Dollar Pricing Redux

 

BIS Working Papers

No 653

 

“Dollar pricing redux”

by Camila Casas, Federico Díez, Gita Gopinath, Pierre-Olivier Gourinchas

 

Monetary and Economic Department

August 2017

 

JEL classification: F1, F2, F3, F4

Keywords: dominant currency, terms of trade, pass-through, monetary policy

 

BIS Working Papers are written by members of the Monetary and Economic Department of the Bank for International Settlements, and from time to time by other economists, and are published by the Bank. The papers are on subjects of topical interest and are technical in character. The views expressed in them are those of their authors and not necessarily the views of the BIS.

This publication is available on the BIS website (www.bis.org).

 

© Bank for International Settlements 2017. All rights reserved. Brief excerpts may be reproduced or translated provided the source is stated.

ISSN 1020-0959 (print)

ISSN 1682-7678 (online)

 

Foreword

The 15th BIS Annual Conference took place in Lucerne, Switzerland, on 24 June 2016. The event brought together a distinguished group of central bank Governors, leading academics and former public officials to exchange views on the topic “Long-term issues for central banks”. The papers presented at the conference and the discussants’ comments are released as BIS Working Papers 653 to 656.

BIS Papers no 92 contains the opening address by Jaime Caruana (General Manager, BIS) and remarks by Kevin Warsh (Hoover Institution and Stanford Graduate School of Business)

 

Dollar Pricing Redux 

Camila Casas Banco de la Rep´ublica

Federico D´ıez Federal Reserve Bank of Boston

Gita Gopinath Harvard University

Pierre-Olivier Gourinchas University of California at Berkeley

(This Version: August 10, 2016)

 

Abstract

A country’s exchange rate is at the center of economic and political debates on currency wars and trade competitiveness. The real consequences of exchange rate fluctuations depend critically on how firms set prices in international markets. Recent empirical evidence has challenged the dominant ‘producer currency’ pricing and ‘local currency’ pricing paradigms in the literature. In this paper we propose a new paradigm, consistent with the empirical evidence and characterized by three features: pricing in dollars, strategic complementarity in pricing and imported inputs in production. We call this the ‘dollar pricing’ paradigm and contrast its theoretical predictions with prior approaches in a general equilibrium New Keynesian model. We then employ novel data for Colombia to evaluate the implications of exchange rate fluctuations associated with commodity price shocks and show that the findings strongly support the dollar pricing paradigm.


We thank Richard Baldwin and Philip Lane for numerous useful suggestions. The views expressed in this paper are those of the authors and do not indicate concurrence by other members of the research staff or principals of the Board of Governors, the Federal Reserve Bank of Boston, or the Federal Reserve System. The views expressed in the paper do not represent those of the Banco de la Rep´ublica or its Board of Directors. All remaining errors are our own.

 

1 Introduction

A country’s nominal exchange rate is at the center of fierce economic and political debates on spillovers, currency wars and trade competitiveness. This is because in the presence of price rigidities nominal exchange rate fluctuations are associated with fluctuations in relative prices and therefore have consequences for real variables such as the trade balance, consumption and national output.

It is well known that the implications of a nominal exchange rate movement for real variables depend critically on the currency in which prices are rigid. The first genera­tion of models and leading paradigm in international macroeconomics assumes that prices are sticky in the currency of the producing country, so called ‘producer currency pricing’ (PCP). In that paradigm, the law of one price holds and a depreciation reduces the price of exports relative to imports, improving competitiveness. This paradigm is developed in the seminal contributions of Mundell (1963) and Fleming (1962), Svensson and van Wi- jnbergen (1989) and Obstfeld and Rogoff (1995). A second generation pricing paradigm grew out of the pervasive evidence that the law of one price fails to hold. In the original works of Betts and Devereux (2000) and Devereux and Engel (2003) prices are instead assumed to be sticky in the currency of the market in which they are sold, so called ‘local currency pricing’ (LCP). In this case a depreciation leads to an appreciation of the terms of trade, raising the prices of exports relative to imports. Both these paradigms, which assume symmetry in terms of price setting, have been extensively studied in the literature with regards to their predictions for international spillovers and optimal monetary policy. A survey of this research is contained in the latest handbook chapter by Corsetti et al. (2010).

Recent empirical work using granular data on prices in international trade questions the validity of both paradigms. Firstly, there is very little evidence that the best description of pricing in international markets is either PCP or LCP. On the contrary, the vast majority of trade is invoiced in very few currencies, with the U.S. dollar having an outsized role. This is documented in Goldberg and Tille (2008) and Gopinath (2015). Prices are also found to be rigid for significant durations in their currency of invoicing, for countries for which this data are available, as documented by Gopinath and Rigobon (2008), Fitzgerald and Haller (2012).

Secondly, exporters price in markets characterized by strategic complementarities in pricing that give rise to variations in desired mark-ups[1], and use imported inputs to pro­duce.[2] These features help rationalize why exporters coalesce on few currencies (Gopinath et al. (2010)). The work horse general equilibrium open economy models set aside these features by assuming that labor is the only input in production and demand is CES giving rise to constant desired price mark-ups over marginal cost.

In this paper we do the following: First, consistent with the empirical evidence, we build a model of a ‘dollar pricing’ paradigm (DP) by combining the features of dollar price rigidity, strategic complementarities in pricing, and production with imported inputs, in a general equilibrium new Keynesian model. We contrast the predictions of this paradigm with those of LCP and PCP. Second, we employ the universe of customs data for Colombia alongside firm production information to compare the model’s predictions for the response of exchange rates, import and export prices and quantities to a commodity price shock, to those in the data.

We model a small open economy commodity exporter[3] that trades goods and assets with the rest of the world. There is a manufacturing sector where firms have pricing power in world markets and a homogenous goods (commodities) sector where the economy is a price taker. The demand for the output of the manufacturing sector is modeled using a Kimball aggregator that gives rise to strategic complementarities in pricing and variable mark-ups. The manufacturing sector output is produced using labor and imported inputs. Wages adjust sluggishly a la Calvo and are set in the home currency.

We first demonstrate that when prices and wages are fully flexible, the greater is the strategic complementarity in pricing, and the higher the share of imported inputs in pro­duction, the lower is the pass-through of real exchange rate changes (relative price of non­traded goods) into prices of manufactured goods, and consequently, lower is the sensitivity of quantities exported to real exchange rate changes. The intuition for this is as follows: with more strategic complementarities, firms prefer to adjust their markup in response to a cost shock of a given size. The higher is the share of imported intermediates the lower is the effect of the price of non-traded goods on marginal costs, and therefore the smaller the cost shock. Both forces mute the response of the price of manufactured goods and therefore the response of exports to real exchange rate changes.

Further, the greater is the strategic complementarity in pricing, higher is the volatility of the real exchange rate and consumption, while at the same time lower is the volatility of the manufacturing sector terms of trade to commodity price shocks. The intuition for this is as follows: Suppose that the commodity exporting country experiences a negative shock to commodity prices. To satisfy the external resource constraint the country needs to reduce its consumption of traded goods and use of imported intermediate inputs. The extent to which it needs to do this, and consequently the extent of real depreciation required, depends inversely on its ability to raise export income through higher exports from the manufacturing sector that in turn benefits from lower marginal costs following the real depreciation. Because the pass-through from the real depreciation to export prices and hence to export quantities is lower the greater the strategic complementarity in pricing it calls for a greater depreciation of the real exchange rate.

We then contrast the dynamics of the model under the three cases of DP, PCP, and LCP. We focus on the responses of the manufacturing terms of trade, the price and quantity of manufacturing exports and imports, and the mark-ups of manufacturing firms. In response to a negative commodity price shock the terms of trade mildly depreciates in the case of DP, strongly depreciates in the case of PCP and strongly appreciates in the case of LCP. When we compare the pass-through’s of exchange rate changes into export and import prices expressed in home currency across pricing regimes we obtain the following contrasting results: In the case of DP both pass-through’s are quantitatively close to one and decline slowly over time. In the case of PCP, the pass-throughs initially diverge sharply, with the pass-through for export prices close to zero and the pass-through for import prices close to one. Further, the pass-through into home currency export prices increase over time, in contrast with the DP case. In the case of LCP there is a similar divergence in the pass-through rates as in the case of PCP, with the difference that it is pass-through into export prices which start high and close to one, while pass-through into import prices start low and close to zero.

Consistent with the pass-through estimates for prices, in the case of DP the pass­through into export quantities is small with a gradual increase over time, while there is a sharp decline in imports quantities. That is, exchange rate depreciations in the case of DP result in a larger response in imports than in exports. In the case of PCP, there is a sharp increase in exports and a small increase in imports initially with a gradual decrease over time. In the case of LCP both exports and imports increase somewhat over time. In the case of mark-ups, following a decline in commodity prices, there is a sharp increase in mark-ups in the case of LCP, a sharp decline in mark-ups in the case of PCP and a moderate increase in mark-ups in the case of DP.

We then proceed to analyze data for Colombia, an oil-exporting country that is repre­sentative of emerging markets in that it relies heavily on dollar invoicing. We find that the manufacturing terms of trade has low sensitivity to the exchange rate. A 1% depreciation of the exchange rate is associated with a 0.3% depreciation of the manufacturing terms of trade. The pass-through into import and export peso prices starts out close to 1 for import prices and to 0.8 for export prices and then declines some over time. While there is a significant drop in quantities imported, the impact on exports is insignificantly different from zero. Overall we find that the data strongly supports the DP paradigm.

Our paper is related to a small literature that examines the consequences of dollar pricing. These include Corsetti and Pesenti (2005), Goldberg and Tille (2008), Goldberg and Tille (2009) and Devereux et al. (2007). Unlike the models in these papers, which are static with one period ahead price stickiness, and assume constant desired mark-ups and production functions that use only labor, we combine dynamic pricing, variable mark-ups and imported inputs use in production, all of which are important ingredients to match the facts on pricing in international trade.

Burstein and Gopinath (2014a) survey the evidence on variable mark-ups. 

2 The fact that most exporters are also importers is now well documented in the literature by Bernard et al. (2009), Kugler and Verhoogen (2009), Manova and Zhang (2009) among others. This is also re­flected in the fact that value added exports are significantly lower than gross exports, particularly for manufacturing, as documented in the works of Johnson (2014) and Johnson and Noguera (2012)

3 The advantage of focusing on a commodity exporter are twofold. First, commodity prices can arguably be considered exogenous for most commodity exporters. Second, as is well known from Chen et al. (2010), commodity currencies are closely tied to commodity prices.

2 Invoicing Currency

The volume of global merchandize trade has grown tremendously over the last several decades, and the vast majority of this trade is denominated in very few currencies. Figure 1 taken from Gopinath (2015) uses invoicing data for 55% of world imports, and 57% of world exports to demonstrate this fact.[4] Figure 1(a) represents the share of imports that are imported from the US, Eurozone, or the rest of the world; versus the share of imports that are invoiced in dollars, euros, or other currencies. Figure 1(b) does the same for exports. If the world were best described as PCP we should observe that import trade shares and invoicing shares line up with each other. That is, the share of dollar invoicing in world imports should be equal to the share of world imports originating from the U.S.[5] If the world was LCP then we should expect to see that the bar for invoicing shares is dominated by “other currencies,” which is not what we see in the data. In the data, the dollar share as an invoicing currency is estimated to be around 4.7 times its share in (the sample of) world imports. In the case of the euro, its share is more closely aligned at 1.2, however this is because trade within the euro area is also included. If the sample is restricted to only the trade of euro area countries with non-euro area countries, as we document later, this ratio falls strictly below one. The invoicing facts for exports, Figure 1(b), resemble those for imports. The dollars share as an invoicing currency is estimated to be around 3.1 times its share in (the sample of) world exports, while for the euro it is 1.2 times.

Tables 1 and 2 report, for a list of 45 countries, the share of imports and exports that are invoiced in dollars, in euros, and in the country’s own currency as well as the country’s share of imports (exports) coming from (going to) the United States. From the tables it is apparent that the dollar almost always has the largest invoicing share. The euro also has large shares but only in the case of European countries (both, members and non-members of the monetary union). Additionally, only a handful of countries have sizable shares of trade invoiced in their own currency, mostly international reserve currencies; however, these shares are significantly below the share of the dollar for the US. Finally, note that the dollar share column is greater than the US trade share column for every single country—that is, regardless of how much a country trades with the US, the share of its exports and imports invoiced in dollars is greater than the actual trade with the US.[6]

These facts point to a world where a disproportionate share of trade is invoiced in very few currencies, and the dollar has an outsized role. The dominant benchmarks of LCP and PCP would suggest a world of symmetry in invoicing, with all currencies shares lined up more closely with their country’s shares in world trade. Importantly, as the share of world trade attributable to emerging markets has grown over the last fifteen years to constitute 33% of world exports and 37% of world imports (Figure 2), the share of invoicing in dollars has gained greater prominence.[7] To explore the implications of this, in the next section we model a world where a single currency, the dollar, is used in international trade and compare the predictions to those of LCP and PCP.

4 Invoicing facts have also been reported by Goldberg (2013), Goldberg and Tille (2009) and Ito and Chinn (2013).

5 The U.S. is excluded from the sample because of course there is no “U.S.” trade counter-party, which would only serve to artificially increase the non-US trade and the dollar invoicing share.

6 In the case of countries in the euro area, the US trade share is computed considering only the trade that takes place with countries outside of the monetary union.

7 We follow the IMF classification to define emerging markets.

3 Model
 

In this section we model a three sector small open economy that trades goods and assets with the rest of the world. There is a manufacturing sector where firms have pricing power in world markets and a homogenous goods sector where the economy is a price taker. There is also a non-traded sector that is modeled as an endowment. The single nominal exchange rate denoted Et is expressed as home currency per unit of foreign currency. Henceforth we shall refer to the home currency as peso and the foreign currency as dollar. We compare three pricing scenarios: producer currency pricing, local currency pricing, and dollar-pricing.

3.1 Households

The small open economy is populated with a continuum of symmetric households. House­holds are indexed by h E [0,1], but we often omit the index h to simplify exposition. In each period household’s consume a bundle of imported traded goods CT, and non-traded goods CN. Each household also sets a wage rate Wt(h) and supplies labor Nt(h) in order to satisfy demand at this wage rate. They own all domestic firms and profits are rebated lump sum to them. The per-period utility function is given by,

where Nt represents labor supply. The consumption aggregator C is Cobb-Douglas over the traded and non-traded good, C = CTC^-7. Households solve

 

subject to the per-period budget constraint, expressed in home currency,

where Pt = (1-1T)i-7 (PT,t)YPN— is the price of the home consumption bundle and

PtCt = PT,tCT + PN,tCN,t measures household consumption expenditures. nMt and nO,t represents profits generated in the manufacturing and homogenous goods sector, respec­
tively. Households are also endowed each period with non-traded goods YN whose price is 
PN,t.

Households trade a riskless bond denominated in dollars with the rest of the world, that pays out a gross interest rate Rf. As is common in the literature we assume the foreign gross interest rate evolves according to,

 

where ф > 0. Bf is steady state debt denominated in foreign currency. Households also have access to a full set of domestic state contingent securities (in pesos) that they can trade among themselves and that are in zero net supply. Denoting S the set of possible states of the world, Qf(s) is the period-t price of the security that pays one pesos in period t + 1 and state s Є S.

Intra-temporal optimality requires that the allocation of consumption across traded and non-traded goods satisfy,

 

Inter-temporal optimality conditions for dollar bonds and peso bonds are given by,

 

where is the inverse of the price of a peso bond at time t that delivers one peso for sure in every state of the world in period t + 1.

Households are subject to a Calvo friction when setting wages in pesos: in any given period, they may adjust their wage with probability 1 — 9w, and maintain the previous- period nominal wage otherwise. The optimality condition for wage setting is given by:

 

where 0t, s = fts-t C?f:t-lc pt is the stochastic discount factor between t and s used to discount profits and Wt(h) is the optimal reset wage in period t. This implies that Wt(h) is preset as a constant markup over the expected weighted-average between future marginal rates of substitution between labor and consumption and aggregate wage rates, during the duration of the wage. This is a standard result in the New Keynesian literature, as derived, for example, in Gall (2008)

3.2 Homogenous good sector
 

Output in the homogenous sector is produced using labor, NO, subject to diminishing returns.[8]

8 The subscript O for the homogenous goods sector is a reference to ‘oil’.

 

where 0 < v <1. The  labor input  NO t is a CES aggregator of the individual varieties supplied by each household, .The output is sold on international markets at the world dollar price, PO,t, and is taken as given by the homogenous goods producers. Profits in this sector are given by,

 
 
and are non-zero because of diminishing returns. The optimal amount of labor hired is determined by the condition that equates the marginal product of labor to the wage rate.
 
 
3.3 Manufacturing Sector
Each firm in the manufacturing sector uses labor, NM (w), and imported intermediate in­puts, X(w), to produce a unique variety The production function is,
 
 
where zM,t captures aggregate productivity in the manufacturing sector. The labor input NM,t is a CES aggregator of the individual varieties supplied by each household,

We assume that all of the manufacturing output is exported, and the external demand for variety u is modeled using a Kimball aggregator that gives rise to strategic complemen­tarities in pricing. We adopt the Klenow-Willis (2006) formulation:

 

where Pf (u) is the price set by the manufacturing firm expressed in the currency of the foreign country (dollar). The parameter £ is the key parameter that influences the variabil­ity of desired mark-ups. In the limit as £ ^ 0 we obtain the CES benchmark of constant mark-ups when prices are fully flexible given by a/(a — 1). The time varying elasticity of demand is defined as at(u) = d log(Cft (u))/d ln(p/ (u)/pf).

Consistent with the small open economy assumption, the foreign price level Pf and aggregate consumption level, Cf are taken as exogenous. The firm’s per-period profits are:

 
3.4 Pricing

We contrast the outcomes under three pricing assumptions in international markets: pro­ducer currency pricing, local currency pricing, and dollar pricing. In each case we consider a Calvo pricing environment where firm’s are randomly chosen to reset prices with prob­ability 1 — 9p. Firms choose prices to maximize the expected net present value of profits conditional on no price change,

 
3.4.1 Dollar Pricing (DP)

Manufacturing exports: When prices are set in dollars, the optimal (dollar) reset price, Pf (w), satisfies the equation,

 

Define the (log) manufacturing export price index,

Imports: While these prices are taken as exogenous under the small open economy assumption, to capture DP, we assume that the price of imported goods is fixed in dollars at P¯∗ t . Accordingly, the home import price is:

 
3.4.2 Producer Currency Pricing (P CP)

Manufacturing exports: When prices are set in the producer’s currency, the optimal (peso) reset price, P/f (w) satisfies the equation,

The (log) manufacturing export price index is now,

 

where et is the log of the exchange rate, and p f h,t evolves according to

 

Imports: Under P CP, the price of imported goods is fixed in the currency of the producing country. Accordingly,

 

As is evident, the distinction between DP and P CP shows up on export prices, while it is the same in the case of import prices, given our environment where the ROW is the dollar block.9

3.4.3 Local Currency Pricing (LCP)

Manufacturing exports: The manufacturing pricing problem is identical to the case of dollar pricing.

Imports: Following LCP, we assume the price of imported goods is fixed in the home currency. Accordingly,

 

Definition 1 We solve for the rational expectations competitive equilibrium of the DP, P CP and LCP models, where

a) DP satisfies equations (2-19), P CP satisfies equations (2-16) and equations (20-22), and LCP satisfies equations (2-18) and equation (23).

b) Markets clear such that CN,t = YN , Nt = NO,t + NM,t, and Bh t = 0.

c) The monetary block of the model is described by a money supply rule, mt = ρmmt−1 + m,t and money demand mt − pt = ηcct − ηrr h t . 10
d) (log)World price of the homogenous goods sector evolves according to an AR(1), p ∗ o,t = ρop ∗ o,t−1 + o,t
 
Before we proceed to simulate the dynamics of the model it is useful to study the features of this economy in the long-run, when prices and wages are fully flexible.
 
4 Long-run

To provide an analytical characterization we consider the case where the utility function for consumption is in logs(ac = 1), the Frisch elasticity of labor     supply   is infinite (^ = 0), and trade is balanced. In the flexible price case we set E = 1 (without loss of generality) and express all prices relative to the price of the imported good, which we choose to be the numeraire.

The real exchange rate is simply a function of the relative price of non-traded goods, specifically PN-7, where PN is the price of the non-traded good expressed in units of the imported good.[9]

Under monopolistic pricing, and with the specification of Kimball demand chosen (12), (log) markup when prices are set flexibly satisfy.

9 This follows from the small open economy assumption.

 

and the price elasticity of the mark-up r = 5^/5 ln(Ptf /Pf) is given by,

 

Taking a log-linear approximation around a symmetric point, and combining the intratemporal condition eq. (4), combined with the flexible wage and price setting conditions, W P 1−γ N = C and P f M = e µ f W1−α αα(1−α) 1−α , and market clearing in the non-traded sector, CN = YN , we obtain the following proposition

Proposition 1 The greater is the strategic complementarity in pricing, r, and the higher the share of imported inputs in production, a, the lower is the pass-through of real exchange rate changes into prices of manufactured goods, and consequently, lower is the sensitivity of quantities exported to real exchange rate changes.

 

A real appreciation of the country’s currency raises the marginal cost of the firm, how­ever the extent to which it does so is decreasing in the share of imported intermediate inputs in production. This is captured by the term (1 — a). Conditional on the change in marginal costs, (1 — a)pN, the extent to which the firm raises its export prices depends on its desired mark-up. Given fixed world manufacturing prices, the greater the degree of strategic com­plementarity in pricing, r, the smaller the desire of the firm to pass-through the marginal cost increases into higher prices. As r ^ 0 and a ^ 0, we arrive at the benchmark of hundred percent pass-through from real exchange rate fluctuations. However, given that the data supports r > 0 and a > 0, pass-through is much more incomplete.

In the next proposition we consider a particular exogenous shock that drives changes in the real exchange rate, namely shocks to the relative price of the homogenous sector output Po. The external resource constraint CT + XT = POYO + Pf Cf is used to pin down the required change in the real exchange rate. The following proposition follows:

Proposition 2 The greater is the strategic complementarity in pricing, higher e and there­fore r, higher is the volatility of the real exchange rate and consumption. By contrast, higher is r lower is the volatility of the non-commodity terms of trade, pf, and of manufacturing exports to commodity price shocks.

The intuition for this proposition is as follows: Suppose there is a negative shock to the price of oil, which is a negative terms of trade shock. To satisfy the external resource constraint the country needs to cut down on its consumption of traded goods and imports of intermediate inputs. The extent to which it needs to do this, and consequently the extent of real depreciation (decline in relative price of non-traded goods) required depends on its ability to raise export income through higher exports from the manufacturing sector that benefits from lower marginal costs following the real depreciation. As follows from Proposition 1, the higher is r lower is the gain in exports for a given depreciation of the real exchange rate. A higher r therefore calls for a larger depreciation.[10] A larger depreciation therefore goes along with a smaller depreciation of the (manufacturing) terms of trade. A more volatile real exchange rate therefore goes hand in hand with a less volatile manufacturing terms of trade.

We now proceed to a dynamic analysis that contrasts the three pricing models.

10 The comparative statics with respect to a do not yield itself to closed form solutions because the steady state values sO and sM also depend on a.

5 Dollar Pricing Redux

We now solve for the dynamics of the three pricing models employing log linear approxi­mations around a zero inflation steady state. The reset price equations for manufacturing output (17) and (20) give rise to the below pricing dynamics for the aggregate manufac­turing price index under DP (LCP) and PCP regimes, respectively.

 

where £ = log ^31. In the case of DP the marginal cost on the right hand side is expressed in dollars, mct — et, while in the case of PCP, it is expressed in the home currency. Because this is a two country environment DP and LCP have the same pricing equations for manufacturing exports. However, this does not imply that the chosen prices are the same as we demonstrate below, because of differences in the evolution of marginal costs across the two cases.

The higher the degree of variable mark-ups, that is higher is r, the lower the sensitivity of prices (in the currency in which they are set) to changes in marginal costs, as would be the case with a higher degree of price-stickiness 0p. At the same time a higher r raises the sensitivity to the foreign manufacturing price index. The evolution of the aggregate price indices follow, p{,t p{,t-i = (1 — 0p)(pfit p[t-1) for i = f,h.

 

5.1 Impulse Responses

In this section we study the response to a one time negative price shock to the homogenous good sector. As described in Section 4, such a terms of trade shock, in the presence of incomplete financial markets, generates a depreciation of the real exchange rate. In the presence of nominal rigidities this is partly accomplished by a depreciation of the nominal exchange rate.

Figures 3(c), 3(e), 3(f) and 4(d) contrast the impulse response of the terms of trade, manufacturing exports, imports and manufacturing firm’s mark-up to the shock. As is evident these impulse responses display in almost all cases important qualitative and quan­titative differences. First consider the non-commodity terms of trade, 3(c), defined as the ratio of the price of manufacturing exports to the price of imports (totM), all expressed in dollar terms, pPt~. In the case of DP there is a very slight depreciation of the terms of trade in contrast to a sharp depreciation of the terms of trade in the case of PCP and a sharp appreciation of the terms of trade in the case of LCP.

In the case of DP and LCP because prices are sticky in dollars the decline in dollar prices is gradual in both cases, however it declines by more in the case of LCP as depicted in figure 3(d). This is because in the case of DP the depreciation of the exchange rate raises the home currency cost of inputs, and consequently the decline in marginal costs that the depreciation affords is smaller as compared to the case of LCP for which imported input prices are sticky in the home currency. The small response of dollar export prices, and the constancy of dollar import prices together generate a small depreciation of totM. The sharp appreciation of the totM in the case of LCP follows from the decline in import prices denominated in dollars, a consequence of stickiness in local currency. As for PCP the sharp depreciation of totM follows from the stickiness of export prices in the home currency. The differential quantity response for manufacturing exports, depicted in 3(e), follows from the differential export price responses in the three cases.

The current account to GDP deteriorates in all three cases, figure 4(a), which follows from the decision of households to borrow in the face of a transitory negative shock. The non-commodity trade balance, figure 4(c) improves owing the fall in imports and increase in exports, however this improvement is the smallest for the case of DP owing to the smallest response in manufacturing exports.

Figure 4(d) plots the diverging responses of mark-ups of manufacturing sector firms, with a sharp increase in mark-up in the case of LCP, a sharp decline in mark-up in the case of PCP and a moderate increase in mark-up in the case of DP. Once again the difference between LCP and DP arises from the lack of sensitivity of imported input costs to the exchange rate in the case of LCP, which in turn generates a sharper increase in mark-ups.

5.2 Exchange Rate Pass-through

It is common to estimate the pass-through of exchange rates into a country’s import and export prices to understand the pricing behavior of firms. Burstein and Gopinath (2014b) provide a recent survey of this literature. One specification involves regressing the change in (log) prices on current and lagged changes in the (log) exchange rate.

In figures 5(a)-5(f), we plot the cumulative sum of the coefficients of the exchange rates from the following regressions estimated on the model simulated data,

 

Figures 5(a)-5(c) depicts the pass-through of exchange rate changes into the home currency price of exports (dashed line), where Ast = Apf + Aet, and the home currency price of imports (solid line), ApT,t, for the three cases of DP, PCP and LCP. Each point represents J^j=0 / There are large differences across the three cases. In the case of DP both pass-through’s start out close to one and the pass-through into export prices declines towards 0.66. The flat line for import pass-through follows from the constancy of Pft. The important feature is that both pass-through’s are quantitatively close and they decline over time.

In the case of PCP, the contemporaneous effect for pass-through’s diverge sharply, with the pass-through for export prices below 0.2 and the pass-through for import prices close to one. The pass-through into home currency export prices increase over time, in contrast with the DP case. In the case of LCP there is a similar divergence in the pass-through rates for early lags as in the case of PCP, with the important difference that the lines are flipped around, that is pass-through into export prices start high, while pass-through into import prices start low.

Figures 5(d)-5(f) plot the cumulative coefficients for pass-through into quantities, with the dashed line representing exports (manufacturing only) and the solid line imports. Con­sistent with the pass-through estimates for prices, in the case of DP the pass-through into export quantities is small with a gradual increase over time, while there is a sharp decline in imports. In other words, exchange rate depreciations in the case of DP result in a bigger response in imports than in exports. In the case of PCP, there is a sharp increase in exports and a small increase in imports initially with a gradual decrease over time. The initial small positive arises because of the demand for imported inputs used to produce exports that rises sharply. In the case of LCP both exports and imports increase over time.

Armed with these contrasting predictions we turn to empirical evidence in the next section to examine how the facts compare with the predictions.[11]

6 Empirical Evidence

To investigate the relationship between firms, exchange rates and the macroeconomy we use data for Colombia. Besides data availability, there are other good reasons to study Colombia. Firstly, because the Colombian exchange rate (peso) is a commodity currency and fluctuations in the peso[12] are closely tied to fluctuations in commodity prices[13] we are less subject to the long standing critique of standard general equilibrium models of exchange rates that have trouble explaining exchange rate fluctuations at quarterly frequencies using observable shocks, the so called exchange rate disconnect puzzle (Meese and Rogoff (1983)). In the case of Colombia there is a strong negative co-movement between the exchange rate and commodity prices at quarterly frequencies. Figure 7 displays the relation between the Colombian peso (solid black line) and the overall terms of trade (dashed blue line), where the two closely track each other. The correlation between the two series is -0.62, and the regression coefficient is -1.15 with an R2 of 0.38. The Colombian economy therefore provides a useful lens through which to evaluate the performance of open economy macro models. Secondly, the Colombian economy is representative of a large number of economies that rely extensively on dollar invoicing, with 98% of its exports invoiced in dollars. We provide further details in the next section.

11 Here, we briefly speak to the question of endogenous currency choice. That is, given the model’s assumptions and its specific parameterizations, if firms in the SOE could choose which currency to price in, what would it be? As derived in Gopinath et al. (2010) a proxy measure for this decision is the pass­through conditional on a price change. If this pass-through is more (less) than 0.5 (this is an argument based on second order approximations) the firm will choose to price in dollars (peso). In the model generated data this conditional pass-through estimate is 0.86 and consequently the firm would choose to price in dollars. The use of imported inputs and strategic complementarity in pricing are both essential to making dollar pricing optimal.

12 The Colombian peso officially switched to a floating status in 1999.

13 These commodity prices are also exogenous to the economy because while mining output makes up 58.4% of total exports for Colombia, it is small relative to world commodity markets. For example, Colombia’s oil production was 1.1% of world oil production in 2014.

6.1 Data Sources

We combine two different firm-level datasets: one has information on firms’ foreign market participation, while the other contains detailed balance sheet and operational information.

The data on international trade are from the customs agency (DIAN), and the depart­ment of statistics (DANE), and includes information on all importers and exporters. We have access to the data through the Banco de la Republica. The data on exports include the exporting firm’s tax identification number, the 10-digit product code (according to the Nandina classification system, based on the Harmonized System), the FOB value (in U.S. dollars) and volume (net kilograms) of exports, the country of destination, and the currency of each transaction, among other details. The data on imports include the importing firm’s tax identification number, the 10-digit product code, both the FOB and CIF value and the volume of imports, and the countries of origin and purchase, among other details. Both datasets are available on a monthly basis, and for our analysis we aggregate exports and imports at the annual or quarterly level. Trade data is available for the period between 2000 and 2015.

The data on firms’ production and input consumption are from “Superintendencia de Sociedades,” the agency in charge of supervising corporations. Specifically, the data are from the “Sistema de Information y Riesgo Empresarial” (SIREM) database. The SIREM includes relatively large firms, and firms in financial trouble.[14] The data are at an annual frequency and are self-reported by the firms. We have access to public information such as balance sheets, as well as to confidential data included in the annexes filed by the firms.[15] Thus, we are able to observe a great amount of details about each firm that are not usu­ally available in other datasets. These variables include the firm’s tax identification, the income obtained from the sales of each product (at the 4-digit level, according to the ISIC classification system), the purchase of inputs (broken down between domestic and foreign materials), investments, detailed labor information, and some financial information such as financial obligations with domestic and foreign banks. Additionally, we observe organiza­tional variables such as whether the firm is a standalone firm, an affiliate, a headquarters with affiliates, or part of a conglomerate, and the firm’s location. Our firm data cover the period 2005-2014.

Throughout the paper, we focus on manufacturing firms producing noncommodity, trad­able goods.[16] For our analysis we exclude manufacturers of coke, refined petroleum products and nuclear fuel (ISIC 23), and manufacturers of basic metals (ISIC 27), which include met­als such as gold, silver, platinum, and nickel, since as commodity producers their dynamics are different from those of other manufacturing firms.

Given that both datasets include information on firms from several industries, and that we observe in the data multi-product firms not limited to manufacturing activities, we need to define precisely which observations to use in our estimations. For the empirical exercises that rely exclusively on our trade data, since we have information at the product level, we keep all the observations corresponding to exports and imports of manufactures. In the case of the SIREM, where data is at the firm level, we need to classify each firm (and not each product) as a manufacturer or not. Taking advantage of the detailed income data, we define as manufacturers those firms for which income from manufactures represents at least 50% of the total operational income in any given year, and we exclude those firms with income from the manufacture of coke, refined petroleum products, nuclear fuel, or basic metals.[17]

We merge both datasets using the firms’ tax identification number. After the merge we are left, on average, with 34% of all exporters (and 40% of exporters of manufactures). Since we keep relatively large firms, these represent 65% of the value and 80% of the volume of all exports, and 77% of the value and 63% of the volume of manufacturing exports. In

Table 3, we break down these percentages by year. While we do observe some variations from year to year, the share of exports included in our merged data remains fairly constant.

When we look at the representativeness of our data in terms of imports, we find a similar pattern. In terms of the total number of importers, our merged data contains 27% of all importers and 30% of the importers of manufactures. However, as mentioned above, these represent relatively large firms, and the share of imports is much larger: the firms in our sample represent over 74% of total imports, measured either with the FOB value or the net weight of imported goods. Table 4 shows these percentages broken down by year. Again, the share of imports included in our merged dataset is relatively constant across years.

In Table 5, we break exports across destinations. The table includes countries with a share of manufacturing exports of at least 1%. Based on the value of manufacturing exports, Colombia’s main trading partners are the United States and Venezuela, followed by Ecuador and Peru. The list is completed by other Latin American countries, European countries, Japan and Canada.[18]

Starting in 2007, our exports data includes information on the invoicing currency of each transaction. The distribution of exports across currencies is presented in Table 6. It is evident that the vast majority of Colombian exports are priced in dollars. Although some transactions are negotiated in euros, Colombian pesos, or Venezuelan bolivares among other currencies, the U.S. dollar accounts for over 98% of all exports. Moreover, the distribution is very similar if we look at the value of exports negotiated in each currency instead of the number of transactions. In Table 7 we present the distribution of currencies, broken down by destination groups. Even for exports to the euro zone, or the U.K, the overwhelming invoicing currency is the dollar.

14 Firms must report their financial data if their assets and/or income (adjusted by inflation) are grater than 30,000 times the current legal monthly minimum wage, if their external liability is grater than the total assets, if the financial expenditures are at least 50% of their income, if their cash flow is negative, or if their losses reduce the net equity below 70% of the social capital. During our sample period the average minimum monthly wage was 250 dollars; it oscillated between a minimum of 165 dollars in 2005, and a maximum of 315 dollars in 2012 and 2013.

15We obtained access to the confidential data through the Banco de la Republica.

16 We follow the ISIC classification (Rev. 3.1) to define which goods are manufactures.

17 As alternative definitions of manufacturing firms we considered those whose main product is a manu­facture, or those with positive income from manufacturing. The descriptive statistics presented below and the main results of the pass-through regressions are very similar under all definitions.

18 It is worth noting that among the countries included in “Other”, China, the Netherlands, India, and Switzerland are important trade partners for goods not included in our set of manufactures. In particular, China receives 5% of total exports, and 3% of exports of manufactures including petrochemicals and basic metals.

6.2 Terms of Trade

Figure 7 plots the manufacturing terms of trade (dashed and dotted red line) for Colombia alongside the overall terms of trade and the nominal exchange rate. Unlike the overall terms of trade the manufacturing terms of trade has low sensitivity to the exchange rate. A 1% depreciation of the exchange rate is associated with a 0.33% depreciation of the manufacturing terms of trade. Using model simulated data we find that DP comes closest to this low sensitivity with a depreciation of 0.24%, PCP generates a much greater sensitivity of 0.68%. LCP generates the wrong sign with a 0.5% appreciation following a depreciation of the exchange rate.

6.3 Empirical Pass-through Estimates

Figures 6 and 7 plot the empirical counterpart of eq. (26) using Colombian import and export price and quantity data. Specifically, for export (import) prices we construct unit values using the f.o.b value at the 10-digit product code level for a specific destination (origin) divided by the volume of exports at the same level of disaggregation for each quarter from 2005-2015. To be consistent with the model we restrict exports to and imports from dollarized economies. These include the USA, Panama, Puerto Rico, Ecuador, and El Salvador. We include fixed effects at the firm-country-industry level. The solid lines represent the cumulative 3 over nine quarters and the dashed lines represent two standard error bands, where standard errors are clustered by year. The pass-through into import (solid line) and export (solid with squares) peso prices start out close to one for import prices and to 0.8 for export prices and then decline some over time.

Table 9 reports annual pass-through estimates using annual changes in peso export prices and annual changes in exchange rates. In columns 1-2 we just include the exchange rate and the same firm-country-industry fixed effects as in the figures. In columns 3-4, the estimates are reported using controls for origin and destination producer price index (PPI), as controls for costs and allowing for strategic complementarity in pricing. In columns 5­6 we use controls for the firms productivity estimated using Gandhi et al. (2016)[19] We also estimate pass-throughs for the subset of manufactured goods that are differentiated products as defined by Rauch (1999). These are reported in the even-numbered columns, and it is clear that the results are essentially unchanged. We repeat the exercise for import prices in Table 7. Regardless of the controls used, pass-through into export peso prices is close to 0.8 and for import prices is close to 1 over one year. The destination country’s PPI also shows up as significant for pricing decisions. In line with strategic complementarity in pricing a higher price in the destination country all else equal is associated with the exporting firm charging a higher price. When compared to the model generated figures 5(a)-5(f) it is clear that the pass-through patterns in the data support strongly the DP model as opposed to PCP or LCP.

Figure 7 and Tables 11 and 12 perform the same pass-through analysis but where the dependent variable is quantities instead of prices. The quantities exported depend on the price charged by the exporting firm relative to the price charged by its competitors (PPI*) and the level of demand in the destination country. The price charged by the firm in turn depends on the exchange rate, its own costs (PPI) and the competitors price level (PPI*). We therefore estimate a specification that has export quantities on the left hand side, and the exchange rate, domestic and foreign PPI and foreign demand on the right hand side. We do the same for imports but with domestic demand replacing foreign demand.

The figure suggests that while there is a significant drop in quantities imported, the estimates for exports is highly noisy and insignificantly different from zero. Table 11 reports the pass-through into export quantities. We observe that while the coefficient on the exchange rate are negative when we do not include any controls, they become statistically insignificant once we add foreign PPI and GDP to control for costs and foreign demand. Similarly, Table 12 reports the pass-through into import quantities. In this case, we find that the pass-through is always negative and close to -1.1. This pattern is again consistent with the predictions of DP and contrary to those of LCP and PCP.

19 We estimate TFP following the method proposed by Gandhi et al. (2016), that exploits the information contained in the firm’s first-order conditions to account for unobserved productivity through the observed demand for flexible inputs, while eliminating the productivity term from the estimation itself to avoid the simultaneity bias. We use information on sales and input usage from our SIREM data to estimate firm-specific productivity for the period between 2005-2013.

7 Conclusion

In this paper we examine the combined features of dollar pricing, strategic complemen­tarities in pricing and imported input use in production to understand the consequences for economies of shocks that generate fluctuations in the exchange rate. We demonstrate that these predictions when compared to the data outperform the dominant paradigms of producer and local currency pricing in the literature. In future work we plan to extend the model to a multi-country environment so as to capture the consequences of using the dollar in third party transactions, an important feature of the data.

References

Bernard, A. B., Jensen, J. B., and Schott, P. K. (2009). Importers, Exporters and Multi¬nationals: A Portrait of Firms in the U.S. that Trade Goods. In Producer Dynamics: New Evidence from Micro Data, NBER Chapters, pages 513-552. National Bureau of Economic Research, Inc.
Betts, C. and Devereux, M. (2000). Exchange rate dynamics in a model of pricing-to- market. Journal of International Economics, 50(1):215-44.
Burstein, A. and Gopinath, G. (2014a). International prices and exchange rates. In Gopinath, G., Helpman, E., and Rogoff, K., editors, Handbook of International Eco¬nomics, volume 4 of Handbook of International Economics, pages 391 - 451. Elsevier.
Burstein, A. and Gopinath, G. (2014b). International prices and exchange rates. In Gopinath, G., Helpman, E., and Rogoff, K., editors, Handbook of International Eco¬nomics, volume 4, pages 391 - 451. Elsevier.
Chen, Y.-C., Rogoff, K. S., and Rossi, B. (2010). Can exchange rates forecast commodity prices? The Quarterly Journal of Economics, 125(3):1145-1194.
Corsetti, G., Dedola, L., and Leduc, S. (2010). Chapter 16 - optimal monetary policy in open economies? volume 3 of Handbook of Monetary Economics, pages 861 - 933. Elsevier.
Corsetti, G. and Pesenti, P. (2005). The Simple Geometry of Transmission and Stabilization in Closed and Open Economies. NBER Working Papers 11341, National Bureau of Economic Research, Inc.
Devereux, M. and Engel, C. (2003). Monetary policy in the open economy revisited: Price setting and exchange rate flexibility. Review of Economic Studies, 70:765-84.
Devereux, M. B., Shi, K., and Xu, J. (2007). Global monetary policy under a dollar standard. Journal of International Economics, 71(1):113-132.
Fitzgerald, D. and Haller, S. (2012). Exchange rates and producer prices: Evidence from micro data. Working Paper, Stanford University.
Fleming, J. M. (1962). Domestic financial policies under fixed and under floating exchange rates (politiques finacirieures intrieures avec un systme de taux de change fixe et avec un systme de taux de change fluctuant) (poltica financiera interna bajo sistemas de tipos de cambio fijos o de tipos de cambio fluctuantes). Staff Papers (International Monetary Fund), 9(3):369-380.
Gali, J. (2008). Monetary Policy, Inflation and the Business Cycle: An Introduction to the New Keynesian Framework. Princeton University Press.
Gandhi, A., Navarro, S., and Rivers, D. (2016). On the identification of production func¬tions: How heterogeneous is productivity? (Unpublished; University of Wisconsin- Madison).
Goldberg, L. (2013). The international role of the dollar: Does it matter if this changes? In Westermann, F. and Cheung, Y.-W., editors, Global Interdependence, De-Coupling, and Re-Coupling. MIT Press and CES IFO.
Goldberg, L. and Tille, C. (2009). Macroeconomic interdependence and the international role of the dollar. Journal of Monetary Economics, 56(7):990-1003.
Goldberg, L. S. and Tille, C. (2008). Vehicle currency use in international trade. Journal of International Economics, 76(2):177-192.
Gopinath, G. (2015). The international price system. In Jackson Hole Symposium, vol¬ume 27. Kansas City Federal Reserve.
Gopinath, G., Itskhoki, O., and Rigobon, R. (2010). Currency choice and exchange rate pass-through. American Economic Review, 100(1):306-336.
Gopinath, G. and Rigobon, R. (2008). Sticky borders. Quarterly Journal of Economics, 123(2):531-575.
Ito, H. and Chinn, M. (2013). The rise of the ‘redback’ and china’s capital account liber¬alization: An empirical analysis on the determinants of invoicing currencies.
Johnson, R. C. (2014). Five facts about value-added exports and implications for macroe¬conomics and trade research. Journal of Economic Perspectives, 28(2):119-42.
Johnson, R. C. and Noguera, G. (2012). Accounting for intermediates: Production sharing and trade in value added. Journal of International Economics, 86(2):224 - 236.
Kugler, M. and Verhoogen, E. (2009). Plants and imported inputs: New facts and an interpretation. American Economic Review, 99(2):501-07.
Manova, K. and Zhang, Z. (2009). China’s Exporters and Importers: Firms, Products and Trade Partners. NBER Working Papers 15249, National Bureau of Economic Research, Inc.
Meese, R. and Rogoff, K. (1983). Empirical exchange rate models of the seventies: Do they fit out of sample? Journal of International Economics, 14(1):3-24.
Mundell, R. A. (1963). Capital mobility and stabilization policy under fixed and flexible exchange rates. The Canadian Journal of Economics and Political Science / Revue canadienne d’Economique et de Science politique, 29(4):475-485.
Obstfeld, M. and Rogoff, K. (1995). Exchange rate dynamics redux. Journal of Political Economy, 103:624-60.
Rauch, J. (1999). Networks versus markets in international trade. Journal of International Economics, 48:7-35.
Svensson, L. and van Wijnbergen, S. (1989). Excess capacity, monopolistic competition, and international transmission of monetary disturbances. Economic Journal, 99. 
 

 
 
 
Comments by Philip Lane

This is an excellent paper. It fills an important gap in the international macroeconomics literature by providing a workhorse model that is suitable for the analysis of small open economies. The traditional view of smallness was that a country faces externally determined international prices. In the new open economy macroeconomics (NOEM) literature, the primary focus was on the endogenous determination of international prices: with monopolistic competition, the exporting firms in even small countries had sufficient market power to set prices22. In widely used versions of the NOEM model, smaller economies are structurally similar to larger economies, but just account for a smaller share of world population and world output.

The model here takes seriously the concept of smallness. In particular, import prices are externally determined. While export prices are decided by domestic firms, the assumption of strategic complementarity in pricing decisions means that firms optimally choose to limit the extent of price adjustments in response to shocks. Accordingly, the terms of trade are quite stable, in contrast with more usual setups in the NOEM literature that generate substantial movements in the terms of trade.

In a multi-country world, it is natural that a few dominant currencies emerge as a unit of account for international transactions. While the US dollar is the dominant international currency for many countries, the euro dominates cross-border transactions among members of the monetary union and vis-a-vis a number of neighbouring countries. It is also the case that the list of dominant currencies might shift over time. For instance, if China shifted to a monetary regime that tolerated greater volatility of the Chinese renminbi against the dollar, it is possible that various small Asian economies might switch from dollar pricing to RMB pricing.

The model in this paper provides a natural framework for thinking about the role of dominant currencies: it is quite intuitive that dominant currency pricing should be the benchmark pricing convention for most small economies, compared with the local currency pricing (LCP) and producer currency pricing (PCP) models that have been most studied in NOEM literature. Of course, the LCP and PCP models remain highly relevant for thinking about the interactions across large blocs (Europe versus the United States versus Asia) in global models.

The main findings of the model are supported by the empirical work based on high-grade Colombian data, but also accord more broadly with the general experience of many small economies. In particular, we know that the relative price of imports to domestic goods is a more important mechanism inducing expenditure switching, with movements in international export prices a much more limited channel by which exchange rates operate, especially in the short run. The capability to match the data is enabled by allowing for international trade in intermediate inputs, since imported intermediates act to moderate the impact of exchange rate

 

There were some exceptions. The textbook by Obstfeld and Rogoff (1996) included a small open economy model in which international prices were exogenous but the relative price of non-traded goods was determined endogenously. This model was applied in Lane (1997) to examine the implications of variation in the relative share of non-tradables in consumption for steady state inflation rates.

movements on competitiveness indicators. Under dominant currency pricing, the impact of bilateral exchange rate movements between pairs of non-dominant currencies has to be understood in the context of the co-movement patterns with the dominant currency, which is clearly demonstrated in the quantitative work laid out by the authors.

In terms of the future research agenda, the model could be applied to revisit some classic topics in international macroeconomics. For instance, it would be interesting to re-examine the transfer problem: the real exchange rate adjustment required to support the generation of trade surpluses for countries with large net external liabilities. Applying the model to understand the relative properties of alternative exchange rate regimes (ie floats, pegs and monetary unions) would also be an interesting project, as would the related question of optimal reserves holdings.

It could also be useful to examine various model extensions. For instance, various financial frictions could be studied, including a trade finance restriction in funding imported intermediates and trade-offs between domestic-currency and foreign- currency debt issuance.

In relation to dominant currencies, it is also important to recognise that there can be differences between the dominant currency in trade and the dominant currency in finance, which can give rise to asymmetries between the trade and financial impact of exchange rate movements (Devereux and Lane (2003); Lane and Shambaugh (2010)). The authors might be able to develop further the model to allow for differences in dominant currencies across the trade and financial dimensions.

Finally, it may be worthwhile to also consider some alternative setups. For many purposes, it might be simpler to assume that exporting firms are just price takers in international markets, with domestic price dynamics captured by a segmented non­traded sector (Obstfeld and Rogoff (1996); Lane (1997); Clancy and Merola (2016)). Such a setup might generate quite similar results to those obtained under the more elaborate model outlined in this paper.

References
 
Clancy, D and R Merola (2016): "Eire mod: A DSGE model of Ireland", Economic and Social Review, vol 47, no 1, pp 1-31.
Devereux, M and P Lane (2003): "Understanding bilateral exchange rate volatility", Journal of International Economics, vol 60, no 1, pp 109-32.
Lane, P (1997): "Inflation in open economies", Journal of International Economics, no 42, pp 327-46.
Lane, P and J Shambaugh (2010): "Financial exchange rates and international currency exposures", American Economic Review, vol 100, no 1, 518-40.
Obstfeld, M and K Rogoff (1996): Foundations of international macroeconomics, MIT Press.

Comments by Richard Baldwin∗

Let me start by congratulating the authors on a well written and brave paper that aims to help fill in some of the shortcomings in open macroeconomic modelling literature. I have visited a number of central banks in emerging markets and repeatedly heard complaints that the standard macro toolkit is too centred on advanced economies - that emerging markets face a set of questions that are not naturally addressed with the standard toolkit. In my view, and remember I am not a macroeconomist, this paper adds a model with some special emerging market features and thus broadens the toolset.

As Philip Lane has focused on the macro aspects, I will focus on the trade aspects using the diagram below.

 

The model focuses on an open, quasi-small nation, which uses a national currency (the peso) to set its wages and the price of its non-traded good. The nation exports both a commodity (ie oil) and manufactured goods. It imports intermediate goods, X, that are used only in the manufacturing sector and directly in domestic consumption (very much in the spirit of Krugman and Venables (1995)).

The country is small in terms of the oil market and the market for imported inputs (X goods) in the sense that the dollar denominated oil price is fixed. The import price for X-sector (intermediate) goods is fixed either in pesos or US dollars (the authors consider both cases). The country, however, is large in terms of trade when it comes to manufacturing since its firms set prices for their differentiated manufactured exports. The authors assume a foreign demand structure for the nation's M-sector exports that makes pricing to market optimal (referred to as strategic complementarity in pricing).

In terms of stickiness, the oil price is fully flexible and exogenous, while the manufacturing price is sticky (since wages are) and endogenous. The price of the non-traded good is endogenous but fully flexible since it is produced without any labour input (it is modelled as an endowment good that is available in a fixed quantity in each period).

Pricing in which currency?

The microeconomics of the model are not yet complete since the macroeconomics of the model require the authors to specify the currency in which prices are set. They consider two standard assumptions - that all prices are either set in the consumer's currency (so-called local currency pricing), or all prices are set in the producer's currency (producer currency pricing). Note that the assumptions are not quite consistent since oil is priced in dollars even in the local currency pricing case (which seems reasonable since it is meant to represent a commodity like oil).

The authors also consider a hybrid case where the nation's manufacturing exporters price in their customers' currency (the dollar), but the foreign exporters of the imported intermediate price in their own currency (the dollar). They refer to this as a dollar pricing case because every traded good is priced in dollars. To clarify, the array of pricing assumptions made in the model in the three cases are shown in the table.

 

Comments on the trade aspects of the model

A number of assumptions are quite ad hoc from the trade perspective. For instance, while foreign manufactured goods (the X-sector goods) are both used in domestic production and consumption, the nation's own manufactured goods are only exported. Likewise, the export commodity (the O-sector good) is only exported and not imported.

While the lack of local consumption of the manufactured good seem innocuous (they could probably deal with it easily by assuming a segmented market structure so domestic and export prices are set separately), the local consumption of oil seems more problematic. The exchange rate and local currency price of oil strikes me as a first order macro issue for oil-exporting nations.

Using data from Columbia, the authors suggest that the dollar pricing hybrid is a better fit with reality than other more standard assumptions of local or producer currency pricing.

Concluding remarks

This is an informative paper that expands on the range of macroeconomic models by including a model that has features that are particularly important to commodity- exporting emerging markets. The only comment I would make beyond praise is to say that the authors could justify the dollar pricing assumption as really just a simple extension of the small country assumption. That is, the country is too small to force the exchange rate volatility onto its export customers (in the O and M sectors) and is too small for foreign producers to consider absorbing the exchange rate volatility that peso pricing would entail.

References
Krugman, P and A Venables (1995): "Globalization and the inequality of nations," The Quarterly Journal of Economics, vol 110, no 4, pp 8.