Exiting from QE by Fumio Hayshi and Junko Koeda

Transcription

1 Exiting from QE by Fumio Hayshi and Junko Koeda Federal Reserve Bank of San Francisco March 28 th 2014 Roger E. A. Farmer, Distinguished Professor, UCLA

2 What this Paper Does Uses a structural VAR with endogenous regime switching to study QE in Japan. I will focus on three questions Did the economy behave differently during periods of QE? How did policy behave during QE and non QE periods? Was QE effective? 2

3 Japan is Scary No growth in Industrial Production since QE QE2 QE3 Is this the future of the US? Log of Industral Production 3

4 Japan and QE QE1 QE3 QE Policy rate reached zero in March 1999 Three periods of QE since then QE1 1999m m07 QE2 2001m m06 QE3 2008m12-present Policy Rate Excess Reserves 4

5 Vars, Svars and Models E F( XYX,,, Y, U ) = 0 Theorists build models. X is a vector of endogenous variables Y is a vector of policy variables Primes denote the future U is a vector of shocks 5

6 Linearized Models Lead to Vars X = AX + BY + V Y = CX + DY + V 1 2 Private sector reduced form Policy sector reduced form Identification question How is U related to V? Hayashi-Koeda answer Reduced form private sector Model Policy sector: Two Taylor Rules 6

7 Regime Switching Models E F (,,,,, ) S XYX Y U S = 0 Model depends on regime S 7

8 Regime Switching Models Lead to Markov Switching Vars X = A X + B Y + V S1 S1 S1 Y = C X + D Y + V S1 S1 S1 1 2 E VV T = Ω S1 X = A X + B Y + V S2 S2 S2 Y = C X + D Y + V S2 S2 S2 1 2 E VV T = Ω S 2 Pr S = Si = G XY, ( ) ( ) 8

9 Estimation Estimate private sector by least squares separately in each regime Estimate policy rules and switching with maximum likelihood X = A X + B Y + V S1 S1 S1 Y = C X + D Y + V S1 S1 S1 1 2 Estimated by Least squares Estimated by Maximum likelihood 9

10 Data: The output gap and excess reserves Output Gap Excess Reserves 10

11 Data: Inflation and the Policy Rate Twelve Month Inflation Rate One Month Inflation Rate Policy Rate 11

12 Points I will Make There is a not a lot of evidence that the inflation process changes much across regimes There is strong evidence of a change in the persistence of the output process There is some evidence that QE affects output 12

13 The inflation process 13

14 Private Sector Equations 14

15 Private Sector Equations 15

16 Private Sector Equations 16

17 The output process 17

18 Private Sector Equations 18

19 Private Sector Equations 19

20 Private Sector Equations 20

21 Private Sector Equations 21

22 Twelve month or one month inflation? 22

23 Private Sector Equations 23

24 QE Period Normal Period INFL GAP INFL GAP INFL(-1) ( ) ( ) [ ] [ ] GAP(-1) ( ) ( ) [ ] [ ] C ( ) ( ) [ ] [ ] XRES(-1) ( ) ( ) [ ] [ ] R_POL(-1) ( ) ( ) [ ] [ ] R-squared INFL(-1) ( ) ( ) [ ] [ ] GAP(-1) ( ) ( ) [ ] [ ] C ( ) ( ) [ ] [ ] R_POL(-1) ( ) ( ) [ ] [ ] DUM ( ) ( ) [ ] [ ] R-squared

25 QE Period Normal Period INFL GAP INFL GAP INFL(-1) ( ) ( ) [ ] [ ] GAP(-1) ( ) ( ) [ ] [ ] C ( ) ( ) [ ] [ ] XRES(-1) ( ) ( ) [ ] [ ] R_POL(-1) ( ) ( ) [ ] [ ] R-squared INFL(-1) ( ) ( ) [ ] [ ] GAP(-1) ( ) ( ) [ ] [ ] C ( ) ( ) [ ] [ ] R_POL(-1) ( ) ( ) [ ] [ ] No price puzzle here DUM ( ) ( ) [ ] [ ] R-squared

26 Some Comments on Counterfactuals The authors conduct counterfactuals This is a minefield for ordinary Svars It is a minefield with nuclear landmines for regime switching Svars 26

27 Some words of praise The method of regime dependent Vars is an interesting extension to the Svar literature The finding of expansionary QE under regime switching is important There is a job for theorists to understand the mapping from structural models to Svars 27