How Exchange Rates Pass Through to Inflation

Suppose an exchange rate is quoted as units of domestic currency per one unit of foreign currency. A rise in that quote means the domestic currency has weakened: it now takes more domestic currency to buy the same foreign-currency unit. If the quote is written the other way around, the direction reverses, so a report must identify both the currency pair and the quote convention before calling a move an appreciation or depreciation.

Exchange-rate pass-through, often shortened to ERPT, is the estimated response of a specified price measure to a specified exchange-rate change over a specified horizon. The price measure might be an import-price index at the border, a producer-price index, or a consumer-price index. The horizon might be the immediate month, four quarters, or a longer interval. Those choices matter: there is no single pass-through number that applies to every country, product, currency pair, and period.

Pass-through is also different from merely converting an unchanged foreign invoice into domestic currency. That conversion is arithmetic. Pass-through asks how much the prices recorded along a broader pricing chain actually move after firms, distributors, and consumers respond. The European Central Bank (ECB) describes direct effects through imported final consumer goods and indirect effects through imported inputs used in domestic production. [ECB’s overview of exchange-rate pass-through]({source:ecbExchangeRatePassThrough2016})

Start with a deliberately simple calculation. A foreign supplier invoices a product at 100 foreign-currency units. Before the exchange-rate move, assume one foreign unit costs 1.00 domestic-currency units. The invoice is therefore worth 100 domestic units. If the quote rises to 1.10 domestic units per foreign unit and the foreign invoice price stays at 100, the converted border value becomes:

100 foreign units × 1.10 domestic units per foreign unit = 110 domestic units

The converted invoice is 10% higher. This is a hypothetical currency conversion, not a forecast of a retail price or a measured inflation rate. It holds the foreign-currency price fixed and ignores freight, insurance, duties, taxes, hedges, contract timing, and any change in the supplier’s price. If the exchange rate is quoted as foreign units per domestic unit, a weakening domestic currency appears as a fall in that quote instead.

An importer may have contracted earlier, hedged the currency exposure, or bought inventory before the move. The amount paid today can therefore differ from the amount implied by converting a current spot quote. A company may also renegotiate with its supplier or accept a smaller margin to keep a local-currency price stable. The invoice calculation isolates one link in the process; it does not describe every firm’s cash flow or pricing decision.

There are two broad routes. The direct route starts with imported final goods that households buy. If a retailer pays more in domestic currency for a finished imported product, that cost can eventually affect the shelf price and the consumer-price index. It may do so only after the retailer sells existing inventory, reviews its price list, or renegotiates a contract.

The indirect route starts with imported inputs. A local factory may use imported fuel, components, packaging, or software services. A weaker domestic currency can raise those input costs. The factory may pass some of the increase to wholesalers or retailers, who may in turn change prices. This chain can take longer and can affect domestic goods as well as products that are imported in finished form. The ECB calls the direct route first-stage pass-through and the input-cost route second-stage pass-through. [ECB’s explanation of direct and indirect effects]({source:ecbExchangeRatePassThrough2016})

The effect generally becomes weaker and slower as it moves from import prices at the border through producer prices to final consumer prices. Domestic labor, rent, transport, taxes, wholesale and retail services, and other local costs also contribute to a final price. A change in one imported component therefore does not mechanically reset the full price of a product. The ECB’s euro-area review finds a pronounced difference between estimates for import prices and consumer prices, while also documenting variation across sectors and countries. [ECB’s 2020 review of transmission to euro-area inflation]({source:ecbExchangeRateTransmissionInflation2020})

The illustration below summarizes that chain. It is conceptual artwork; the fading cost effect is not a measured coefficient and the scene contains no country, company, or price data.

<!-- learn:illustration -->

Imagine a representative consumer-price basket in which one group of goods has a 5% weight. Assume imported inputs account for 60% of the price of that group. A 10% currency-driven increase in those imported costs occurs, but firms pass through only 40% of that increase to the group’s price during the period being measured.

The group-level price change in this simplified example is 60% × 40% × 10% = 2.4%. Its approximate contribution to the whole basket is then 5% × 2.4% = 0.12%. Equivalently, multiplying all four shares gives 5% × 60% × 40% × 10% = 0.12%. The assumptions are invented to show the arithmetic; they are not estimates for any country or product.

This is a small-basket approximation with fixed weights. It ignores substitution toward other goods, changing quantities, taxes, domestic input costs, margin adjustments, and changes in the basket itself. It also assumes the 40% pass-through applies to the imported-cost component in the stated period. If all of the imported-cost increase were passed through immediately, the same simplified inputs would imply a 0.30% basket-level change instead. If no increase reached the shelf price in that period, the immediate contribution from this route would be zero, even though a later adjustment could still occur.

The 0.12% result describes an illustrative change in the price-index level attributable to one component under the assumptions. It is not a promise that the annual inflation rate will rise by 0.12 percentage points, and it says nothing by itself about how long the effect lasts. Inflation is the rate at which a price index changes. If the index rises once and then stays at its new level, the higher price level remains, but that one-time step does not keep adding to the inflation rate every month.

The currency used to invoice a good is one reason. If a transaction is priced in the buyer’s domestic currency, a short-lived currency move may not change the invoice at all. If it is priced in the exporter’s currency or a third currency, the conversion can matter more immediately. Contracts and price lists are often reset infrequently, so a move can reach prices with a lag. Over a longer horizon, firms may renegotiate even if the original contract was in the buyer’s currency. [ECB research on global value chains and invoicing currency]({source:ecbGlobalValueChainsExchangeRatePassThrough2019})

Inventories and financial hedges change timing too. A retailer can sell goods bought at an earlier exchange rate before paying for replacement stock at the new rate. An importer may have fixed a conversion rate with a forward contract for part of its expected purchases. Those arrangements can delay or reduce the cost change that appears in accounts for a particular month. They do not remove the economic exposure forever: they shift it across purchases, firms, or dates.

Exporters and distributors can adjust markups. A foreign producer that wants to defend market share may lower its foreign-currency price when the buyer’s currency weakens. The domestic importer can also absorb some of the higher border cost in its margin, or pass more of it on when demand is strong or alternatives are limited. Local distribution costs matter because the final consumer price includes services performed in the domestic market. These choices can differ across products even when they face the same exchange-rate move.

Competition, supply chains, and the amount of imported content all affect the result. A good with many substitute suppliers may be priced differently from a specialized input with few alternatives. A domestically assembled product can still depend on imported components, while a product assembled abroad can include inputs from the destination country. The ECB’s analysis of global value chains finds that production links and the currency of invoicing help explain why pass-through differs across sectors and countries. [ECB’s global-value-chain analysis]({source:ecbGlobalValueChainsExchangeRatePassThrough2019})

The composition of imports matters. Energy, food, and other frequently traded goods can respond differently from services or manufactured products. A country that imports a large share of what households consume may have a more direct route from foreign costs to its consumer basket. Imported components used by local producers create additional indirect exposure. Trade openness alone is not enough to predict the final consumer-price response; the type of goods, input links, and local cost shares matter too.

The starting inflation environment and expectations matter as well. The IMF authors’ 2023 working paper examines a large sample of advanced and emerging economies and reports higher pass-through in periods of higher inflation and elevated uncertainty. It also finds different responses depending on the shocks associated with exchange-rate movements. These are research findings for the paper’s sample and method, not a universal rule or a forecast for a particular currency. The authors state that an IMF working paper presents their views and does not necessarily represent the IMF, its Executive Board, or management. [IMF Working Paper 23/86]({source:imfStateDependentExchangeRatePassThrough2023})

Monetary policy can influence whether a temporary cost increase spreads into broader prices and expectations. A credible central bank may help keep longer-term inflation expectations anchored, while firms and workers can behave differently if they expect a currency-driven price increase to persist. This does not mean a central bank can erase the original import bill. Policy affects the broader response through demand, financing conditions, and expectations, and it can also change the exchange rate itself.

A depreciation can affect output as well as prices. It may make exports less expensive to foreign buyers, while raising the cost of imported inputs used by local businesses. The net effect depends on how firms source inputs, how customers respond, and what shock moved the exchange rate. It is not safe to infer that every depreciation raises output, lowers output, or produces the same inflation response.

Empirical estimates differ because researchers may choose different exchange-rate measures, price indexes, countries, time periods, model specifications, controls, and response horizons. A bilateral rate against one trading partner can tell a different story from a trade-weighted nominal effective rate. A border import-price index is not the same outcome as a consumer-price index. Some studies estimate a reduced-form association between exchange rates and prices; structural studies try to identify how prices respond after a particular type of shock.

The ECB’s 2020 review illustrates why the result must be read with its scope. Using consistent reduced-form data for the euro area and EU countries, it reports that a 1% euro depreciation was associated on average with about a 0.30% rise in total import prices within a year and about a 0.04% rise in the headline HICP price index over the same horizon. The paper says the consumer-price estimates are not always statistically distinguishable from zero, and its figures are specific to its data and method. They are not a conversion rule to apply to another country, a current forecast, or a claim that every product’s price moves by those amounts. [ECB’s empirical estimates and definitions]({source:ecbExchangeRateTransmissionInflation2020})

The same review distinguishes a reduced-form exchange-rate pass-through estimate from a price-to-exchange-rate ratio after an identified shock. The second approach tries to account for feedback among prices, the exchange rate, output, and policy. This distinction matters because an exchange rate is not an external dial that changes for only one reason. Interest-rate decisions, global demand, commodity prices, risk sentiment, and domestic news can move it while also affecting prices through other channels.

For that reason, a simple chart showing the currency and CPI moving together does not establish how much of the CPI change the currency caused. To interpret an estimate, check the exact exchange-rate definition, whether a rise means appreciation or depreciation, which price stage is measured, the response horizon, the sample, the model, and the uncertainty interval. Treat the coefficient as a study result under stated assumptions, not as a household inflation calculator.

First identify the exchange-rate quote and measure. A bilateral quote may be expressed as domestic currency per foreign unit or the reverse. A nominal effective exchange-rate index combines multiple trading partners, but the index provider’s convention determines which direction represents a stronger currency. Do not describe the move until that convention is clear.

Next follow the pricing chain rather than jumping straight to the consumer index. Import-price data can show the border response; producer-price data can indicate cost changes for domestic sellers; CPI or HICP measures final consumer prices. These series cover different goods, services, transactions, and stages, and they may be released on different schedules. Look for the currency of invoicing and distinguish contract prices from spot-rate conversions where the data allow it.

Then state the time window and the kind of change. A one-time depreciation can lift the price level as firms adjust, while the inflation rate may rise only during the adjustment period. Repeated depreciation or a broader wage-and-price response can create a different path. A year-over-year measure can also reflect earlier price changes that happened before the latest exchange-rate move.

Finally separate the observed change from the explanation. Identify foreign-currency prices, freight and commodity costs, imported input shares, domestic margins, taxes, demand, and any policy response that could also matter. A defensible summary names the measure, dates, quote convention, price stage, and research horizon. It does not multiply the latest exchange-rate change by one published pass-through estimate and call the result a forecast.

For the exchange-rate measures themselves, see nominal and real effective exchange rates. For consumer-price indexes, see CPI, PCE, and the GDP deflator. The distinction between broad and underlying price measures is covered in headline vs. core inflation.