALIAN NATIONAL UNIVERSITY

A comment on Wu and Xia (2015), and the case for two-factor

  • Upload
    lehanh

  • View
    217

  • Download
    4

Embed Size (px)

Citation preview

Page 1: A comment on Wu and Xia (2015), and the case for two-factor

Crawford School of Public Policy

CAMACentre for Applied Macroeconomic Analysis

A comment on Wu and Xia (2015), and the case for two-factor Shadow Short Rates

CAMA Working Paper 48/2015December 2015

Leo KrippnerReserve Bank of New Zealand and Centre for Applied Macroeconomic Analysis (CAMA), ANU

Abstract

Shadow Short Rates (SSRs) estimated from shadow/lower-bound term structure models (SLMs) can be useful for monitoring of the stance of unconventional monetary policy and for quantitative analysis, but only if they are relatively robust. I show from several perspectives that SSRs from three-factor SLMs, which includes Wu and Xia (2015) SSRs, are not robust, and how that arises from the inherent flexibility of three-factor SLMs. Such SSRs should therefore be avoided. However, I also show that estimated SSRs from two-factor SLMs are relatively robust. Hence, two-factor SLM SSRs appear to be good candidates for monitoring and quantitative analysis, but ideally with appropriate robustness checks including alternative monetary policy metrics.

| T H E A U S T R A L I A N N A T I O N A L U N I V E R S I T Y

Page 2: A comment on Wu and Xia (2015), and the case for two-factor

Keywords

Shadow Short Rates, zero lower bound, unconventional monetary policy, term structure models.

JEL Classification

E43, G12, G13

Address for correspondence:

(E) [email protected]

ISSN 2206-0332

The Centre for Applied Macroeconomic Analysis in the Crawford School of Public Policy has been established to build strong links between professional macroeconomists. It provides a forum for quality macroeconomic research and discussion of policy issues between academia, government and the private sector.

The Crawford School of Public Policy is the Australian National University’s public policy school, serving and influencing Australia, Asia and the Pacific through advanced policy research, graduate and executive education, and policy impact.

| T H E A U S T R A L I A N N A T I O N A L U N I V E R S I T Y

Page 3: A comment on Wu and Xia (2015), and the case for two-factor
Page 4: A comment on Wu and Xia (2015), and the case for two-factor
Page 5: A comment on Wu and Xia (2015), and the case for two-factor
Page 6: A comment on Wu and Xia (2015), and the case for two-factor
Page 7: A comment on Wu and Xia (2015), and the case for two-factor
Page 8: A comment on Wu and Xia (2015), and the case for two-factor

2008 2009 2010 2011 2012 2013 2014 2015

Perc

ent

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7Interest rate data

Federal Funds Rate3-month Tbill rateWX 3x1-month f orward rateWX 25 bp LB parameter

Page 9: A comment on Wu and Xia (2015), and the case for two-factor
Page 10: A comment on Wu and Xia (2015), and the case for two-factor
Page 11: A comment on Wu and Xia (2015), and the case for two-factor
Page 12: A comment on Wu and Xia (2015), and the case for two-factor
Page 13: A comment on Wu and Xia (2015), and the case for two-factor
Page 14: A comment on Wu and Xia (2015), and the case for two-factor
Page 15: A comment on Wu and Xia (2015), and the case for two-factor
Page 16: A comment on Wu and Xia (2015), and the case for two-factor

Time to maturity0 2 4 6 8 10

No

unit

0

0.2

0.4

0.6

0.8

1

WX shadow forward rate factor loadings

Factor 1Factor 2Factor 3

Page 17: A comment on Wu and Xia (2015), and the case for two-factor
Page 18: A comment on Wu and Xia (2015), and the case for two-factor
Page 19: A comment on Wu and Xia (2015), and the case for two-factor

0 2 4 6 8 10

% p

oint

s

-2

0

2

4rLB = 25 bps (full)

FR dataLB FRSh. FR

0 0.5 1 1.5 2 2.5-0.2

0

0.2

0.4

0.6

0.8rLB = 25 bps (zoomed)

SSR (-1.42%) rLB

0 2 4 6 8 10

% p

oint

s

-2

0

2

4rLB = 19 bps (full)

FR dataLB FRSh. FR

0 0.5 1 1.5 2 2.5-0.2

0

0.2

0.4

0.6

0.8rLB = 19 bps (zoomed)

SSR (-0.59%) rLB

0 2 4 6 8 10

% p

oint

s

-2

0

2

4rLB = 14 bps (full)

FR dataLB FRSh. FR

0 0.5 1 1.5 2 2.5-0.2

0

0.2

0.4

0.6

0.8rLB = 14 bps (zoomed)

SSR rLB

Time to maturity0 2 4 6 8 10

% p

oint

s

-2

0

2

4rLB = 0 bps (full)

FR dataLB FRSh. FR

Time to maturity0 0.5 1 1.5 2 2.5

-0.2

0

0.2

0.4

0.6

0.8rLB = 0 bps (zoomed)

SSR rLB

Page 20: A comment on Wu and Xia (2015), and the case for two-factor

0 2 4 6 8 10

No

unit

0

0.5

1

rLB = 25 bps

0 2 4 6 8 100

0.5

1

rLB = 19 bps (est.)

Time to maturity0 2 4 6 8 10

No

unit

0

0.5

1

rLB = 14 bps

Time to maturity0 2 4 6 8 10

0

0.5

1

rLB = 0 bps

Page 21: A comment on Wu and Xia (2015), and the case for two-factor
Page 22: A comment on Wu and Xia (2015), and the case for two-factor
Page 23: A comment on Wu and Xia (2015), and the case for two-factor
Page 24: A comment on Wu and Xia (2015), and the case for two-factor
Page 25: A comment on Wu and Xia (2015), and the case for two-factor
Page 26: A comment on Wu and Xia (2015), and the case for two-factor
Page 27: A comment on Wu and Xia (2015), and the case for two-factor

0 2 4 6 8 10

% p

oint

s

-4

-2

0

2

rLB = 25 bps (full)

Y C dataLB YCSh. Y C

0 0.5 1 1.5 2 2.5-0.2

0

0.2

0.4

0.6

0.8rLB = 25 bps (zoomed)

SSR (-5.47%) rLB

0 2 4 6 8 10

% p

oint

s

-4

-2

0

2

rLB = 16 bps (full)

Y C dataLB YCSh. Y C

0 0.5 1 1.5 2 2.5-0.2

0

0.2

0.4

0.6

0.8rLB = 16 bps (zoomed)

SSR (-3.64%) rLB

0 2 4 6 8 10

% p

oint

s

-4

-2

0

2

rLB = 14 bps (full)

Y C dataLB YCSh. Y C

0 0.5 1 1.5 2 2.5-0.2

0

0.2

0.4

0.6

0.8rLB = 14 bps (zoomed)

SSR (-3.38%) rLB

Time to maturity0 2 4 6 8 10

% p

oint

s

-4

-2

0

2

rLB = 0 bps (full)

Y C dataLB YCSh. Y C

Time to maturity0 0.5 1 1.5 2 2.5

-0.2

0

0.2

0.4

0.6

0.8rLB = 0 bps (zoomed)

SSR (-1.92%) rLB

Page 28: A comment on Wu and Xia (2015), and the case for two-factor

0 2 4 6 8 10

No

unit

0

0.5

1

rLB = 25 bps

0 2 4 6 8 100

0.5

1

rLB = 16 bps (est.)

Time to maturity0 2 4 6 8 10

No

unit

0

0.5

1

rLB = 14 bps

Time to maturity0 2 4 6 8 10

0

0.5

1

rLB = 0 bps

Page 29: A comment on Wu and Xia (2015), and the case for two-factor
Page 30: A comment on Wu and Xia (2015), and the case for two-factor
Page 31: A comment on Wu and Xia (2015), and the case for two-factor
Page 32: A comment on Wu and Xia (2015), and the case for two-factor
Page 33: A comment on Wu and Xia (2015), and the case for two-factor
Page 34: A comment on Wu and Xia (2015), and the case for two-factor
Page 35: A comment on Wu and Xia (2015), and the case for two-factor
Page 36: A comment on Wu and Xia (2015), and the case for two-factor

2008 2009 2010 2011 2012 2013 2014 2015

WX

ETZ

(yea

rs)

4

3

2

1

0

25 bps19 bps (est.)14 bps0 bps

2008 2009 2010 2011 2012 2013 2014 2015

K-AN

SM(2

) SSR

(% p

oint

s)

-5

0

SSR (RHS)

QE3taper

Page 37: A comment on Wu and Xia (2015), and the case for two-factor
Page 38: A comment on Wu and Xia (2015), and the case for two-factor