Spillover Effects of U.S. Monetary Policy and Policy Uncertainty on Chinese Economy

Transcription

1 Spillover Effects of U.S. Monetary Policy and Policy Uncertainty on Chinese Economy Steven Wei Ho, Ji Zhang, and Hao Zhou PBC School of Finance Tsinghua University

2 Motivation Background U.S. unconventional monetary policy zero lower bound LSAP QEs tapering impact on emerging market, especially China China is likely to pass U.S. as world s leading economic power

3 Motivation Questions How do U.S. monetary policy and policy uncertainty shocks affect China, especially during and after the Great Recession? What s the transmission channel? Which one is more important? Is there any structural change in the past few years?

4 This Paper What Do We Do? Factor-Augmented VAR U.S. monetary policy U.S. policy uncertainty both before and at the ZLB

5 This Paper Main Results At the ZLB, U.S. monetary policy shocks do have significant impact on Chinese monetary policy and real economy U.S. policy uncertainty shocks are less important Comparing the cases before and at the ZLB, Dynamics of Chinese macro variables are different under the same shocks Relative importance of the two shocks changes Indicate structural changes in both U.S. policy transmission and Chinese economy

6 Model FAVAR Bernanke et al(25, QJE) ( Ft Y t ) = Φ(L) ( Ft 1 Y t 1 F t : factors (T K), X t : macro variables (T N, N > K), Y t : policy indicators. ) + ν t, (1) X t = Λ f F t + Λ y Y t + e t. (2)

7 Model Why FAVAR? Overcomes some shortcomings of structural VAR: use more information no arbitrary choice of a specific data series impulse responses for all variables

8 Data Data 162 Chinese macroeconomic series (real activity variables, policy variables, price variables) U.S. policy rate: Wu-Xia shadow rate (Wu and Xia (214)) U.S. policy uncertainty: news-based measure (Baker, Bloom and Davis (213)) sample period: 2M1:214M2

9 Data Ajustment New Year effect (except policy variables) (Fernald et al (213)) seasonal adjustment (US Census Bureau X13 Software) EM algorithm for missing values (Stock and Watson (22, JEBS)) fast/slow moving variable (asset price vs. wages) Transformation for stationarity: no transformation, first difference, logarithm, first log difference (Bernanke et al(25, QJE), Stock and Watson (22, JEBS))

10 Estimation Estimation and Identification Bernanke et al(25, QJE) factors are constructed to be orthogonal to policy rate and uncertainty pc t : principal components of all series, pc s t : principal components for slow-moving series run the regression: factors are constructed to be pc t = b pc pc s t + b Y pcy t + η pc t F t = pc t b Y pcy t

11 Estimation Estimation and Identification Bernanke et al(25, QJE) estimate the observation equation by OLS (2 lags) estimate the transition equation by OLS identification is achieved through recursive assumption two subsamples: 2M1-28M9, 28M12-214M2

12 R 2 R 2 Variables Full Name R 2 before ZLB at ZLB SSEI Shanghai Stock Exchange Composite Index PE ratio (SSE All) PE ratio (SSE All Stocks) PE ratio (SSE A) PE ratio (SSE A-Shares ) PE ratio (SSE Fin) PE ratio (SSE Finence) PE ratio (SSE RE) PE ratio (SSE Real Estate) PE ratio (SSE Const) PE ratio (SSE Construction) PE ratio (SSE Manu) PE ratio (SSE Manufacturing) Loan The Total Amount of Loans Loan Rate (1yr) Nominal Loan Rate (1 Year) HH DR (1yr) Household Saving Deposit Rate (1 Year) SHIBOR (1d) SHIBOR (1 Day) Bond Index (Inter Bank ST) Inter-Bank Short Term Government Bond Index FES (USD 1yr) Foreign Exchange Swap (USD 1 Year) CPI CPI InvestRE Investment in Real Estate NHS New Housing Start Comm Bldg Sales Commodity Building Sales PMI Manufacturing Purchase Management Index (Manufacturing) PMI new orders Purchase Management Index (New Orders) Auto Sales (DM) Automobile Sales (Domestically Made) Macro index Macroeconomic Index

13 IRFs Outline after 28M12 (ZLB is binding in the U.S.) IRFs to U.S. monetary policy shock IRFs to U.S. policy uncertainty shock before 28M12 (ZLB is not binding in the U.S.) IRFs to U.S. monetary policy shock

14 IRFs IRFs to USMP Shock at the ZLB U.S. Monetary Policy Shock at ZLB Policy Rate.2 USPU.5 Loan Loan Rate (1yr).15 HH DR (1yr).2 SHIBOR (1d) Bond Index (Inter Bank ST).2 FES (USD 1yr) 2 CPI

15 IRFs IRFs to USMP Shock at the ZLB U.S. Monetary Policy Shock at ZLB Policy Rate.2 USPU.4 SSEI PE ratio (SSE All).1 PE ratio (SSE A) PE ratio (SSE Fin) PE ratio (SSE RE) PE ratio (SSE Const).15 PE ratio (SSE Manu)

16 IRFs IRFs to USMP Shock at the ZLB U.S. Monetary Policy Shock at ZLB Policy Rate.2 USPU.5 InvestRE NHS.6 Comm Bldg Sales.2 PMI Manufacturing PMI new orders.2 Auto Sales (DM).5 Macro index

17 IRFs Is It Interest Rate Channel? Impulse Responses of the Interest Rates at the ZLB.5 Treasury rate:6 month.5 Treasury rate:1 year Treasury rate:5 year Treasury rate:15 year Treasury rate:1 year Treasury rate:3 year

18 IRFs Is It Hot Money? U.S. Monetary Policy Shock at ZLB.4 hot money

19 IRFs Possible Channel U.S. monetary policy shock does not affect Chinese economy through market determined interest rate channel Responses of Chinese policy and international capital inflow are important U.S. MP capital inflow / people s expectations Chinese policy rates real economy

20 IRFs IRFs to USPU Shock at the ZLB U.S. Policy Uncertainty Shock at ZLB.6 Policy Rate.15 USPU.1 Loan Loan Rate (1yr).15 HH DR (1yr).3 SHIBOR (1d) Bond Index (Inter Bank ST).3 FES (USD 1yr) 1 CPI

21 IRFs IRFs to USPU Shock at the ZLB U.S. Policy Uncertainty Shock at ZLB.6 Policy Rate.15 USPU.15 SSEI PE ratio (SSE All).4 PE ratio (SSE A).1 PE ratio (SSE Fin) PE ratio (SSE RE).2 PE ratio (SSE Const).15 PE ratio (SSE Manu)

22 IRFs IRFs to USPU Shock at the ZLB U.S. Policy Uncertainty Shock at ZLB.6 Policy Rate.15 USPU.2 InvestRE NHS.4 Comm Bldg Sales.3 PMI Manufacturing PMI new orders.4 Auto Sales (DM) 1.5 Macro index

23 IRFs IRFs to USMP Shock before the ZLB U.S. Monetary Policy Shock before ZLB.1 Policy Rate.3 USPU.3 Loan Loan Rate (1yr).5 HH DR (1yr).15 SHIBOR (1d) Bond Index (Inter Bank ST).15 FES (USD 1yr) 1 CPI

24 IRFs IRFs to USMP Shock before the ZLB U.S. Monetary Policy Shock before ZLB.1 Policy Rate.3 USPU.1 SSEI PE ratio (SSE All).1 PE ratio (SSE A).1 PE ratio (SSE Fin) PE ratio (SSE RE).1 PE ratio (SSE Const).1 PE ratio (SSE Manu)

25 IRFs IRFs to USMP Shock before the ZLB U.S. Monetary Policy Shock before ZLB.1 Policy Rate.3 USPU.4 InvestRE NHS.3 Comm Bldg Sales.15 PMI Manufacturing PMI new orders.6 Auto Sales (DM).4 Macro index

26 IRFs Why Different at and before the ZLB? change in U.S. monetary policy transmission responses of policy rate change in Chinese economy bonds market liberalization of the interest rates managed floating exchange rate

27 IRFs USMP and US Interest Rates (Zhang(214)) Year Treasury Year TIPS 2 Year Breakeven Year Treasury Year TIPS Year Breakeven Year Treasury Year TIPS Year Breakeven

28 VD Variance Decomposition Variance Decomposition Ratio (MP/PU) Variables before ZLB at ZLB 1m 2m 6m 12m 1m 2m 6m 12m SSEI PE ratio (SSE All) PE ratio (SSE A) PE ratio (SSE Fin) PE ratio (SSE RE) PE ratio (SSE Const) PE ratio (SSE Manu) Loan Loan Rate (1yr) HH DR (1yr) SHIBOR (1d) Bond Index FES (USD 1yr) CPI InvestRE NHS Comm Bldg Sales PMI Manufacturing PMI new orders Auto Sales (DM) Macro index

29 Conclusion Results At the ZLB, U.S. monetary policy shocks do have significant impact on Chinese real economy The impact transmits not through market interest rate channel Possible transmission mechanism: hot money and people s expectations Chinese policy rates real economy U.S. policy uncertainty shocks are less important Comparing the cases before and at the ZLB, Dynamics of Chinese macro variables are different under the same shocks Relative importance of the two shocks changes Indicate structural changes in both U.S. policy transmission and Chinese economy

30 THANK YOU!