National Bank of Poland Asymmetry of the exchange rate pass-through: An exercise on the Polish data Jan Przystupa Ewa Wróbel 0th Annual NBP - SNB Seminar June 4, 03 Zurich
PLN/USD fluctuations Band +/-5% Floating exchange rate (April 998) EU accession (May 004)
The exchange rate Exchange rate vs. inflation 00 90 80 70 crow ling peg to a currency basket crowling peg to a currency basket + crawling band floating exchange rate 0 90 70 60 50 50 40 30 0 0 00 fixed exchange rate 99-0 99-0 99-07 993-04 994-0 994-0 995-07 996-04 997-0 997-0 Inflation (left scale) 998-07 999-04 000-0 000-0 00-07 00-04 003-0 NER (right scale) 003-0 004-07 005-04 006-0 006-0 007-07 008-04 009-0 009-0 30 0 90 70 50
MOTIVATION Understanding how prices respond to exchange rate is of key importance for any open economy. An extra stimulus: FAQ: What is an impact of currerent depreciation (appreciation) on consumer prices? Does it depend on the scale of the ex changes or on the phase of the business cycle? What is the pricing policy of the foreign exporting firms towards the Polish market? To answer these questions we propose a complex investigation of the exchange rate channel of the monetary transmission mechanism for an open economy with the floating exchange rate regime.
First: we assessed the level of the exchange rate pass-through. Based on McCarthy (999) where the impact of a sequence of supply, demand and exchange rate shocks on the import, producer and consumer prices is examined. Since it is a popular method we used it for the sake of comparability with other studies and to have an idea about the level of the pass-through effect in Poland. Flowchart of the pass-through effect. Money (M3) Interest rate oil price (Brent, in USD) (supply shock) output gap (demand shock) NEER (30% USD; 70% EUR) CPI PPI Import prices (in domestic currency)
McCarthy s model (999) πm - index of the import transaction prices expressed in the domestic currency; εm - unexpected change of the import price ( π m ) + αε s + αε d + α ε e m m = Et 3 ε π + πw - price index of the sold production of industry (PPI); εw - unexpected change of the production price ( π w ) + β ε s + β ε d + β ε e + β ε m w π + w = Et 3 4 ε πc - consumption price index (CPI); εc - unexpected change of the CPI Coefficients explain a part of the shock assigned to the corresponding variable ( π c ) + γ ε s + γ ε d + γ ε e + γ ε m + γ ε w c π + c = Et 3 4 5 ε
Exchange rate pass-through. Estimation based on the McCarthy s SVAR PT ( z t ) h = z e t, t+ h t, t+ h Pass-through effect: changes of the variable z (import, production, consumption prices) from t to t+h being the response on the exchange rate changes between t and t+h Pass-through effect after Import transaction prices (PM) for Price index of the sold production of industry (PPI) Consumption price index CPI) quarters 4 quarters 8 quarters Est.0 Est. Est.0 Est. Est.0 Est. 0.5 0.46 0.69 0.67 0.79 0.7 0.7 0.9 0.50 0.33 0.59 0.38 0.7 0.0 0.36 0.6 0.4 0.8
Exchange rate pass-through. Estimation based on the McCarthy s SVAR Time decomposition of the pass-through effect. Time decomposition of the pass-through effect for (total P-T=00) Quarter after shock Q 0 Q Q Q 3 Q 4 -Q 8 Import transaction prices (PM) 7 49 5 4 5 Price index of the sold production of industry (PPI) 35 9 0 4 Consumption price index CPI) 0 4 3 7 0
Assessing the level of the exchange rate pass-through to the import prices and pricing to market (PTM) with cointegration technics. Cointergating vector, exchange rate - NEER p = α + α e + α p + α p + α y + ε IMP F H H t 0 t t 3 t 4 t t VECM (cointegration Johansen), - t stat in [ ] unrestricted α - 0.9 [5.9] α.44 [.] α (If significant then PTM exists) 3 0.95 [5.9] α (If significant then PT is full) Chi - square=0.56, p. 0.45 restricted: α = α - 0.8 [7.50] α 0.8 [7.50] α.3 [7.65] 3 restricted: α = restricted: α = α =, 3 α = (Law of one price) Chi - square=9.64, p. 0.008 α =0 (Unit homogeneity) Chi - square=., p. 0.0 restricted: α = α = (Unit coeff. at EX & FP) Chi - square=0., p. 0.006
EX pass-through to the import prices Cointegrating vector, bilateral exchange rate EUR/PLN p = + e + p + p + y + α α α α α ε IMP F H H t 0 t t 3 t 4 t t VECM (cointegration Johansen),t- stat in [ ] Fully modified least squares t - stat in [ ] unrestricted α - 0.7 [.47] -0.68 [4.73] α 0.4 [0.35] 0.74 [.09] α 0.83 [.7] 0.8 [4. 4] 3 restricted: α = α Chi - square=0.03, p. 0.86 Chi - square=0.08, p. 0.9 α - 0.78 [5.8] -0.67 [7.5] α 0.78 [5.8] 0.67 [7.5] α 0.7 [7.] 0.84 [5.9] 3 restricted: α = α = Chi - square=7.4, p. 0.06 Chi - square=.46, p. 0.00 restricted: α = α =, α 3 =0 Chi - square=7.49, p. 0.058 Chi - square=60, p. 0.000 restricted: α = α = Chi - square=6.87, p. 0.03 Chi - square=40.0, p. 0.000
4 4 4 4 IMP F H IMP t β0 β t β, i t i β3, i t i β4, i t i β5, i t i t i= 0 i= 0 i= 0 i= p = + EC + e + p + p + p + v Table A. Dynamic import price equation Variable Coefficient t-stat β 0.00 0.6 0 β -0.464-3.0 β 0.5 3.94,0 β PTM 0.898.8 4,0 Return to equilibrium EX pass-through (short term) Dynamic import price equations Table A. Dynamic import price eq uation: appreciation of the EUR /PLN, usa ble obs.: 4. Variable Coefficient t - stat, p - value in ( ) A β - 0.003-0.6 (0.874) 0 A β - 0.654 -.96 (0.008) A β,0 0.55.58 (0.3) A β 4,0.9.8 (0.09) A t = t { if e 0 0 otherwise Table A 3. Dynamic import price equ ation: depreciation of the EUR /PLN, usable obs.: 0 Variable Coefficient t - stat β - 0.0055-0.39 D 0 D β - 0.5 -.03 D β,0 0.59.07 D β 4,0 0.77. D t = t { if e 0 0 otherwise
Dynamic import price equations 4 4 4 4 IMP F H IMP t β0 β t β, i t i β3, i t i β4, i t i β5, i t i t i= 0 i= 0 i= 0 i= p = + EC + e + p + p + p + v Table A4. Dynamic import price equation: positive EC, usable obs.:. Variable Coefficient t-stat β + 0.0083 0.08 EC 0 EC β + -0.3-0.63 EC β +,0 0.66 4.7 EC β + 4,0 0.003 0.003 Table A5. Dynamic import price equation: negative EC, usable obs.: 3. Variable Coefficient t-stat β -0.09 -.6 EC 0 EC β -0.65 -.0 EC β,0 0.6 0.8 EC β 4,0.0 3.39 EC+ import prices are lower than the equilibrium level determined by exporters prices and domestic prices EC- import prices are higher than the equilibrium level determined by exporters prices and domestic prices
Exchange rate pass-through models based on the Phillips curve π = α Eπ + ( α α ) π + α ( e ) + α y + ε qi qi qi qi qi r qi tk,, k t t+, k, k t, k t 3, k t t where: π stands for inflation (CPI); qi is a variable (i= 4) stands for: i= output gap (y); i= nominal effective exchange rate ( e); i=3 volatility of the nominal effective exchange rate (s); i=4 inflation (π): actual inflation inflation target (π*); k=,; Threshold estimated with SETAR (Self-Exciting Threshold AutoRegresive) model k= for qi > τ (τ = threshold); k= for qi τ k=3,4; Investigate nonreversibility of the linear functions through segmenting the variables. k=3 for qi > qi- k=4 for qi qi- r et is a nominal effective exchange rate (in log) plus foreign inflation (HICP in the euro zone; in log)
The economic interpretation of the threshold model and model based on nonreversibility of the linear functions % 3 0 Threshold models Nonreversible linear models t t0 recession expansion Nonreversible linear models t>t0 EU accession Threshold models - Russian crisis Threshold models - -3 997Q 997Q3 998Q 998Q3 999Q 999Q3 000Q 000Q3 00Q 00Q3 00Q 00Q3 slump 003Q 003Q3 004Q 004Q3 005Q 005Q3 006Q 006Q3 prosperity 007Q 007Q3 008Q 008Q3 output gap threshold business cycle
The economic interpretation of the threshold model and model based on nonreversibility of the linear functions Inflation - target inflation (π - π * ) Early recession = t r a cost t r = a t cos θ θ Late expansion Early expansion Output gap (y) Late recession Two-leaf clover curve
Asymmetry of the exchange rate passthrough to CPI related to: The asymmetry of the exchange rate pass-through to the consumer prices Threshold models (τ = threshold) Nonreversible linear models variable > τ variable τ t > t 0 t t 0 Output gap (y) nominal effective exchange rate ( e) τ = 0.4% 0.9 0.79 τ =.08% 0.065 0.39 0.74 0.09 0.08 0.38 Volatility of the nominal effective exchange rate (s) Inflation (π) Pass-through (general) τ = 4.3% 0.47 0.549 τ = level of official inflation target 0.95 0.0 0.9 0.39 0.4 0.60 0.83
The asymmetry of the exchange rate pass-through to the consumer prices 3 Recession Expansion 0,4 EU accession 0,35 0,3 0 - Russian crisis Average pass-through 0,5 0, 0,5 pass-through effect Max. pass-through 0, - -3 997Q 997Q3 Average pass-through 998Q 998Q3 999Q Min. pass-through 999Q3 000Q 000Q3 00Q 00Q3 00Q 00Q3 Average pass-through 003Q 003Q3 004Q 004Q3 005Q 005Q3 006Q 006Q3 007Q 007Q3 008Q 008Q3 output gap pass-through effect (right scale) proxy of the business cycle This is coherent with the behavior of enterprises in the business cycle, conditioning their investment decisions on expected profits with a maximum in the early expansion and a minimum in the early recession. The enterprises propensity to change prices follows profit expectations. 0,05 0
Exchange rate pass-through and inflation. I q`08 II q'08 III q'08 IV q'08 I q`09 II q'09 III q'09 IV q'09 I q`0 II q'0 III q'0 IV q'0 I q` II q' III q' IV q' I q` II q' III q' IV q' I q`3 Linear PT -0,65-0,684-0,59-0,56 0,84 0,44 0,8-0,05-0,0-0,070 0,056 0,0-0,83-0,35-0,047 0,77,4 0,50-0,06-0,49-0,60 Asym.PT -0,573-0,707-0,783-0,57,707,646,99-0,404-0,64-0,339 0,05 0,065-0,90-0,46-0,067 0,509 0,885 0,356-0,08-0,77-0,7 Figures shown in the table indicate by how many percentage points inflation in a given quarter would be higher (-) or lower (+) than the counterfactual inflation that assumes no exchange rate changes. A year on year impact is calculated as an average impact in the consecutive four quarters.
THANK YOU