Why Exchange-Rate Pass-Through Changes with the Regime
This series examines exchange-rate pass-through: how a movement in a currency reaches consumer prices and inflation. Its central question is not simply how large the pass-through rate is, but why the answer changes across countries, time periods, policy frameworks, markets, and parts of the consumer basket. That distinction supports a practical judgment: when someone quotes a pass-through estimate, the reader should ask which exchange-rate movement, which prices, which horizon, and which economic state produced it.
The question matters whenever a currency moves sharply. A depreciation can make imported fuel, food, machinery, or intermediate inputs more expensive in local currency. An appreciation can reduce those costs. But the change in an importer's costs is not automatically the change in the final price paid by households. Firms may absorb part of the movement in their margins, adjust prices with a delay, change suppliers, or respond differently when customers are already expecting inflation. Domestic demand, wages, and competition can amplify or offset the initial cost change. The exchange rate is therefore a transmission mechanism, not a switch that converts a percentage movement into an identical percentage change in the consumer price index.
The number is not the mechanism
“Pass-through” sounds like a single proportion: a 1 percent exchange-rate movement produces, perhaps, a 10 percent or 10 percentage-point price response. That shorthand is useful only after several choices have been made. An analyst might measure the response of import prices, producer prices, headline consumer prices, core prices, goods, or services. The response might be observed immediately, over one year, or over a longer period. It might describe an average movement, an unexpected policy shock, or an exchange-rate change that occurred because policymakers were reacting to expected inflation.
Those choices change the object being measured. Import prices are close to the first border crossing in the transmission chain. Consumer prices are farther downstream: they include domestic distribution, taxes, wages, rents, services, and firms' pricing decisions. Goods often incorporate imported costs more quickly than services, whose prices depend more heavily on domestic labor and capacity. A short-horizon estimate can therefore be small even when a later response is larger, while a headline estimate can conceal very different movements inside its components.
The direction also requires care. In ordinary language, a weaker domestic currency tends to raise the local-currency cost of imports, while a stronger currency tends to lower it. Statistical studies may define the exchange-rate index in a way that reverses the sign. For that reason, a credible statement about pass-through must identify the convention and, when necessary, describe the magnitude and direction in words rather than detach a coefficient from its definition.
The useful question is thus conditional: under what circumstances do firms treat an exchange-rate movement as a cost that must be reflected in prices, and under what circumstances do they wait, absorb it, or offset it elsewhere? Inflation itself changes the environment. When prices are generally stable, a firm may regard a currency movement as temporary or unusual. When inflation is already high, changing prices may be more frequent, and an exchange-rate movement may fit an established pattern of rising costs. The same nominal movement can then meet a different pricing system.
The diagram separates the movement itself from the conditions that govern its journey. The imported-cost channel is intuitive, but it is not the only route. An exchange-rate change can alter demand for domestic production and labor, affecting the amount of slack in the economy and the pressure on prices. Meanwhile, inflation, volatility, market structure, credibility, and labor conditions affect how strongly and how quickly the two channels appear in the consumer-price data. They are not interchangeable explanations; each describes a different part of the transmission environment.
Two views of the same transmission problem
The broad evidence in the Bank for International Settlements study asks why pass-through differs across countries and periods. It examines 98 economies over four decades and first estimates country-specific, period-specific relationships between nominal effective exchange rates and consumer-price inflation. The purpose of retaining those differences is important. If every country were forced to share one coefficient, the analysis could report an average but could not explain why the average fails in a particular economy or decade.
The BIS study then relates the estimated differences to macroeconomic and structural conditions. Its approach is useful for this series because it does not assume in advance that every relationship is linear. It finds that the size of the economy and the level of inflation are especially strongly associated with the estimated pass-through, followed by product homogeneity and exchange-rate volatility. The study also examines policy-related conditions, including inflation gaps, exchange-rate arrangements, and fiscal health. These results do not turn each factor into a proven cause. They establish a map of where the response tends to be different and invite a closer explanation of the mechanisms behind that map.
The second lens is narrower but more institutional: Singapore. The Monetary Authority of Singapore manages the nominal effective exchange rate within a basket, band, and crawl framework. In this setting, the exchange rate is not merely a market price that happens to affect inflation; it is the principal monetary-policy instrument. The framework is designed around expected inflation and the output gap, so a change in the Singapore dollar can be both a force acting on prices and a policy response to information about future prices and growth.
That dual role creates a problem for interpretation. Suppose the monetary authority appreciates the currency when it expects inflationary pressure to rise. A simple correlation may show that appreciation coincides with higher inflation, making the exchange rate appear less powerful than it really is. Alternatively, a movement may reflect conditions that also affect prices directly. To learn how the exchange rate transmits, the analyst must distinguish an unexpected policy-induced movement from a movement that was already responding to the outlook. Singapore makes this issue visible because the policy instrument and the transmission variable are closely connected.
The Singapore case also clarifies that “consumer prices” are not a single mechanism. An appreciation can lower the domestic-currency price of imported goods and services. It can also reduce demand for domestic factors of production, narrowing a positive output gap and easing price pressure. Those channels need not have the same timing. Goods can respond as imported costs change; services may respond more slowly because wages, capacity, and local demand matter more. A national pass-through number can therefore be accurate as an average while still being a poor guide to what happens in a particular sector.
What changes when the state changes
The word “regime” should not be reduced to a binary choice between a fixed and a floating exchange rate. It is a shorthand for the surrounding state in which a movement is interpreted and transmitted. A currency that rarely moves may be anchored by a credible arrangement, making an eventual adjustment look unusually persistent. A currency that moves constantly may signal weak fundamentals and repeated cost pressure. A currency in a successful float may move enough to matter but also enough that firms expect some reversals. Those situations can produce different pricing decisions even when the measured volatility is similar over a short window.
Policy credibility is part of that state. If inflation remains close to a central bank's target and the broader monetary-fiscal framework is trusted, firms may expect a temporary exchange-rate movement to have limited persistence in overall inflation. If inflation is already far from target, or if large fiscal deficits raise doubts about future price stability, the same movement may be incorporated more quickly into prices. This is not a claim that a central bank can choose a pass-through rate at will. It is a reason to treat credibility and fiscal conditions as part of the environment that shapes private expectations.
Labor-market conditions add another layer. Tight labor markets can sustain domestic demand and wage pressure even when an exchange-rate appreciation lowers imported costs. They can also change the relative response of goods and services. In Singapore, the evidence points to a weaker exchange-rate-to-inflation transmission under tight labor-market conditions, especially for services. That finding does not mean labor tightness is the sole cause, nor does it mean appreciation ceases to matter. It means that the response is conditional: the same policy movement can have less pricing force when domestic conditions generate their own pressure or alter the timing of adjustment.
The evidence in this series therefore supports a disciplined form of skepticism. A quoted coefficient is not meaningless, but it is incomplete until its scope is restored. Ask whether the movement was a market change or an identified shock. Ask whether the endpoint is an import price or a consumer price, and whether goods and services are being combined. Ask whether the horizon captures the relevant adjustment. Then ask what was happening to inflation, volatility, market structure, fiscal credibility, and labor conditions.
The practical payoff is greater than a more cautious vocabulary. If pass-through is a regime-dependent function, policymakers cannot infer the effect of a new exchange-rate movement from an old average alone. Businesses cannot assume that a historical currency cost will reappear unchanged in their prices. Readers cannot compare a global one-year CPI estimate with a local, shock-based response over nine months as if the two numbers measured the same thing. The next step is to examine how these estimates are constructed and why their timing and scale differ. That is where the apparently simple question—how much of an exchange-rate movement reaches prices?—becomes an empirical question about a particular economy in a particular state.