Commodity price shocks and impefectly credible macroeconomic policies

Transcription

1 Commodity price shocks and impefectly credible macroeconomic policies Juan Pablo Medina (IMF) Claudio Soto (Central Bank of Chile) November 2012 uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

2 Export price volatility and output volatility Positive correlation between export price volatility and output volatility Positive correlation between the share of primary exports and output volatility Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

3 Policy links between commodity prices and activity Fiscal policy Commodity exports are a relevant source of government nancing, in particular, in emerging economies The elasticity of public revenues to commodity price shocks lies between 0,2 and 0,7 for emerging markets (Ehrhart and Guerineau, 2012; Medina 2010) and between 0,1 and 0,2 for commodity exporting advanced economies (Medina 2010) Then, commodity price uctuation a ect the volatility of public revenues If the government pursue a balanced budget, then public expenditure will be more volatile ) more output volatility Government may commit to an a-cyclical behavior, but need to be credible and transparent Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

4 Policy links between commodity prices and activity Figure: Public revenues linked to commodity exports (%GDP) source: Medina (2010) uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

5 Policy links between commodity prices and activity Monetary policy Commodity price increases may lead to exchange rate appreciations If central bank dislike currency appreciation, monetary authority may be tempted to deviate from systematic behavior and pursue a more expansive policy A commitment to stick to a rule may increase e ectiveness of policy, but need to be credible and transparent Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

6 This paper We develop a DSGE where macro policies may lack credibility We distinguish credibility from transparency Transparency allow private agents to learn fast the true policy rules followed by the authorities Simulate impulse-response to a commodity price shock under alternative con gurations Present some new empirical evidence Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

7 Outline Sketch of model Model simulation Empirical evidence Conclusions Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

8 Model Small open economy DSGE model uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

9 Model Small open economy DSGE model Ricardian and non-ricardian households uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

10 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

11 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages Firms produce using capital and labor uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

12 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages Firms produce using capital and labor Commodity sector Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

13 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages Firms produce using capital and labor Commodity sector Stochastic endowment Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

14 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages Firms produce using capital and labor Commodity sector Stochastic endowment Commodity price determined in the international market Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

15 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages Firms produce using capital and labor Commodity sector Stochastic endowment Commodity price determined in the international market All the production is exported Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

16 Model Small open economy DSGE model Ricardian and non-ricardian households Sticky prices and wages Firms produce using capital and labor Commodity sector Stochastic endowment Commodity price determined in the international market All the production is exported A share of the revenues belong to the government (taxes, property) uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

17 Policy rules Monetary policy " i t + ϕi # it (1 ϕi )ϕ πt+1 π = E t exp ζ 1 + i 1 + i 1 + π m,t {z } with ζ m,t N 0, σ 2 m 1+ei t : systematic behavior uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

18 Policy rules Monetary policy " i t + ϕi # it (1 ϕi )ϕ πt+1 π = E t exp ζ 1 + i 1 + i 1 + π m,t {z } 1+ei t : systematic behavior with ζ m,t N 0, σ 2 m Fiscal policy P G,t G t Y t S t P = R G,t + τ + t Y Co,t χp P Y,t Y t Y Co exp ζ t P Y,t Y G,t t {z } eg t : systematic behavior with ζ G,t N 0, σ 2 G uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

19 Policy rules Monetary policy " i t + ϕi # it (1 ϕi )ϕ πt+1 π = E t exp ζ 1 + i 1 + i 1 + π m,t {z } 1+ei t : systematic behavior with ζ m,t N 0, σ 2 m Fiscal policy P G,t G t Y t S t P = R G,t + τ + t Y Co,t χp P Y,t Y t Y Co exp ζ t P Y,t Y G,t t {z } eg t : systematic behavior with ζ G,t N 0, σ 2 G Parameters σ 2 m and σ 2 G determine the reputation of both authorities uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

20 Policy rules under imperfect credibility Monetary policy " i t + ϕi it (1 ϕi )ϕ πt+1 π P (1 ϕi # )ϖ π = E Co,t t 1 + i 1 + i 1 + π Pt {z } perceived systematic behavior exp ζ m,t Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

21 Policy rules under imperfect credibility Monetary policy " i t + ϕi it (1 ϕi )ϕ πt+1 π P (1 ϕi # )ϖ π = E Co,t t 1 + i 1 + i 1 + π Pt {z } perceived systematic behavior exp ζ m,t Fiscal policy P G,t G t = P Y,t Y t + χ Y t P Co,t St P p Co + t Y Co,t ϖ G exp ζ P Y,t Y G,t t R G,t + τ Y t Pt {z } perceived systematic behavior Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

22 Imperfect credibility Private agent assign di erent probabilities that either the monetary authority or the scal authority are deviating from their systematic behavior uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

23 Imperfect credibility Private agent assign di erent probabilities that either the monetary authority or the scal authority are deviating from their systematic behavior There are four possible cases: (1) non is deviating; (2) the monetary authority is deviating; (2) the scal authority is deviating and (4) both are deviating uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

24 Imperfect credibility Private agent assign di erent probabilities that either the monetary authority or the scal authority are deviating from their systematic behavior There are four possible cases: (1) non is deviating; (2) the monetary authority is deviating; (2) the scal authority is deviating and (4) both are deviating We denote by pr i,t the probability of each one of the four cases above in period (i = 1,..., 4) uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

25 Imperfect credibility Private agent assign di erent probabilities that either the monetary authority or the scal authority are deviating from their systematic behavior There are four possible cases: (1) non is deviating; (2) the monetary authority is deviating; (2) the scal authority is deviating and (4) both are deviating We denote by pr i,t the probability of each one of the four cases above in period (i = 1,..., 4) Monetary and scal policy do not deviate from their rules, but private agents learn this slowly over time uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

26 Lack of transparency Monetary policy dev i,t = i t ei t + e 1,t P = ζ m,t (pr 1,t + pr 4,t ) (1 ϕ i )ϖ π ln Co,t where e 1,t N 0, σ 2 1 P t + e 1,t uan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

27 Lack of transparency Monetary policy dev i,t = i t ei t + e 1,t P = ζ m,t (pr 1,t + pr 4,t ) (1 ϕ i )ϖ π ln Co,t where e 1,t N 0, σ 2 1 Fiscal policy dev G,t = g t eg t + e 2,t P = ζ G,t + (pr 3,t + pr 4,t ) ϖ G Γ ln Co,t where e 2,t N 0, σ 2 2 P t P t + e 1,t + e 2,t Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

28 Learning Based on the information available up to t the vector of states ζ tjt 1 : 1, private agents infer ζ tjt 1 = pr 1,tjt 1 pr 2,tjt 1 pr 3,tjt 1 pr 4,tjt 1 ζ m,tjt 1 ζ G,tjt 1 The law of motion of the endogenous variables in the model is given by: x t = Px t 1 +! 4 Q Co,j pr j,tjt 1 pco,t + Q m ζ m,tjt 1 + Q G ζ G,tjt 1 j=1 Using information available in, private agents update their inference regarding ζ tjt using the Kalman lter and then project ζ t+1jt Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

29 Model calibration Parameter Value Description h 0.7 Habit persistency λ 0.7 Share of non-ricardian households 1 φ 0.25 Frequency of price adjustments G /Y 0.12 Share of public consumption χ 0.4 Share of revenues from commodities owned by the gov. Y Co /Y 0.1 Share of commodity production Parameter imperf. credibility and lack of transp. ϖ π 0.05 ϖ G 0.5 σ 2 i 0.05 σ 2 G 0.10 σ σ Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

30 Simulations Commodity price shock of 10% We compare the case of full credibility with the other three cases discussed above For the cases where there is lack of credibility, private agents begin assigning a probability of 25% of each of one of them Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

31 Impulse response to a commodity price shock Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

32 Evolution of beliefs Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

33 Monetary policy transparency Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

34 Evolution of beliefs Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

35 Monetary policy transparency and reputation Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

36 Monetary policy transparency and reputation, and scal policy transparency Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

37 Empirical evidence We estimate a set of cross-country regressions of the form: σ 2 y,i = α 0 + α 1 σ 2 P x,i + α 2σ 2 P x,i q i + ε i where σ 2 y,i is the volatility of output growth, σ 2 P x,i is the volatility of the export/commodity price relevant to country i, and q i is a variable that measures either credibility or transparency of policies Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

38 Empirical evidence Measures of credibility Run auxiliary regressions to estimate monetary policy rules for the period ; We use the standard deviation of the residuals as our proxy for monetary credibility For scal policy credibility we use the standard deviation of measures of the structural balance of government for the period from the IMF Measures of transparency Monetary policy: Dincer and Eichengreen (2009) index Fiscal policy: Open Budget Index (OBI) Additionally, we consider dummy variables for whether countries have IT regime or whether they have a scal rule in place according to the classi cation of the IMF Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

39 Results I Dependent variable: output growth volatility (1) (2) (3) (4) (5) SD Comm. price ( ) ( ) ( ) ( ) ( ) SD. Comm. price X Mon. Transp. 4.12e 05* (2.16e 05) SD Comm. price X Mon. Cred. 3.46e 07*** (1.21e 07) SD Comm. price X Fiscal Transp. 3.27e 07 (2.69e 06) SD Comm. price X Fiscal Cred. 8.42e 05* (3.22e 05) SD Comm. price X Inf. Target ( ) Constant ** *** *** *** (0.0126) ( ) ( ) (0.0377) ( ) Observations R squared Robust standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1 SD Comm. price = Standard Deviation of Commodity Price Growth Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

40 Results II Dependent variable: output growth volatility (1) (2) (3) (4) (5) (6) SD Exports Defl ** *** ** *** ** ( ) ( ) ( ) ( ) ( ) ( ) SD Exports Defl. X Mon. Transp * (6.10e 05) SD Exports Defl. X Mon. Cred. 4.22e 07 (7.72e 07) SD Exports Defl. X Fiscal. Transp. 9.47e 06 (6.71e 06) SD Exports Defl. X Fiscal. Cred *** (7.94e 05) SD Exports Defl. X Fiscal Rule *** ( ) SD Exports Defl. X Inf. Target *** ( ) Constant *** *** *** *** *** *** ( ) ( ) ( ) ( ) ( ) ( ) Observations R squared Robust standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1. SD Exports Defl. = Standard Deviation of Deflator of Exports Growth Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

41 Results III Dependent variable: output growth volatility (1) (2) (3) (4) (5) (6) SD Exports Defl. X % Prim. Exports/GDP 8.49e 06** 1.07e 05*** 5.24e e e e 06 (3.85e 06) (3.24e 06) (4.47e 06) (1.57e 05) (3.25e 06) (5.93e 06) SD Exports Defl. X % Prim. Exports/GDP 1.88e 06* X Mon. Transp. (1.08e 06) SD Exports Defl. X % Prim. Exports/GDP 6.85e 09 X Mon. Cred. (1.98e 08) SD Exports Defl. X % Prim. Exports/GDP 1.17e 07 X Fiscal Transp. (1.23e 07) SD Exports Defl. X % Prim. Exports/GDP 4.14e 06** X Fiscal Cred. (1.86e 06) SD Exports Defl. X % Prim. Exports/GDP 2.07e 05*** X Inf. Target (6.29e 06) Constant *** *** *** *** *** *** ( ) ( ) ( ) ( ) ( ) ( ) Observations R squared Robust standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1 SD Exports Defl. = Standard Deviation of Deflator of Exports Growth; % Prim. Exports/GDP = Share of Primary Export to Total Product Export Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27

42 Conclusions Isolating the economy from terms of trade movements requires enough degrees of exchange rate exibility and a scal policy that shields public spending from uctuations in revenues Well designed monetary and scal rules that induce these types of systematic behavior is a necessary condition to reduce output volatility However, the mechanic implementation of these rules is not enough: these rules will not be e ective if monetary and scal authorities lack credibility and are not transparent The empirical evidence in this paper tends to support that view Institutional arrangement that enhance the transparency of policies and good reputation will make macro policies more e ective in isolating the economy from commodity price uctuations Juan Pablo Medina (IMF), Claudio Soto (Central Commodity Bank ofprice Chile) shocks () and impefectly credible macroeconomic November policies / 27