Abstract: Using unique Japanese firm-level production network data combined with international trade data, we investigate the upstream/downstream propagation effects of exchange rate shocks on the performance of indirect exporters/importers. We construct firm-specific export and import effective exchange rates to account for the variations of exchange rate exposure across trading firms. We find significant and robust responses in sales and profitability of indirect exporters to exchange rate shocks of downstream exporting firms, suggesting the upstream propagation effect of exchange rate shocks.
We use Japanese micro data on exporters and their suppliers (indirect exporters), and importers and their buyers (indirect importers) to study the direct and propagation effects of exchange rate shocks on firm performance. Therefore, it allows us to estimate the effects of exchange rate changes on the sales and profitability of direct and indirect exporters/importers, and to investigate how exchange rate shocks propagate along domestic supply chains. Recent studies have linked exchange rate pass-through or elasticity to firm-level characteristics.
Third, we show that the spillover effects of exchange rate shocks are asymmetric across firms along various dimensions such as trading mode (direct vs. indirect), firm size (large vs. small), industry affiliation (manufacturing vs. wholesalers), and firm advantage (first tier vs. second level).
Construction of firm-specific effective exchange rates
Firms in the Kikatsu data show significant variations in export and import intensity, that is, the share of exports to total sales and the share of imports to total procurement. This suggests that the impact of a given exchange rate shock may differ significantly between firms and the indirect exporters and importers. To capture the variation of firms' exchange rate exposure on the export and import market, we define export (import) exchange rate exposure as the interaction of the firm's export (import) share and its regional export-weighted (import-weighted) effective exchange rate changes , in other words.
The export share is exports' share of total sales, and the import share is imports' share of total purchases. Firm-specific effective exchange rates provide significant variation across firms that can be used to identify the impact of exchange rate shocks for both exporters (importers) and indirect exporters (importers).
3 Empirical analysis
Specifications
EXEERft is the change in export real exchange rate and IMEERft is the change in import real exchange rate defined in section 2.3. In equation (2), exshare is defined as the share of exports to sales, and imshare is the share of imports to the total input of intermediate products. The exchange rate exposure IMEERftimshareft1 and EXEERftexshareft1 reflect the impact of exchange rate changes on firm performance through import cost and export price channels, respectively.
The coefficients 1 and 2 capture the responses of firm sales and profitability to fluctuations in real exchange rates and are the key parameters to be estimated. 1 is expected to be negative and 2 is expected to be positive if the Japanese yen had a depreciation. In terms of control variables, firm TFP estimated by the Levinsohn and Petrin (2003) approach is included to control for firm-specific productivity shocks.９ We also include firm-level regional trade (exports and imports) weighted real oDP changes defined as .
９ The results remain strong when we control for registry labor productivity (value added per employee). As these are long panel data, we include industry-year fixed effects to control for time-varying industry-specific factors such as input composition and prices. Direct and indirect exporter/importer level - in domestic production chains, a firm's sales and profitability depend not only on shocks and direct exchange rate exposure, but also on exchange rate movements in its firms in upstream and downstream.
Therefore, estimation results at the direct exporter level are likely to be biased by omitted variables. To account for direct and spillover/spillover effects, we need to examine the effects of the exchange rate on sales and profitability at the buyer and supplier levels using matched TSR-Kikatsu data. Since we use the average of the firm characteristics of upstream and downstream firms, the number of observations is significantly reduced compared to the original linkage ratio data.
Results
Recalling the definition of the upward and downward propagation effects in Figure 3, the results in Table 3 confirm our predictions. First, changes in the export exchange rate and its interaction terms with the lagged share of firms' exports are positively related to sales and profitability growth with statistical significance. On the other hand, firms' direct exposure to the import exchange rate is negatively related to profitability growth, but not significant.
The degree of sales and profitability responses to exchange rate changes depends on export and import intensity. These results are consistent with those in Table 2, the cases of direct exporters and importers. More importantly, we find that firm sales and profitability respond differently to upstream exchange rate shocks and downstream exchange rate shocks.
Companies' sales and profitability responses are not significant for import effective exchange rate exposure (import cost channel) from upstream suppliers. The depreciation or appreciation of yen against foreign currency in upstream importers is unlikely to affect firms' sales and profitability. On the contrary, we observe that the export exchange rate exposure of downstream customers is positively associated with firms' sales and profitability growth with high statistical significance.
The effects are greater if a company's downstream direct exporters have a larger export share on average (export price channel). This implies that indirect exporters are likely to increase sales in response to the depreciation of the yen against foreign currencies among downstream exporters, even after controlling for their TFP and demand shocks. This new and direct evidence shows that the propagation effect of exchange rate shocks occurs through domestic production chains, especially from direct exporters downstream to indirect exporters upstream (upstream propagation effect) in the case of Japan.
4 Discussion and robustness checks
- Pure indirect exporters/importers
- Firm size
- Including small firms in TSR data
- Manufacturing vs wholesalers
- Second degree propagation effects
In this case, we expect their sales and profitability to be more responsive to the exchange rate exposure of their downstream customers/exporters. The results in columns 1–2 show that large firms tend to respond to their direct export exposure to the exchange rate, and their sales and profitability improved significantly with yen depreciation, especially for export-intensive firms. Large firms also respond to downstream customer export exchange rate shocks in terms of sales but not profitability.
The coefficient of direct (downstream) export exchange rate exposure is at one percent significance level. Same as large firms, the SMEs' responses of sales and profitability to upstream import exchange rate shocks are not significant. On average, the upstream propagation effect of downstream exchange rate exposure on both sales and profitability is smaller than direct fixed exchange rate.
Similar to previous results, both large firms and SMEs respond to export exchange rate exposure of downstream customers rather than to import exchange rate exposure of upstream suppliers. The upstream spillover effect is larger for SMEs than large firms in terms of the magnitude of the coefficient of downstream exchange rate exposures. These results make sense as firms respond to direct exchange rate exposure rather than their downstream customers.
On the import side, firms also respond to import exchange rate exposure and direct import exchange rate exposure. Compared to large companies, SMEs are more responsive to lower exchange rate exposure. The results show that manufacturing firms respond strongly to exchange rate exposure both at the beginning and at the end of the supply chain when both their suppliers and customers are producers.
5 Conclusion
Comparing their coefficients, we can see that the magnitude of the first-order effect is larger than the second-order effect in all columns. It suggests that the upstream propagation effects gradually diminish as they spread through production chains. On the other hand, the downstream propagation effect from the upstream import side is not very obvious.
Although we can observe the first-degree effect, the second-degree effect is not statistically significant.
Evidence from China, Journal of International Economics International trade and domestic production networks, RIETI Discussion Paper 17-E-116. Wei (2014) Offshore Value Chains and Effective Exchange Rates at the Country-Sector Level, NBER Working Paper 20236. The bilateral exchange rate (local currency/US dollar), oDP measured at constant price, and consumer price indices (CPI) are obtained from the PWT9 database.
Note: TSR-Kikatsu 1 data contains exporters/importers (identified by export/import information in Kikatsu data) and their suppliers/buyers in the manufacturing and wholesale sectors in TSR data. Note: Firm-level export exchange rate exposure EXEERft exshareft1 equals the change in the export real exchange rate and the firm's export share (exports/sales). Firm-level import exchange rate exposure IMEERft imshareft1 equals the change in the import real exchange rate and the firm's import share (imports/total resource).
Note: TSR data refers to the original TSR data in 2013, which contains the trade status of companies, which is categorized into export, import, export and import, and domestic (non-export and non-import) companies. The TSR-Kikatsu 1 match contains exporters/importers (identified by export/import information in Kikatsu data) and their suppliers/buyers in the manufacturing and wholesale sectors in TSR data. TSR-Kikatsu 2 matching is limited to firms that have information on buyer-supplier linkages in TSR data and firm-level variables (including export/import information) reported in Kikatsu data.
