How exchange rate fluctuations impact imports and domestic prices: Evidence from Rwanda

In recent years, the Rwandan Franc (RWF) has depreciated significantly against a wide basket of currencies: based on year-on-year, the Rwandan Franc depreciated by 5.3% against the US dollar in 2021, higher than the 4.3% recorded in June 2020. With respect to regional currencies, the RWF depreciated against the Kenyan Shilling, Tanzanian Shilling, and Burundian Franc by 4.1%, 5.3%, and 1.7%, respectively in the year ending in June 2020. The RWF shows a similar pattern of depreciation in the previous fiscal years (NBR 2016, 2017, 2018, 2019). The current account deficit, largely driven by the trade deficit, is accounted as the main driver of this nominal depreciation.

Given this recent extended period of depreciation of the RWF, an important question from the policy perspective of the National Bank of Rwanda (BNR) is the extent to which this has affected import activity as well as the domestic activity of firms in Rwanda. Such a nominal depreciation could lead to an increase in import prices in RWF terms, affecting firms’ importing behaviour. This may propagate to the domestic economy through importers’ domestic sales and purchasing decisions.

Data and methodology

In an IGC study, we empirically investigate this by studying the impacts of exchange rate fluctuations on import volume and prices, as well as importers’ domestic activity and prices. We leverage two data sets in Rwanda: (1) the customs data that records the universe of import transactions by Rwandan firms, and (2) the EBM (electronic billing machines) data, which records the item-level firm-to-firm and firm-to-consumer transactions by VAT-registered firms in Rwanda. The latter is a particularly unique dataset that can potentially provide new insights on exchange-rate pass-through. In Rwanda, VAT-registered businesses are mandated to submit itemised receipts of all their sales through EBMs. The itemised price information contained in the EBM receipts can be a potentially useful data source to guide monetary and exchange rate policies. For our purpose, the microdata sheds light on the heterogeneity of firm-level responses to nominal exchange rate shocks, which previous studies relying on macrolevel data have not been able to address.

One important caveat for our analysis is that it is based on customs data and VAT data, implying that the informal sector is not included in the sample. from the samples. Studying how the informal economy responds differently from the formal sector is left for future research.

Main findings

1. Exchange-rate pass-through to import prices

We document an exchange-rate pass-through to import prices of 10% to 40% depending on the price measure used and on whether we use exchange rates based on the country of import origin or invoice currency. Further exploring the richness of the microlevel data, we document various heterogeneity patterns across time periods, sectors, and firms. In particular, we find that small firms and firms importing intermediate goods appear to be affected more (i.e., experience higher pass-through) by exchange rate shocks. The results on high-frequency versus low-frequency importers are mixed, depending on how we define the frequency measure. We also examined whether the impacts of exchange rate fluctuations are different across different periods and did not find evidence on that. Finally, on the extensive margins, we document suggestive evidence that firms reduce the number of products they import and the number of countries they source from in response to currency depreciation, and some degree of substitution may happen across the origin countries.

2. Impacts on importers’ domestic prices, purchases, and sales

We then study importers’ domestic purchases and sales, as well as their domestic prices, using the EBM data. There are two main takeaways from this. Firstly, we find that importers respond by imperfectly substituting for domestic purchases as a response to the exchange rate fluctuations: importers that are more strongly hit by exchange rate shocks reduce import expenditures more, and in compensation, increase domestic purchases more. However, this substitution is imperfect – for a one percentage point reduction of import expenditure, the domestic purchase increases by about 0.3 percentage points. Secondly, the pass-through of exchange rate fluctuations to importers’ buyers is limited. There are no impacts of exchange rate fluctuations on sales prices. This is partly because of the import substitution as indicated above. Thirdly, we find that large firms, and firms in commerce industries, are less affected by the exchange-rate fluctuations. Together, the results suggest that importers act as a “shock absorber” in mediating the pass-through of exchange rate fluctuations to the domestic economy and downstream consumers.

Policy implications

The project deepens the understanding of how the devaluation of the Rwandan Franc affects import and domestic prices. Some argue that the depreciation of the Rwandan Franc has led to the increase of domestic prices, while others argue that there are no substantial impacts on consumer prices because firms can effectively substitute to domestic intermediate goods suppliers. Which of these arguments is closest to reality is ultimately an empirical question. Our results show that the latter is the case.

Beyond the implications of the devaluation, this project also provides evidence of strong substitution by Rwandan firms between domestic and foreign inputs. This finding suggests that there is a large scope for import substitution policies (such as the Made-in-Rwanda Initiative).

Lastly, our analysis provides a first case of effectively using the EBM data set for policy analysis, demonstrating that this type of rich administrative data can be effectively utilised in policymaking.