Banque nationale de Belgique SA · EURONEXT:BNB
Banque nationale de Belgique SA — Investor Presentation (October 2018)
34 slides from Banque nationale de Belgique SA's investor presentation, October 2018
Slide 1: Dominant Currencies
Dominant Currencies How firms choose currency invoicing and why it matters Mary Amiti Oleg Itskhoki Jozef Konings NY FRB Princeton Leuven and Liverpool Preliminary and Incomplete National Bank of Belgium October 2018 1 / 19
Slide 2: Motivation
Motivation • Currency use in international trade is central for 1 international transmission of shocks 2 optimal monetary (exchange rate) policy • Accumulated evidence in favor of endogenous currency choice — active firm-level decision — slow changes in the roles of individual currencies over time •
Slide 3: Main Findings
Main Findings 1 Little role for PCP in imports / DCP in exports (ex-EU trade) 2 Instead, two dominant currencies: — regional Euro: dominates by count of (smaller) transactions — global US dollar: dominates by value of trade 2 / 19
Slide 4: Main Findings
Main Findings 1 Little role for PCP in imports / DCP in exports (ex-EU trade) 2 Instead, two dominant currencies: — regional Euro: dominates by count of (smaller) transactions — global US dollar: dominates by value of trade 3 Substantial variation in currency use within country×sectors. Firm-level c
Slide 5: Related Literature
Related Literature 1 Theory of currency choice and ERPT — Engel (2006) — Gopinath, Itskhoki and Rigobon (2010) — Mukhin (2018) 2 Firm-level analysis of exchange rate pass-through — Berman, Martin and Mayer (2012) — Amiti, Itskhoki and Konings (2014, 2018) 3 Empirical analysis of currency choice and
Slide 6: Theoretical Framework
THEORETICAL FRAMEWORK 4 / 19
Slide 7: Model Environment
Model Environment • Consider a problem of a Belgian exporter i serving a given industry s in foreign country k • Denote with e = ek the log euro exchange rate with country k; an increase in e corresponds to a depreciation of the euro • The desired export price of the firm in foreign currency is: p̃i
Slide 8: Desired Pass-through
Desired Pass-through • Desired markup: p̃i = µ̃i + mci • Price change decomposition (following AIK 2018): 1 Γi dp̃i = dmci + (dzk∗ + de) + εi 1 + Γi 1 + Γi — zk∗ is the competitor price index in destination market k — Γi ≡ −∂ µ̃i /∂pi is the elasticity of the desired markup, increasing in the firm’s
Slide 9: Desired Pass-through
Desired Pass-through • Desired markup: p̃i = µ̃i + mci • Price change decomposition (following AIK 2018): 1 Γi dp̃i = dmci + (dzk∗ + de) + εi 1 + Γi 1 + Γi — zk∗ is the competitor price index in destination market k — Γi ≡ −∂ µ̃i /∂pi is the elasticity of the desired markup, increasing in the firm’s
Slide 10: Desired Pass-through
Desired Pass-through • Desired markup: p̃i = µ̃i + mci • Price change decomposition (following AIK 2018): 1 Γi dp̃i = dmci + (dzk∗ + de) + εi 1 + Γi 1 + Γi — zk∗ is the competitor price index in destination market k — Γi ≡ −∂ µ̃i /∂pi is the elasticity of the desired markup, increasing in the firm’s
Slide 11: Prices Stickiness
Prices Stickiness • Prices are set before the uncertainty in Ω is realized Specifically, in currency ` the preset price is: p̄i` = arg maxp̄` E Πi (p̄i` + e` − e|Ω) i • Ex post, firm can reset price with a Calvo probability θ (θ is increasing with time interval) — Gopinath, Itskhoki & Rigobon (2010)
Slide 12: Prices Stickiness
Prices Stickiness • Prices are set before the uncertainty in Ω is realized Specifically, in currency ` the preset price is: p̄i` = arg maxp̄` E Πi (p̄i` + e` − e|Ω) i • Ex post, firm can reset price with a Calvo probability θ (θ is increasing with time interval) — Gopinath, Itskhoki & Rigobon (2010)
Slide 13: Currency Choice
Currency Choice • Optimal currency choice ` for the preset prices solves: n o ` = arg max maxp̄` E Πi (p̄i` + e` − e|Ω) ` i 7 / 19
Slide 14: Currency Choice
Currency Choice • Optimal currency choice ` for the preset prices solves: n o ` = arg max maxp̄` E Πi (p̄i` + e` − e|Ω) ` i • Lemma 2 (Currency Choice) The optimal currency choice is second-order equivalent to: n o n o ` = arg min var p̃i` = arg min var p̃i∗ + e − e` . ` ` — the optimal currency ` e
Slide 15: Currency Choice
Currency Choice • Optimal currency choice ` for the preset prices solves: n o ` = arg max maxp̄` E Πi (p̄i` + e` − e|Ω) ` i • Lemma 2 (Currency Choice) The optimal currency choice is second-order equivalent to: n o n o ` = arg min var p̃i` = arg min var p̃i∗ + e − e` . ` ` — the optimal currency ` e
Slide 16: Currency Choice
Currency Choice • Optimal currency choice ` for the preset prices solves: n o ` = arg max maxp̄` E Πi (p̄i` + e` − e|Ω) ` i • Lemma 2 (Currency Choice) The optimal currency choice is second-order equivalent to: n o n o ` = arg min var p̃i` = arg min var p̃i∗ + e − e` . ` ` — the optimal currency ` e
Slide 17: Estimating Equation I
Estimating Equation I Currency choice • Currency choice: ( 0 (PCP), if ψ̃i ≈ 0 ιi = 1 (LCP or DCP), otherwise, i.e. if ψ̃i 0 Γi and recall desired pass-through ψ̃i = ϕi + 1+Γ i [(1 − Ψ∗k ) − ϕi ] 8 / 19
Slide 18: Estimating Equation I
Estimating Equation I Currency choice • Currency choice: ( 0 (PCP), if ψ̃i ≈ 0 ιi = 1 (LCP or DCP), otherwise, i.e. if ψ̃i 0 Γi and recall desired pass-through ψ̃i = ϕi + 1+Γ i [(1 − Ψ∗k ) − ϕi ] • Estimating equation: P{ιi = 1} = αsk + βϕi + γSi + δῑ−i (∗) — Euro (PCP) less likely for import-inten
Slide 19: Estimating Equation II
Estimating Equation II Exchange rate pass-through • Exchange rate pass-through: ∂pi ψi = E = (1 − θ) · ιi χi +θ · ψ̃i , ∂e |{z} |{z} sticky flex where χi = 1 for LCP and χi ∈ [0, 1] for DCP 9 / 19
Slide 20: Estimating Equation II
Estimating Equation II Exchange rate pass-through • Exchange rate pass-through: ∂pi ψi = E = (1 − θ) · ιi χi +θ · ψ̃i , ∂e |{z} |{z} sticky flex where χi = 1 for LCP and χi ∈ [0, 1] for DCP • Estimating equation for ERPT (augmenting AIK 2014): h i E∆pi ≈ ask + bϕi + cSi + dιi ∆e (∗∗) — ERPT into for
Slide 21: Data
DATA 10 / 19
Slide 22: Dataset
Dataset • We merge 3 micro-level datasets on Belgian firms: 1 NBB and Customs: New data on currency choice of Belgian firms at the firm-product-country-month level for both imports and exports from February 2017 to March 2018 — CN 8-digit level (over 10,000 products) — Only extra-EU trade. We focus
Slide 23: Key Variables
Key Variables • Currency use: ιikt = 0 for Euro and ιikt = 1 for non-Euro, by firm-industry-destination-time • Export price change: in euros by firm-product-destination-time Export Valueikt ∆pikt = ∆ log Export Quantityikt • Import intensity: at the firm-year level Total non-EU import valueit ϕit ≡
Slide 24: Stylized Facts
STYLIZED FACTS 12 / 19
Slide 25: Currency Use in Trade
Currency Use in Trade Outside EU Exports Imports Count Value share Count Value share share All Diff Non-diff share All Diff Non-diff Euro 0.691 0.388 0.425 0.324 0.418 0.414 0.512 0.283 Dollar 0.207 0.475 0.374 0.645 0.495 0.501 0.354 0.701 Other 0.102 0.137 0.201 0.031 0.087 0.084 0.134 0.017 1 Eur
Slide 26: Euro and US Dollar
Euro and US Dollar Imports vs exports (a) Euro (b) Dollar 1 1 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 13 / 19
Slide 27: Dominant Currencies
Dominant Currencies (a) Imports (b) Exports 1 1 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 14 / 19
Slide 28: Variance Decomposition
Variance Decomposition Currency choice in exports Dep.var.: ιikt (1) (2) (3) (4) (5) (6) Adjusted R 2 0.126 0.386 0.461 0.539 0.563 0.681 Fixed effects: country X X X X HS-4 industry X HS-8 industry X country×HS-4 X X firm X X — Value-weighted projections of ιikt on different sets of fixed effects •
Slide 29: Empirical Results
EMPIRICAL RESULTS 16 / 19
Slide 30: Currency Choice
Currency Choice Linear probability regressions Dep. var.: P{ιikt = 1} (1) (2) (3) (4) (5) (6) (7) ϕi 0.403∗∗ 0.275∗∗ 0.239∗ (0.187) (0.130) (0.140) ϕEi 0.222 0.161 0.110 0.186 (0.282) (0.193) (0.206) (0.280) ϕX i 0.637∗∗∗ 0.419∗∗∗ 0.398∗∗∗ 0.565∗∗∗ (0.165) (0.130) (0.142) (0.171) log Empli 0.098∗∗∗
Slide 31: Exchange rate pass-through
Exchange rate pass-through Dep. var.: ∆pikt (1) (2) (3) (4) (5) ∆ekt 0.002 −0.044 −0.048 0.001 −0.048 (0.031) (0.030) (0.032) (0.032) (0.032) ∆ekt · ϕi 0.609∗∗∗ 0.366∗∗ 0.350∗∗ (0.187) (0.163) (0.161) ∆ekt · ϕEi 0.332 0.144 (0.266) (0.274) ∆ekt · ϕX i 0.846∗∗∗ 0.534∗∗∗ (0.253) (0.192) ∆ekt · Sik 0.1
Slide 32: Dynamics of ERPT
Dynamics of ERPT by currency of invoicing 1.2 LCP 1 DCP PCP 0.8 0.6 0.4 0.2 0 -0.2 -0.4 4 6 8 10 12 Regression horizon, months — Value-weighted projection coefficients of ∆pikt on ∆ekt at different time horizons by bins of currency choice 18 / 19
Slide 33: Summary
SUMMARY 19 / 19
Slide 34: Summary and Future Work
Summary and Future Work • Two dominant currencies: global US dollar and regional Euro • Currency choice is shaped by firm-level characteristics, which determine desired ERPT • Currency choice feeds back into the dynamics of ERPT, and still matters at an annual horizon • There are strategic complemen
Banque Nationale De Belgique — Investor Presentation (October 2018) on 6ix