- Email: [email protected]
Exchange rate pass-through to domestic producer prices: Evidence from Korean firm-level pricing survey
Accepted Manuscript Exchange rate pass-through to domestic producer prices: Evidence from Korean firm-level pricing survey JaeBin Ahn, Chang-Gui Park PII: DOI: Reference:
S0165-1765(14)00199-2 http://dx.doi.org/10.1016/j.econlet.2014.05.022 ECOLET 6355
To appear in:
Economics Letters
Received date: 28 March 2014 Revised date: 22 May 2014 Accepted date: 26 May 2014 Please cite this article as: Ahn, J., Park, C.-G., Exchange rate pass-through to domestic producer prices: Evidence from Korean firm-level pricing survey. Economics Letters (2014), http://dx.doi.org/10.1016/j.econlet.2014.05.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
*Highlights (for review)
The imported-inputs channel is a major channel of exchange rate pass-through. The firm-level degree of exchange rate pass-through is nonlinear and asymmetric. Nonlinear pass-through stems from industry-level price movements. Asymmetric pass-through is driven by capacity constrained firms.
*Title Page
T itle: E xchange Rate Pass-T hrough to Domestic Producer Prices: E vidence from K orean F irm-L evel Pricing Survey
JaeBin A hn1, International Monetary Fund C hang-G ui Par k 2, Bank of Korea A bstract: This note studies exchange rate pass-through to the prices of domestically produced goods, exploring the firm-level pricing survey conducted by the Bank of Korea. The data reveal the imported inputs channel of, as well as nonlinear and asymmetric, exchange rate-pass-through.
K eywords: exchange rate pass-through, imported-inputs channel, nonlinearity, asymmetry. J E L : E3, F1, F3, F4
1
Corresponding Author: Asia and Pacific Department, International Monetary Fund, 700 19th st NW, Washington, DC 20431, USA, email: [email protected], tel: 1-202-623-5998
2
Economic Research Institute, the Bank of Korea, 39, Namdaemunno 3-Ga, Jung-Gu, Seoul, Republic of Korea, email: [email protected]
*Manuscript Click here to view linked References
1
Introduction
A growing number of central banks use firm-level pricing surveys to broaden understanding of actual price-setting behaviors (e.g., Fabiani et al., 2006; ´ Olafsson et al., 2011). The current note is one outcome of the Bank of Korea’s firm-level pricing survey in 2012, expressly designed to explore firm-level exchange rate pass-through to the price of domestically produced goods. The main strength of this survey data is that the degree of exchange rate passthrough is not a product of the estimation procedures, but rather an observable variable obtained at the firm level under four different scenarios— a large depreciation, small depreciation, large appreciation, and small appreciation— separately, providing an excellent tool to evaluate the within-firm pattern of nonlinear and asymmetric exchange rate pass-through. Among a vast array of studies on exchange rate pass-through, this note builds on the literature with an explicit consideration of the imported-input channel as well as nonlinearity and asymmetry (e.g., Amiti et al., forthcoming; Choi and Cook, 2013; Goldberg and Campa, 2010; Greenway, Kneller, and Zhang, 2010; Fauceglia, Shingal, and Wermelinger, 2012; Pollard and Coughlin, 2004; Powers and Riker, 2013).
2
A Model
The following model, a simple variant of the one discussed in Burstein and Gopinath (forthcoming), provides a conceptual background to understanding changes in the producer price of domestic goods in response to exchange rate movements. Firm i in industry j sets an optimal price as the mark-up over marginal cost: pij = μij + mcij ,
(1)
where each term is expressed in log with a lower case letter. The mark-up, μij , is a decreasing function of the firm’s price relative to the aggregate industrylevel price. Marginal cost is specified in log as:
mcij = mcij qijtot , wj , mij (e) = Aij + κij qijtot + β ij mij (e) + γ ij wj ,
(2)
where qijtot is the total output, and wj is the price of domestic inputs. e denotes the local currency value of foreign currency, which converts the price of imported inputs in foreign currency, pm ij , into the local currency, mij . κij 1
governs the degree of returns to scale in the production technology. β ij and γ ij are the share of imported inputs and that of domestic inputs in total costs, respectively. Aij denotes the constant term. Total differentiating equations (1) and (2) gives the log change in the optimal price as:
Δpij = −Γij (Δpij − Δpj ) + κij Δqijtot + β ij Δpm ij + Δe + γ ij Δwj ,
(3)
∂μ
ij where Γij = − ∂(pij −p ≥ 0 is the elasticity of the mark-up with respect to the j) relative price, decreasing in the relative price. Specifying the demand structure in the domestic market as Δqij = −εij (Δpij − Δpj )+Δqj with price elasticity, > 0, equation (3) is rearranged as: εij = − ∂(p∂q(·) ij −pj )
imported-inputs channel Δpij = Δe
m
β ij Δpij +1 Ψij Δe
indirect effect channel
Δp Δwj 1 j + γ ij + Γij + Φij , (4) Ψij Δe Δe
where Ψij = 1+Γij +Φij with Φij = αij κij εij ≥ 0 being the partial elasticity of marginal cost with respect to the relative price, and αij the share of domestic sales in total output. 1 The impact of imported-inputs channel is proportional to the cost share of imported inputs in production, β ij , while mark-up adjustment (Γij ) and changes in marginal cost (Φij ) dampen the magnitude. The marginal cost effect leads to a unique prediction that exchange rate passthrough for capacity constrained firms will be lower when the local currency appreciates than when it depreciates because capacity constrained firms, by definition, find it more costly to increase output in response to a reduction in input costs induced by currency appreciation, captured by a discrete jump upward in κij for these firms. 2 However, such asymmetry will be less pronounced for exporters who can shift a portion of their export sales to the domestic market without incurring an increase in the total output level. This expression is derived from approximating changes in firm-level total output as changes in domestic sales multiplied by the share of domestic sales in total output tot α Δq ). In addition, the price elasticity of demand at the industry(i.e., Δqij ij ij level is assumed 1, yielding Φij = αij κij (εij − 1). 2 This hypothesis may be traced back to Knetter (1994) and Pollard and Coughlin (2004). 1
2
The model also provides an alternative explanation for nonlinear exchange rate pass-through beyond the menu cost model. As long as industry-level prices do not react to small changes in the exchange rate, the indirect effect channel will be in play only when the exchange rate movement is large enough to affect industry-level prices, and it will be amplified as the mark-up elasticity or the marginal cost elasticity is higher.
3
Empirical Evidence
3.1 Data This note explores the firm-level pricing survey data collected by the Bank of Korea in 2012. The key question of particular interest to the paper is how much a firm reflects changes in the exchange rate into the domestic price of the main product, and each respondent chooses a corresponding answer among ”fully”, ”partially”, and ”not at all” under four different scenarios— a large depreciation, small depreciation, large appreciation, and small appreciation. The dataset is based on responses from 693 firms out of 1,678 manufacturing firms selected in the original sample. 3
Table 1 The Degree of Exchange Rate Pass-Through for Importers and Non-importers The sample firms are carefully selected from a population of around 300,000 incorporated firms in Korea, by applying Lavallee and Hidiroglou’s algorithm. A detailed description of the sampling procedure is provided in Park and Song (2013, in Korean), an English translated version of which is available on request. 3
3
summarizes exchange rate pass-through for importers and nonimporters separately. Unsurprisingly, importers tend to change the domestic price more than non-importers in response to changes in the exchange rate. Panel A shows that, when local currency depreciates, 83 percent of nonimporters do not change the domestic price, whereas only 35 percent of importers do not change the price. Similar patterns are found for other cases (Panel B-D). A quick comparison across panels also hints at potential nonlinear and asymmetric exchange rate pass-through. shows that, in response to a large depreciation, firms that use imported inputs more intensively tend to have a higher exchange rate pass-through (Panel A, column 1). The estimated coefficient implies that, from the average marginal effect analysis, a producer with a higher intensity of imported inputs in productive inputs is, compared to a producer with a lower intensity of imported inputs, 17 percentage points more likely to fully reflect changes in adjusting the domestic price of its main product. Adding other firm characteristics lowers the size of the coefficient estimate on imported inputs intensity slightly, but the difference is not statistically significant (Panel A, column 2). A positive coefficient estimate on export intensity, among others, is consistent with the model that exchange rate pass-through decreases in the share of domestic sales in total output. Adding industry fixed effects does not Industry classifications are broadly comparable with the three-digit North American Industry Classification System (NAICS). 5 Given the categorical nature of discrete responses in the questionnaire, a threshold value of 25 percent is arbitrarily chosen to distinguish between high imported inputs intensity firms and low imported inputs intensity firms. All the results are robust at different threshold values. 4, exchange rate pass-though is higher when changes in exchange rate change are large, and such nonlinearity is amplified for firms with a higher share of imported inputs (columns 1 and 2). Similarly, exchange rate pass-through is higher when currency depreciates, but such asymmetric passthrough is not related to the use of imported inputs (columns 3 and 4). The 6
S0165-1765(14)00199-2 http://dx.doi.org/10.1016/j.econlet.2014.05.022 ECOLET 6355
To appear in:
Economics Letters
Received date: 28 March 2014 Revised date: 22 May 2014 Accepted date: 26 May 2014 Please cite this article as: Ahn, J., Park, C.-G., Exchange rate pass-through to domestic producer prices: Evidence from Korean firm-level pricing survey. Economics Letters (2014), http://dx.doi.org/10.1016/j.econlet.2014.05.022 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
*Highlights (for review)
The imported-inputs channel is a major channel of exchange rate pass-through. The firm-level degree of exchange rate pass-through is nonlinear and asymmetric. Nonlinear pass-through stems from industry-level price movements. Asymmetric pass-through is driven by capacity constrained firms.
*Title Page
T itle: E xchange Rate Pass-T hrough to Domestic Producer Prices: E vidence from K orean F irm-L evel Pricing Survey
JaeBin A hn1, International Monetary Fund C hang-G ui Par k 2, Bank of Korea A bstract: This note studies exchange rate pass-through to the prices of domestically produced goods, exploring the firm-level pricing survey conducted by the Bank of Korea. The data reveal the imported inputs channel of, as well as nonlinear and asymmetric, exchange rate-pass-through.
K eywords: exchange rate pass-through, imported-inputs channel, nonlinearity, asymmetry. J E L : E3, F1, F3, F4
1
Corresponding Author: Asia and Pacific Department, International Monetary Fund, 700 19th st NW, Washington, DC 20431, USA, email: [email protected], tel: 1-202-623-5998
2
Economic Research Institute, the Bank of Korea, 39, Namdaemunno 3-Ga, Jung-Gu, Seoul, Republic of Korea, email: [email protected]
*Manuscript Click here to view linked References
1
Introduction
A growing number of central banks use firm-level pricing surveys to broaden understanding of actual price-setting behaviors (e.g., Fabiani et al., 2006; ´ Olafsson et al., 2011). The current note is one outcome of the Bank of Korea’s firm-level pricing survey in 2012, expressly designed to explore firm-level exchange rate pass-through to the price of domestically produced goods. The main strength of this survey data is that the degree of exchange rate passthrough is not a product of the estimation procedures, but rather an observable variable obtained at the firm level under four different scenarios— a large depreciation, small depreciation, large appreciation, and small appreciation— separately, providing an excellent tool to evaluate the within-firm pattern of nonlinear and asymmetric exchange rate pass-through. Among a vast array of studies on exchange rate pass-through, this note builds on the literature with an explicit consideration of the imported-input channel as well as nonlinearity and asymmetry (e.g., Amiti et al., forthcoming; Choi and Cook, 2013; Goldberg and Campa, 2010; Greenway, Kneller, and Zhang, 2010; Fauceglia, Shingal, and Wermelinger, 2012; Pollard and Coughlin, 2004; Powers and Riker, 2013).
2
A Model
The following model, a simple variant of the one discussed in Burstein and Gopinath (forthcoming), provides a conceptual background to understanding changes in the producer price of domestic goods in response to exchange rate movements. Firm i in industry j sets an optimal price as the mark-up over marginal cost: pij = μij + mcij ,
(1)
where each term is expressed in log with a lower case letter. The mark-up, μij , is a decreasing function of the firm’s price relative to the aggregate industrylevel price. Marginal cost is specified in log as:
mcij = mcij qijtot , wj , mij (e) = Aij + κij qijtot + β ij mij (e) + γ ij wj ,
(2)
where qijtot is the total output, and wj is the price of domestic inputs. e denotes the local currency value of foreign currency, which converts the price of imported inputs in foreign currency, pm ij , into the local currency, mij . κij 1
governs the degree of returns to scale in the production technology. β ij and γ ij are the share of imported inputs and that of domestic inputs in total costs, respectively. Aij denotes the constant term. Total differentiating equations (1) and (2) gives the log change in the optimal price as:
Δpij = −Γij (Δpij − Δpj ) + κij Δqijtot + β ij Δpm ij + Δe + γ ij Δwj ,
(3)
∂μ
ij where Γij = − ∂(pij −p ≥ 0 is the elasticity of the mark-up with respect to the j) relative price, decreasing in the relative price. Specifying the demand structure in the domestic market as Δqij = −εij (Δpij − Δpj )+Δqj with price elasticity, > 0, equation (3) is rearranged as: εij = − ∂(p∂q(·) ij −pj )
imported-inputs channel Δpij = Δe
m
β ij Δpij +1 Ψij Δe
indirect effect channel
Δp Δwj 1 j + γ ij + Γij + Φij , (4) Ψij Δe Δe
where Ψij = 1+Γij +Φij with Φij = αij κij εij ≥ 0 being the partial elasticity of marginal cost with respect to the relative price, and αij the share of domestic sales in total output. 1 The impact of imported-inputs channel is proportional to the cost share of imported inputs in production, β ij , while mark-up adjustment (Γij ) and changes in marginal cost (Φij ) dampen the magnitude. The marginal cost effect leads to a unique prediction that exchange rate passthrough for capacity constrained firms will be lower when the local currency appreciates than when it depreciates because capacity constrained firms, by definition, find it more costly to increase output in response to a reduction in input costs induced by currency appreciation, captured by a discrete jump upward in κij for these firms. 2 However, such asymmetry will be less pronounced for exporters who can shift a portion of their export sales to the domestic market without incurring an increase in the total output level. This expression is derived from approximating changes in firm-level total output as changes in domestic sales multiplied by the share of domestic sales in total output tot α Δq ). In addition, the price elasticity of demand at the industry(i.e., Δqij ij ij level is assumed 1, yielding Φij = αij κij (εij − 1). 2 This hypothesis may be traced back to Knetter (1994) and Pollard and Coughlin (2004). 1
2
The model also provides an alternative explanation for nonlinear exchange rate pass-through beyond the menu cost model. As long as industry-level prices do not react to small changes in the exchange rate, the indirect effect channel will be in play only when the exchange rate movement is large enough to affect industry-level prices, and it will be amplified as the mark-up elasticity or the marginal cost elasticity is higher.
3
Empirical Evidence
3.1 Data This note explores the firm-level pricing survey data collected by the Bank of Korea in 2012. The key question of particular interest to the paper is how much a firm reflects changes in the exchange rate into the domestic price of the main product, and each respondent chooses a corresponding answer among ”fully”, ”partially”, and ”not at all” under four different scenarios— a large depreciation, small depreciation, large appreciation, and small appreciation. The dataset is based on responses from 693 firms out of 1,678 manufacturing firms selected in the original sample. 3
Table 1 The Degree of Exchange Rate Pass-Through for Importers and Non-importers The sample firms are carefully selected from a population of around 300,000 incorporated firms in Korea, by applying Lavallee and Hidiroglou’s algorithm. A detailed description of the sampling procedure is provided in Park and Song (2013, in Korean), an English translated version of which is available on request. 3
3