Yoshiyuki Nakazono and Kozo Ueda. December 18, 2010

Size: px

Start display at page:

Download "Yoshiyuki Nakazono and Kozo Ueda. December 18, 2010"

Marcus Walker
5 years ago
Views:

1 .... Policy Commitment and Market Expectations: Survey Based Evidence under Japan s Quantitative Easing Policy Yoshiyuki Nakazono and Kozo Ueda Waseda University and the Bank of Japan December 18, 2010

2 2 / 47 Contents...1 Motivation...2 Model...3 QSS Data...4 Nonparametric Perspective...5 Estimation...6 Concluding Remarks

3 3 / 47 Views expressed in this paper are those of the authors and do not necessarily reflect the official views of the Bank of Japan.

4 4 / 47 Motivation Quantitative easing policy (QEP) by the BOJ March 2001 to March 2006 CPI commitment policy BOJ promised to keep their accommodative policy until the CPI inflation rate became stably zero or higher. Bank of Canada, Riksbank, etc We estimate the effect using survey data.

5 5 / 47 QEP and the CPI Commitment Policy 1 Date Mar Oct Policy Committing that the QEP continues to be in place until the CPI (excluding perishables) inflation registers stably a zero percent or an increase year on year. Enhancing monetary policy transparency. BOJ s commitment is underpinned by the following two conditions. 1. it requires not only that the most recently published CPI inflation should register a zero percent or above, but also that such tendency should be confirmed over a few months. 2. the BOJ needs to be convinced that the prospective CPI inflation will not be expected to register below a zero percent. The above conditions are the necessary condition. There may be cases,however, that the BOJ will judge it appropriate to continue with the QEP even if these two conditions are fulfilled.

6 6 / 47 QEP and the CPI Commitment Policy 2 Date Mar Jul Policy Exit from the QEP by changing the operating target to the uncollateralized overnight call rate. Encouraging the uncollateralized overnight call rate to remain at effectively zero percent. Encouraging the uncollateralized overnight call rate to remain at around 0.25 percent.

7 7 / (%) ZIRP QEP 10-year JGB yield 5-year JGB yield 2-year JGB yield O/N call rate

8 Actual and expected inflation (%) Core CPI (%) CPI 1-year QEP CPI 2-year CPI 10-year Core CPI Core CPI (base year of 2005) III IV I II III IV I II III IV I II III IV I II III IV / 47

9 9 / 47 Past Empirical Studies 1 Supporting evidence of the effect of the CPI commitment policy Survey by Ugai (2006) Baba et al. (2005) and Oda and Ueda (2007) develop a macro-finance model. Two findings Interest rates became lower for the short- to medium-term range by 0.4 to 0.5 % points. A threshold existed at about 1 % of the CPI inflation rate.

10 10 / 47 Past Empirical Studies 2 Difficulty Inflation expectations are unobsevable. The low yield curve may simply be the result of low inflation expectations.

11 11 / 47 Contribution We resolve the difficulty by using a direct measure of inflation expectations. Panel survey, QSS (QUICK Survey System), provided by QUICK corp. Market participants Views on the future interest rates and inflation rates The QSS enables us to directly evaluate the effect of CPI commitment policy on market participants expectations.

12 12 / 47 What We Find, albeit Tentative The CPI commitment policy lowered market participants expectations on interest rates. In the short- to medium-term interest rates About 0.2 percent There existed a threshold inflation rate, yielding a kinked curve between interest rates and inflation rates. About 0 percent Consistent with the BOJ s announcement Weaker effects than previous studies due to controling inflation expectations

13 13 / 47 Model QSS Data Nonparametric Perspective Estimation Concluding Remarks

14 14 / 47 Model without CPI Commitment Simple Taylor rule The latent nominal interest rate i t is given by i t = r + π + φ(π t π ) + ε t. (1) In the presence of the ZLB, the interest rate becomes i t = { 0 if i t 0 it if it > 0. (2) Notes Partial equilibrium (unlike Eggertsson and Woodford (2003)) No inertia in the policy rule (unlike Reifschneider and Williams (2000) and Eggertsson and Woodford (2003))

15 15 / 47

16 16 / 47 Model with CPI Commitment With CPI commitment, the interest rate is determined as 0 if it 0 i t = 0 if π t π c. (3) otherwise i t Assume r + π + φ(π c t π ) > 0 (4)

17 17 / 47

18 18 / 47 Medium- to Long-Term Interest Rates A term structure model [ ] = 1 T E T 1 t i t+j. (5) A k-month forecast [ E t it+k T = 1 T E t = 1 T T +k 1 j=k i T t j=0 T +k 1 j=k i t+j ] I(E t π t+j π c ) {r + π + φ(e t π t+j π ) }. (6)

19 19 / 47 Medium- to Long-Term Interest Rates Add a term premium E t i T t+k = 1 T + 1 T T +k 1 j=k T +k 1 j=k α { 1 I(E t π t+j π c ) } I(E t π t+j π c ) {β + φ(e t π t+j π c ) }. (7) We require β > α. (8)

20 20 / 47 Estimation Strategy From the QSS, we know E t i T t+k,and E tπ t+12, E t π t+24, and E t π t+120. Real-time π t Linear interpolation E t π t+m = {(12 m)π t + me t π t+12 } /12 for 1 m 12 {(24 m)e t π t+12 + (m 12)E t π t+24 } /12 for 13 m 24 {(120 m)e t π t+24 + (m 24)E t π t+120 } /96 for 25 m 120 E t π t+120 for 121 m (9).

21 21 / 47 Estimation Strategy We estimate π c, α, β, and φ from E t i T t+k = 1 T + 1 T T +k 1 j=k T +k 1 j=k α { 1 I(E t π t+j π c ) } I(E t π t+j π c ) {β + φ(e t π t+j π c ) }. (10) As for π c, we employ a grid-search method for maximizing the likelihood function.

22 22 / 47 QSS Data Panel data Market participants from securities firms, banks, investment trusts, insurance firms, pension funds, and other private financial institutions About 150 people per month

23 23 / 47 Item Time horizon of forecast Period TIBOR yield (3 months) 1, 3, 6 months 2000M5 2008M11 Newly issued JGB yield (2 years) 1, 3, 6 months 2001M5 2008M11 Newly issued JGB yield (5 years) 1, 3, 6 months 2001M5 2008M11 Newly issued JGB yield (10 years) 1, 3, 6 months 1998M7 2008M11 Newly issued JGB yield (20 years) 1, 3, 6 months 2003M4 2008M11 CPI (excluding perishable) inflation Average of 1, 2, 10 years 2004M7 2008M11

24 24 / 47 Nonparametric Perspective

25 25 / 47 Interest rate expectations vis-a-vis 1-year inflation expectations

26 26 / 47 Interest rate expectations vis-a-vis 2-year inflation expectations

27 27 / 47 Interest rate expectations vis-a-vis 10-year inflation expectations

28 28 / 47 Interest rate expectations vis-a-vis 1-year inflation expectations after the CPI commitment policy ended

29 29 / 47 Interest rate expectations vis-a-vis 1-year inflation expectations during (red solid lines) and after (blue dashed lines) the CPI commitment policy.

30 30 / 47 Estimation

31 31 / 47 Benchmark { β + φ(πt π i t = c ) α (11) α < β φ is positive and significant π c is around 0 percent Adjusted R 2 is the highest for the 2-years yields dependent variables α β φ π c Adj R2 TIBOR 3M Y 3M Y 3M Y 3M Y 3M

32 Estimation of models with and without the CPI commitment α β φ π c π 0 Adj R 2 F test TIBOR benchmark w/o commit 93 = α (S.E.) Y benchmark w/o commit = α (S.E.) Y benchmark w/o commit = α (S.E.) Y benchmark w/o commit = α (S.E.) Y benchmark w/o commit = α (S.E.) / 47

33 33 / 47 Robustness Different models confirm our results...1 Various samples...2 A model excluding food and energy...3 A simple model without a term structure consideration...4 A modified Tobit model Unclear lower bound Tobit and panel (with fixed effect) analysis...5 A model with policy inertia...6 A model with the restriction of φ = 1.1

34 34 / 47 Concluding Remarks Future work Effects of increases in current reserves Recent episode Other countries such as Canada, Sweden, and the United States

35 35 / 47 Current account balances (Source: Bank of Japan)

36 36 / 47 Estimation for the sample periods that include three months just before and after the CPI commitment policy ended α β φ π c π 0 Adj R 2 F test TIBOR w commit w/o commit = α Y w commit w/o commit = α Y w commit w/o commit = α Y w commit w/o commit = α Y w commit w/o commit = α

37 37 / 47 Interest rate expectations vis-a-vis 1-year inflation expectations in the short period just before and just after the CPI commitment policy ended

38 38 / 47 Estimation of models with policy inertia α β φ ρ π c π 0 Adj R 2 F TIBOR w commit w/o commit 08 = α Y w commit w/o commit -12 = α Y w commit w/o commit = α Y w commit w/o commit = α Y w commit w/o commit = α

39 Tobit model using 3M forecast of 2Y yields as a dependent variable and 1Y π e as an explanatory variable i t = { γ + φπt + ε t α (12) α (preset) [#N, #N] γ φ σ ε π c = (α γ)/φ [0, 2096] [413, 1683] [919, 1177] [1242, 854] 39 / 47

40 40 / 47 Interest rate expectations vis-a-vis 1-year inflation expectations excluding food and energy during the CPI commitment policy

41 41 / 47 Interest rate expectations vis-a-vis 1-year inflation expectations excluding food and energy after the CPI commitment policy ended

42 42 / 47 3-months TIBOR yields expectations (top: means; middle: standard deviations / means; bottom: skewness (%) Core CPI (%) TIBOR 1M TIBOR 3M TIBOR 6M Core CPI QEP QEP Core CPI (%) 2.5 TIBOR 1M -.1 TIBOR 3M -0.5 TIBOR 6M -.2 Core CPI QEP Core CPI (%) 3 TIBOR 1M -4 TIBOR 3M -2 TIBOR 6M Core CPI

43 43 / 47 2-years JGB yields expectations (top: means; middle: standard deviations / means; bottom: skewness (%) QEP 2-year 1M 2-year 3M 2-year 6M Core CPI Core CPI (%) QEP Core CPI (%) 2-year 1M 2-year 3M 2-year 6M Core CPI QEP Core CPI (%) 2-year 1M 2-year 3M 2-year 6M Core CPI

44 44 / 47 5-years JGB yields expectations (top: means; middle: standard deviations / means; bottom: skewness (%) Core CPI (%) QEP year 1M 5-year 3M 5-year 6M Core CPI QEP Core CPI (%) year 1M 5-year 3M 5-year 6M Core CPI QEP Core CPI (%) year 1M 5-year 3M 5-year 6M Core CPI

45 45 / years JGB yields expectations (top: means; middle: standard deviations / means; bottom: skewness (%) Core CPI (%) 2.5 ZIRP QEP year 1M 10-year 3M 10-year 6M Core CPI ZIRP QEP Core CPI (%) year 1M 10-year 3M 10-year 6M Core CPI ZIRP QEP Core CPI (%) 3 10-year 1M 10-year 3M 10-year 6M Core CPI

46 46 / years JGB yields expectations (top: means; middle: standard deviations / means; bottom: skewness (%) QEP Core CPI (%) QEP 20-year 1M 20-year 3M 20-year 6M Core CPI Core CPI (%) QEP CORE CPI (%) 20-year 1M 20-year 3M 20-year 6M Core CPI year 1M 20-year 3M 20-year 6M Core CPI

47 Inflation expectations (top: means; middle: standard deviations / means; bottom: skewness (%) Core CPI (%) CPI 1-year QEP CPI 2-year 2.0 CPI 10-year 2.5 Core CPI 1.6 Core CPI (base year of 2005) (%) Core CPI(%) QEP CPI 1-year CPI 2-year CPI 10-year Core CPI (%) QEP CPI 1-year CPI 2-year CPI 10-year Core CPI III IV I II III IV I II III IV I II III IV I II III IV / 47

Comments on "Quantitative Easing and Bank Lending: Evidence from Japan" Takashi Nagahata

Comments on Quantitative Easing and Bank Lending: Evidence from Japan Takashi Nagahata Comments on "Quantitative Easing and Bank Lending: Evidence from Japan" by David Bowman, Fang Cai, Sally Davies, and Steven Kamin Takashi Nagahata ECB Workshop on Non-standard monetary policy measures