شیوه های صورتحساب ارز، نوسانات نرخ ارز و قیمت گذاری به بازار: شواهد به دست آمده از داده های سطح محصول

The paper investigates the long-run relationship between exchange rate volatility and the pricing-to-market policies of international exporting firms when such firms choose between different currencies (exporters, importers or vehicle) for invoicing their trading partners. Distinctively, our analysis is conducted at the level of the product, using data from the Norwegian fishing industry. Dynamic error correction models are formulated to capture the long-run relationship between exchange rate pass-through elasticities and the different currency invoicing strategies. Exchange rate pass-through coefficients vary from 0.07 to 0.98 across products. Moreover, for a given product, pass-through coefficients vary significantly both across and within destination markets, depending upon the invoicing currency chosen. This variation is linked to nominal rigidities and exchange rate uncertainty. The findings also suggest that the choice of invoicing currency may be an important strategic variable facilitating discriminatory pricing by exporting firms. Finally, the results also corroborate theoretical predictions linking pass-through to exchange rate volatility; namely, pass-through is lower the more volatile the exchange rate.

Studies of international financial management practices generally separate the analysis of exchange rate pass-through, exchange rate risk management and choice of currency of invoicing. The central contribution of this paper is to begin to partially integrate these issues. We investigate the relationship between exchange rate pass-through and the pricing policies of international exporting firms when such firms choose between different currencies for invoicing their trading partners. Using a variant of the logarithmic market model (see Hooper & Mann, 1989), the paper empirically estimates how the degree of exchange rate pass-through at a product level alters in relation to differing invoicing practices of the exporting firm. We argue that the results obtained may yield important insights into how firms and MNEs that operate in international markets use currency invoicing decisions as part of their pricing strategy. Moreover, we believe that the findings also provide the first evidence relevant to existing theoretical predictions linking choice of invoicing currency to exchange rate volatility (Friberg, 1998). If exporters have some market power and markets are segmented, an exchange rate change may induce price discrimination across destination markets, or pricing-to-market in Krugman’s (1987) terminology. Alternatively stated, exporters will set different prices, expressed in the exporters’ currency, in different destination markets. This in turn implies that the exchange rate pass-through, defined as the response of the destination (import) market price of the product to an exchange rate change, is incomplete. A destination-specific mark-up adjustment thus absorbs part of the exchange rate change, and deviations from the law of one price will generally occur. Pricing-to-market could depend on either nominal price rigidities and exchange rate surprises (as in Giovannini, 1988) or it could be due to deliberate price discrimination, which in turn could be related to market conditions specific to the destination market. The vast majority of studies focus on the latter issue, and several emphasise the importance of market structure characteristics.1 The extent of pricing-to-market has been explained by both the size of, and the desire to, maintain product market share (Feenstra Gagnon & Knetter, 1996). Slow demand adjustment implies that current price-setting strategy will affect future revenues, suggesting there may be benefits in having a stable short-run price in order to secure market share (Gottfries, 1994 and Krugman, 1987). This raises the possibility that intertemporal links related to the expected permanence of an exchange rate change might influence a firm’s pricing strategy (Froot & Klemperer, 1989). The importance of forward looking exchange rate expectations can also be related to adjustment costs on the supply side (Kasa, 1992). In these models the mechanism for pricing-to-market is (primarily) deliberate price discrimination, and is not linked to either nominal price stickiness or the choice of invoicing currency.2 In contrast, the main theoretical focus of this paper is precisely this latter issue, the potential use of various invoicing currencies as a strategic instrument for pricing-to-market. Our approach also enables us to address the possibility that firms use invoicing decisions to facilitate intra-market price discrimination within a destination market, an issue customarily neglected in previous pass-through studies. An extensive empirical literature on the relationship between exchange rate changes and price adjustments of traded goods has accumulated, both at the level of aggregate imports and exports, and also using industry level data. Goldberg and Knetter (1997) and Menon (1995) provide excellent surveys of these studies. Much of this empirical work relates to large open economies, particularly the US, Japan and Germany. The empirical evidence documenting the existence of pricing-to-market and incomplete exchange rate pass-through for these countries is substantial. Many studies indicate that exporters price discriminate between markets and take destination-specific market conditions into consideration when setting prices (Gagnon and Knetter, 1995, Giovannini, 1988, Kasa, 1992, Knetter, 1989, Knetter, 1993 and Marston, 1990). However, since the production and destination markets vary greatly across different products, we maintain that a study conducted at the aggregate firm or industry level may not capture all the underlying variables that influence strategic pricing practices. Moreover, pass-through estimates vary widely by industry across different markets, making comparison and generalisation of results difficult. A distinctive feature of this paper is that we consider currency invoicing practices and their relationship to pass-through at the level of the individual product. We submit that this micro-level focus is a particularly noteworthy strength of the paper, enabling us to capture empirical regularities, including intra-market price differences, which would not be apparent in studies undertaken at a more aggregated industry level. The products selected for analyses are among the major exported Norwegian fish products. Empirical studies of pass-through at the level of the product rather than the industry are scarce, and no existing study of which we are aware considers the relationship between pass-through and invoicing practices. Other product-level analyses include: Hänninen (1998), studying the long-run influence of exchange rate changes on Finnish sawnwood prices in the UK; Hänninen and Toppinen (1999) whose focus is the long-run exchange rate elasticity of Finnish newsprint and pulp export prices to the UK and Germany; and Adolfson (1999) who analyses the long run pass-through coefficients for automobile and kraft paper exports from Sweden to Germany, France and the UK. However, unlike the present analysis, none of these studies differentiates between the use of different currencies for invoicing exports, and proceed as if all exports are invoiced in the currency of the exporter. Our approach enables us to demonstrate one important new result, namely that there exist significant intra-market differences in the estimates of pass-through obtained according to the choice of invoicing currency adopted by the firm. Price discrimination occurs not just across destination export markets, but also within a given destination market, and is facilitated by a firm’s choice of invoicing currency. This study’s methodology is easily summarised and contrasts markedly with the single equation approach generally used in previous analyses. It formulates a dynamic error correction model that can simultaneously handle the existence of several endogenous variables and ascertain the long run relationship between them (Johansen, 1988 and Johansen, 1991). This methodology is acknowledged to be more robust and efficient than the customary single equation approaches to analysing pass-through, but to date there exist relatively few studies in which it is used to study pricing-to-market effects (Adolfson, 1999, Hänninen, 1998, Hänninen and Toppinen, 1999 and Hung, Kim and Ohno, 1993). We also take steps to ensure that the approach is not subject to simultaneity bias of the kind identified by Kenny and McGettingan (1998). Several hypothesis linked to alternative characterisations of the price determination process are tested by imposing linear restrictions on the specified price-setting relationship. We interpret the degree of pricing-to-market as being reflected in the magnitude of the estimated exchange rate pass-through coefficients across products exported to different destinations, controlling for the choice of invoicing currency. Finally, we note that for small open economies, the conventional theoretical presumption is that exporters are price-takers. This implies that they face an exogenously determined export price in foreign currency and that there is immediate and complete incorporation of both changes in exchange rates and market prices into prices stated in their own currency. In contrast to this belief, several empirical studies indicate that the pricing-to-market hypothesis is relevant also for small open economies (Alexius and Vredin, 1999, Athukolara and Menon, 1995, Gottfries, 1994 and Menon, 1992). Even in such an environment, export producers appear to have some market power and ability to affect prices. Our results support these conjectures. The remainder of the paper’s development is easily outlined. The essential institutional background to fish exports from Norway is given in Section 2. Section 3 outlines the features of the linearised version of the logarithmic mark-up model that is used to develop the various exchange rate pass-through hypotheses. Data composition and sources are detailed in Section 4, and the estimation procedures are presented in Section 5. Section 6 provides a discussion of the test hypotheses, analyses the results, and discusses their broad implications for practitioners. A final section briefly concludes the paper.

This paper investigates the relationship between exchange rate pass-through and the currency invoicing policies of international Norwegian firms when these firms choose between different currencies in which to invoice their exports of fish products. By using a variant of the logarithmic market model (see Hooper & Mann, 1989), we specify a cointegration framework to empirically estimate how the degree of exchange rate pass-through alters in relation to differing invoicing practices of the exporting firms. The dynamic error correction methodology we adopt is much more robust and efficient than single equation approaches to analysing pass-through which are used in the overwhelming majority of previous studies. To our knowledge the product level approach in this paper provides the first empirical estimates relating pass-through to currency invoicing practices available in the literature. There is a clear indication that Norwegian fish exporters possess an ability to set prices, and strong evidence of pricing-to-market behaviour. The degree of pass-through varies by choice of invoicing currency, even for a given product in a given export market destination, implying intra- as well as inter-market price discrimination. This constitutes strong evidence that the choice of invoicing currency is an important strategic pricing decision taken by a firm. The findings corroborate these studies which maintain that exporters effectively price discriminate between markets, and take destination-specific market conditions into consideration when setting prices.23 The results indicate that the pricing-to-market hypothesis is relevant also for exporters from small open economies (Lee, 1997) and suggest that destination market conditions are important in explaining the extent of variation in pricing-to-market practices. Finally, they provide the first evidence in support of theoretical predictions that pass-through is inversely related to exchange rate volatility (Friberg, 1997 and Friberg, 1998).