Money and Banking. Lecture V: Monetary Policy Transmission Mechanisms. Guoxiong ZHANG, Ph.D. November 7th, Shanghai Jiao Tong University, Antai

Transcription

1 Money and Banking Lecture V: Monetary Policy Transmission Mechanisms Guoxiong ZHANG, Ph.D. Shanghai Jiao Tong University, Antai November 7th, 2016

2 Monetary Policy and Its Effects: a Huge Black Box Source:

3 Road Map Is monetary policy effective? monetary policy shocks monetary policy effects Conventional transmission mechanisms of monetary policy traditional interest rate channel asset prices channels credit channels: bank lending channel vs balance sheet channel other credit channels: cash flow, unanticipated inflation/deflation, household liquidity effect Credit channels: from theory to data

4 Monetary Policy Shocks The primary difficulty of identifying the casual effects of monetary policy is that the policy itself depends on the economic conditions (endogeneity issue) Economists have several ways to solve this issue: structural VAR ( Cochrane (1998) ) narrative approaches (Romer and Romer (1989, 2004)) using information from financial markets ( Barakchian and Crowe (2013))

5 Structural VAR and Monetary Policy Shocks Recall the VAR example we have seen in Lecture III: ( ) ( ) ( 1, αy,0 yt α y y,1 =, ) ( ) ( αi y,1 yt 1 α y α i,0, 1 i t α y i,1, + y,p, αy,p i αi i,1 i t 1 α y i,p, αi i,p ) ( yt p i t p ) One common way to identify the monetary policy shock is to assume α y,0 = 0: current monetary policy responds to current economic conditions but does not affect current economic conditions; There are other ways to structurarize the VAR: all of them have to base on some ad hoc assumptions.

6 Narrative Approaches Friedman and Schwartz (1963) was actually the first to use this approach: identify unsusual monetary policy regimes in a narrative way; Their approach was carried over by Romer and Romer (1989): only focused on large contractionary monetary policy shocks, relatively free of endogeneity with respect to output; Romer and Romer (2004) combines narratives with statistical methods: Use narrative accounts of each FOMC meetings to determine intended federal funds rate: eliminating economic conditions effect on market federal funds rate; Use FED s internal forecasts of inflation, output and unemployment to purge part of intended federal funds rate that is response to anticipation of future economic conditions; The rest of the intended federal funds rate should be relatively free of endogeneity and anticipatory actions.

7 Identifying Monetary Policy Shock Through Financial Markets: Barakchian and Crowe (2013) Assume FED s response function to be: S t = f(ω t) + s t To identify the shock s t we rely on the efficient market hypothesis: Assume that Then we have t 1Ŝt = Ep t 1S t + ξ t 1 = E p t 1f(Ω t) + ξ t 1 tŝt = Ep t S t + ξ t = S t + ξ t E p t 1f(Ω t) = f(ω t) ξ t 1 = ξ t s t = t Ŝ t t 1 Ŝ t where t Ŝ t and t 1 Ŝ t can be infered from the federal funds future contracts.

8 Romer and Romer (1989): Monetary Policy Shocks

9 Romer and Romer (1989): Monetary Policy Shock s Effect on Output

10 Romer and Romer (2004): Monetary Policy Shocks

11 Romer and Romer (2004): Monetary Policy Shock s Effect on Output

12 Romer and Romer (2004): Monetary Policy Shock s Effect on Price Level

13 Barakchian and Crowe (2013): Monetary Policy Shocks

14 Barakchian and Crowe (2013): Monetary Policy Shock s Effects

15 Monetary Policy Shocks in China: a Narrative Approach Sun (2013) adopted the method from Romer and Romer (1989) on the study of China s monetary policy shocks; Narratives come from PBoC s quarterly monetary policy implementation report, senior PBoC officials public speech, and national leaders comments on monetary policy; Also focused only on contractionary monetary policy shock.

16 Sun (2013): Monetary Policy Shocks in China

17 Sun (2013): Monetary Policy Shock s Effect on Output

18 Sun (2013): Monetary Policy Shock s Effect on Price Level

19 Monetary Policy Transmission Mechanism: Traditional Interest-Rate Channels As prices are sticky, lower nominal interest rate leads to lower real interest rates; Lowe real interest rates cause higher private investment as well as consumer durable expenditure; As a result aggregate demand is raised by a lower nominal interest rate; When the nominal interest rate hits the zero-lower-bound, the central bank can simply raise the expectation on future inflation by committing expansionary monetary policy in the future; In this way the real interest rate can still be lowered.

20 Monetary Policy Transmission Mechanism: Other Asset Prices Channels Exchange rate channel: r E NX Y ad in fact the relationship between exchange rate and net export is far from clear (J-curve does not really exisits) Tobin s q channel: r P s q I Y ad lower real interest rates make stocks more attractive and hence raises stock prices and in turn lowers firms financing cost; Wealth effect channel: r P s wealth consumption Y ad this channel and the above channel also work for the housing market

21 Monetary Policy Transmission Mechanism: Credit Channels Bank lending channel: r banks financing cost bank loans I Y ad in fact the relationship between exchange rate and net export is far from clear (J-curve does not really exisits) Balance sheet channel: r P s firms net worth adverse selection and moral hazard lending I Y ad there is also a balance sheet channel from the household ( Aladangady (2014) ) It is very difficult to disentangle these two channels from ordinary observational data ( simultaneity problem: supply and demand).

22 Monetary Policy Transmission Mechanism: Other Credit Channels Cash flow channel: r firms cash flow adverse selection and moral hazard lending I Y ad lower real interest rate means lower interest payment to corporate debts Unanticipated inflation channel: r π unanticipated P firms net worth adverse selection and moral hazard lending I Y ad in contrary to debt deflation Household liquidity channel: r P s value of households financial assets likelihood of financial distress consumer durable and housing expenditure Y ad

23 Monetary Policy Transmission Mechanism: From Theory to Data Empirical study on monetary policy transmission mechanism has been focused on the credit channel; Bernanke and Gertler (1995) provides a paramedic view on this channel; Using VAR on aggregate data they found several stylized facts on this channel; The existence of balance sheet channel has been verified by several studies using firm level data; But the existence of bank lending channel is far from conclusive; Essentially all predictions that derived from the bank lending channel can also been derived from the balance sheet channel; Bernanke and Gertler (1995) was not able to solve this issue.

24 Bernanke and Gertler (1995): Interest Rates and Lending Standards

25 Bernanke and Gertler (1995): Interest Rates and Firms Balance Sheets

26 Bernanke and Gertler (1995): Monetary Policy and Firms Performance

27 Attempts to Identify the Bank Lending Channel: Kashyap et al. (2003) Use large scale firm level data; Find that during monetary contraction, on average firms borrow less but issue more commercial papers; Does this necessary mean that there is a bank lending channel?

28 Attempts to Identify the Bank Lending Channel: Kashyap and Stein (2000) Use large scale bank loan data( one million, hand collected); Find that monetary contraction s effect on lending is stronger for bank with less liquid balance sheets; If less liquid banks are indeed less likely to find alternative fund source to compensate the negative effect from a contractionary monetary policy, then the above finding an indirectly prove the existence of a bank lending channel;

29 Attempts to Identify the Bank Lending Channel: Khwaja and Mian (2008) Use a smart way to control the demand side from the firms: sudden and unanticipated nuclear test that causing sanction from the IMF This sanction reduces banks liquidity; They find that bank loans are reduced controlling all fixed effects of the firms; Problem: is the demand side really 100% controlled?