Banque nationale de Belgique SA · EURONEXT:BNB
Banque nationale de Belgique SA — Investor Presentation (September 2017)
38 slides from Banque nationale de Belgique SA's investor presentation, September 2017
Slide 1: Reassessing the role of oil price shocks and monetary policy in the Great Stagflation
Reassessing the role of oil price shocks and monetary policy in the Great Stag‡ation Qazi Haque Nicolas Groshenny Mark Weder September 2017
Slide 2: Our Questions
Our Questions What caused the Great Stag‡ation? I Oil price shocks of 1973-74 and 1979-80 ? I Bad monetary policy? (Why no stag‡ation in aftermath of 2000’s oil shocks?)
Slide 3: Spot Crude Oil Price: West Texas Intermediate (WTI)
Spot Crude Oil Price: West Texas Intermediate (WTI) 200 160 Percent Change from Year Ago 120 80 40 0 -40 -80 1970 1975 1980 1985 1990 1995 2000 2005 Source: Federal Reserve Bank of St. Louis fred.stlouisfed.org myf.red/g/eHBS
Slide 4: CPI and Real GDP, percent change from year ago
Consumer Price Index for All Urban Consumers: All Items Real Gross Domestic Product 15.0 12.5 10.0 Percent Change from Year Ago 7.5 5.0 2.5 0.0 -2.5 -5.0 1970 1975 1980 1985 1990 1995 2000 2005 Sources: BEA, BLS fred.stlouisfed.org myf.red/g/eDnB
Slide 5: CPI and Effective Federal Funds Rate
Consumer Price Index for All Urban Consumers: All Items Effective Federal Funds Rate 17.5 15.0 Percent Change from Year Ago , Percent 12.5 10.0 7.5 5.0 2.5 0.0 -2.5 1960 1970 1980 1990 2000 2010 Sources: BLS, Board of Governors fred.stlouisfed.org myf.red/g/e4Np
Slide 6: Key papers
Key papers Small NK models with passive monetary policy I Clarida, Gali, Gertler (2000); Lubik and Schorfheide (2004) I Coibion and Gorodnichenko (2011); Hirose, Kurozumi, van Zandweghe (2017) NK models with oil I Blanchard and Gali (2009); Blanchard and Riggi (2013) I Bodenstein, Erceg, Guerrieri (
Slide 7: What we do
What we do Estimate NK model with oil over determinacy/indeterminacy region I Sample period: 1966:I-1979:II I Oil price shock treated as observable I Test for indeterminacy
Slide 8: New Keynesian model with oil [Blanchard-Gali (2009)]
New Keynesian model with oil [Blanchard-Gali (2009)] I Oil as input in production and consumption I Real oil price as exogenous I Habit formation in consumption I Real wage rigidity I Trend in‡ation
Slide 9: People
People " # ∞ 1+ ϕ N max E0 ∑ β dt ln Ct t et 1 hC νt t t =0 1+ϕ where 1 χ Ct = Θχ Cm,t Cq,t χ and Z 1 ε ε 1 ε 1 Cq,t = Cq,t (i ) ε di 0 subject to period budget constraint Wt Nt + Rt Bt 1 + Πt Pc ,t Ct + Bt
Slide 10: Household's optimality conditions
Household’s optimality conditions dt Rt dt + 1 = βEt Ct hCt 1 π c ,t +1 Ct +1 hCt γ Wt Wt 1 ϕ 1 γ = νt Nt (Ct hCt 1 ) Pc ,t Pc ,t 1 χ Pc ,t Pq,t St P m,t where St P q,t follows exogenous process
Slide 11: Final good firm
Final good …rm Z 1 max Pq,t Qt Pq,t (i )Qt (i )di Q t (i ) 0 subject to Z 1 ε ε 1 ε 1 Qt = Qt ( i ) ε di 0
Slide 12: Intermediate good firm
Intermediate good …rm 2 P q,t (i ) 3 ∞ Qt +j (i ) λ t +j 4 P q,t n max Et ∑ ξ j βj o 5 +j W t +j (1 α)P m,t +j α P q,t (i ) j =0 λ0 1 α (1 α)P q,t +j A t +j αW t +j Q t + j ( i ) subject to ε Pq,t (i ) Qt + j ( i ) = Qt +j Pq,t +j 1 α Qt (i ) = [At Nt (i )] Mt (i )α
Slide 13: Monetary Policy
Monetary Policy et ρ 1 ρR Rt = Rt R 1 R expfσR εR ,t g where ψπ n et = ( r π ) π c ,t ω π q,t 1 ω xt oψx Yt ψ∆y R π π x Yt 1
Slide 14: Sunspot solutions
Sunspot solutions I Canonical form of LRE model Γ0 ( θ ) $t = Γ1 ( θ ) $t 1 + Ψ ( θ ) εt + Π ( θ ) η t I Sunspot equilibria f)εt + Φζ (θ )ζ $t = Φ(θ )$t 1 + Φε (θ, M t e centered at continuity solution M I Priors for M (θ ) e = M (θ ) + M M where M N (0, 1)
Slide 15: Continuity of impulse responses at boundary (LS 2004)
Continuity of impulse responses at boundary (LS 2004) 8θ 2 ΘI , construct vector θ̃ = g (θ ) that lies on the boundary M (θ ) = arg min discrepancy between ∂Y0t (θ, M (θ )) and ∂Y0t (θ̃, .) ∂et ∂et
Slide 16: Bayesian estimation with Sequential Monte Carlo
Bayesian estimation with Sequential Monte Carlo I Sequence of tempered posteriors [p (XT jθ S , S )]φn p (θ S jS ) π n (θ S ) = R θS [p (XT jθ S , S )]φn p (θ S jS )d θ S I Generate weighted draws from the sequence of posteriors n oN N fπ n (θ )gn =φ 1 ) θ in , Wni i =1 I N = 10000 and Nφ = 200
Slide 17: Calibrated Parameters
Calibrated Parameters β 0.99 ε 11 ϕ 1 α 0.015 / 0.012 χ 0.023 / 0.017 ρs 0.995
Slide 18: Priors
Priors Parameter Range Density Prior Mean St. Dev ψπ Gamma 1.10 0.50 ψx Gamma 0.125 0.10 ψ∆y Gamma 0.125 0.10 ρR Beta 0.50 0.20 ω Beta 0.50 0.20 π Normal 1.0 0.50 R Gamma 1.50 0.25 g Normal 0.50 0.10 ξ Beta 0.50 0.05 γ Beta 0.50 0.20 h Beta 0.50 0.10 Prior probabililty of determinacy is 0.50.
Slide 19: Priors (cont'd)
Priors (cont’d) Name Range Density Prior Mean St. Dev ρd [0,1) Beta 0.70 0.10 ρν [0,1) Beta 0.70 0.10 σs R+ Inv-Gamma 5.00 2.00 σg R+ Inv-Gamma 0.50 0.20 σr R+ Inv-Gamma 0.50 0.20 σd R Inv-Gamma 0.50 0.20 σν R+ Inv-Gamma 0.50 0.20 σζ R+ Inv-Gamma 0.50 0.20 Ms ,ζ R Normal 0.00 1.00 Mg ,ζ R Normal 0.0
Slide 20: Data
Data Sample period: [1966:I - 1979:II] , [1984:I - 2008:II] Observables: ∆yt , Rt , π c ,t , π q,t , ∆wt I Real GDP per capita (quarterly growth rate) I Fed Funds rate I CPI (quarterly growth rate) I Core CPI (quarterly growth rate) I Nominal Hourly Compensation in NFB sector, de‡ated with CPI (quar
Slide 21: Indeterminacy test
Indeterminacy test Log-data density Probability Period Det. Indet. Det. Indet. 1966:I - 1979:II -228.89 -241.06 1 0 1984:I - 2008:II -230.03 -251.05 1 0
Slide 22: Robustness check #1
Robustness check #1 Observables: ∆yt , Rt , π q,t , ∆wt , st st : Spot Oil Price (West Texas Intermediate) de‡ated by core CPI Log-data density Probability Period Det. Indet. Det. Indet. 1966:I - 1979:II -451.89 -458.30 1 0 1984:I - 2008:II -571.05 -608.87 1 0
Slide 23: Robustness check #2
Robustness check #2 Observables: ∆yt , Rt , π c ,t , π q,t , ∆wt1 , ∆wt2 wt1 : Nominal hourly compensation in NFB sector (NIPA), de‡ated by CPI wt2 : Average hourly earnings of production and non-supervisory employees (BLS), de‡ated by CPI Log-data density Probability Period Det. Indet. Det. Indet.
Slide 24: Understanding the 1970's determinacy outcome
Understanding the 1970’s determinacy outcome Determinacy vs Indeterminacy 1966:I - 1979:II Log density Probability Det. Indet. Det. Indet. NK(∆yt , Rt , π c ,t ) [α, χ, γ = 0, ω = 1] -121.1 -118.8 0.09 0.91 BG(∆yt , Rt , π c ,t ) [ω = 1] -123.0 -118.3 0.01 0.99 BG(∆yt , Rt , π c ,t , π q,t ) -157.9
Slide 25: Parameter Estimates 1966:I - 1979:II
Parameter Estimates 1966:I - 1979:II NK (Indet) BG (Indet) BG (Indet) BG (Det) BG (Det) 3 obs 3 obs 4 obs 5 obs 6 obs ψπ 0.94 0.94 0.92 1.55 1.51 (0.11 ) (0.11 ) (0.12 ) (0.19 ) (0.17 ) ψx 0.14 0.21 0.30 0.03 0.03 (0.11 ) (0.14 ) (0.11 ) (0.02 ) (0.03 ) ψ∆y 0.11 0.12 0.10 0.46 0.35 (0.07 ) (0.07 ) (
Slide 26: Indeterminacy region
GNK Model, 3 obs GNK-BG Model, 3 obs GNK-BG Model, 4 obs GNK-BG Model, 6 obs Indeterminacy region
Slide 27: Commodity Price Shock
Commodity Price Shock Headline Inflation Core Inflation 0.4 0.15 =0 0.3 = 0.5 0.1 = 0.9 0.2 0.05 0.1 0 0 -0.1 -0.05 0 5 10 15 0 5 10 15 Output Gap Output Growth 0.1 0.1 0.05 0 0 -0.1 -0.05 -0.2 -0.1 -0.3 -0.15 0 5 10 15 0 5 10 15
Slide 28: Monetary policy active in 70's. No indeterminacy!
Monetary policy active in 70’s. No indeterminacy! What caused the Great Stag‡ation?
Slide 29: Inflation and Output Growth: No Commodity Price Shock
Inflation 12 Data 11 No Commodity Price Shock 10 9 8 7 6 5 4 3 2 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2) Output Growth 8 Data No Commodity Price Shock 6 4 2 0 -2 -4 -6 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2)
Slide 30: Inflation and Output Growth: No Policy Shock
Inflation 12 Data 11 No Policy Shock 10 9 8 7 6 5 4 3 2 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2) Output Growth 8 Data No Policy Shock 6 4 2 0 -2 -4 -6 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2)
Slide 31: Inflation and Output Growth: No Supply-side Shock
Inflation 12 Data 11 No Supply-side Shock 10 9 8 7 6 5 4 3 2 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2) Output Growth 8 Data No Supply-side Shock 6 4 2 0 -2 -4 -6 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2)
Slide 32: Inflation and Output Growth: No Demand Shocks
Inflation 12 Data 11 No Demand Shocks 10 9 8 7 6 5 4 3 2 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2) Output Growth 8 Data No Demand Shocks 6 4 2 0 -2 -4 -6 72(1) 73(1) 74(1) 75(1) 76(1) 77(1) 78(1) 79(2)
Slide 33: What caused the Great Stagflation?
What caused the Great Stag‡ation? 1973-75 episode driven by supply-side shocks Second in‡ation peak fueled by demand shocks
Slide 34: Post-1984 period: What changed?
Post-1984 period: What changed? Why no stag‡ation during 2000’s oil shocks?
Slide 35: Parameter Estimates (6 obs)
Parameter Estimates (6 obs) 1966:I - 1979:II 1984:I - 2008:II ψπ 1.51 3.08 (0.17 ) (0.36 ) ψx 0.03 0.11 (0.03 ) (0.06 ) ψ∆y 0.35 0.69 (0.14 ) (0.15 ) ρR 0.69 0.73 (0.05 ) (0.04 ) ω 0.58 0.14 (0.15 ) (0.05 ) π 1.36 0.95 (0.17 ) (0.09 ) ξ 0.60 0.62 (0.04 ) (0.04 ) γ 0.89 0.46 (0.04 ) (0.12 ) h 0.38 0.
Slide 36: Oil price shock: Pre-79 vs Post-84
0.4 Oil price shock: Pre-79 vs Post-84 Headline Inflation 5 10 Nominal Interest Rate 15 0.3 0.2 0.4 Core Inflation 66:1-79:1I 90% C.l. 84:1-08:1I |] —— 5th pet —— 95th pct Output Growth 10 15 10 15
Slide 37: Commodity Price Shock
Commodity Price Shock Headline Inflation Core Inflation 0.4 0.15 1966Q1-1979Q2 0.3 Post-84 Policy 0.1 Post-84 Wage Rigidity 0.2 0.05 0.1 0 0 -0.1 -0.05 0 5 10 15 0 5 10 15 Nominal Interest Rate GDP 0.15 0.1 0.1 0 0.05 -0.1 0 -0.2 -0.05 -0.3 0 5 10 15 0 5 10 15
Slide 38: Conclusions
Conclusions Monetary policy did not create indeterminacy in 1970’s Trade-o¤ caused by oil price shocks has become less severe Key changes since volative 1970’s: 1. Increased real wage ‡exibility I dampens pass-through of oil price shocks to core in‡ation 2. Central bank focuses more on core in‡ation
Banque Nationale De Belgique — Investor Presentation (September 2017) on 6ix