Manufacturing Costs

The yen weakened for years. A dollar buyer still faced an 11.68% invoice jump in four weeks.

A fixed yen invoice rose 11.68% in USD over July 8 to August 5, 2024; the September 9 eligible archive ends September 4, 2026 and contains 1,274 rolling 20-observation windows. The 95th percentile increase is 4.74%.

Editorial evidence cutoff: September 9, 2026. Published September 29, 2026. Observation periods are stated throughout; older figures are retrospective evidence.

A Japanese equipment purchase can look cheaper over five years and become sharply more expensive before the next payment is due. Between July 8 and August 5, 2024, the dollar cost of an unchanged yen invoice rose 11.68%. Those endpoints were 20 observations apart. No increase in the yen selling price was needed to produce the larger dollar bill.

The longer history points the other way. Between July 6, 2021 and September 4, 2026, the same fixed yen amount became 29.11% cheaper in dollars. Both observations belong to the same record. The purchasing question is which stretch of that record matches the exposure being accepted.

Looking back at the 2024 shock helps a September 2026 equipment buyer distinguish a favorable long currency trend from the exposure on an unpaid balance.

The invoice has a clock of its own

A machine buyer does not necessarily acquire every yen on the date a capital request is approved. The deposit, progress payment and final balance can fall on different dates. If those amounts remain denominated in yen, the buyer is exposed to the exchange rate when the currency is acquired, unless the exposure has already been fixed through the agreement or another arrangement.

That timing distinction changes the relevance of a long-run currency chart. A falling dollar cost over several years can help explain why a project looks attractive. It cannot establish what an unpaid balance will cost four weeks from now. Looking only at annual endpoints removes the very movement that can matter between a signed purchase order and settlement.

There is no evidence here that an actual manufacturer suffered the measured loss. The calculation holds the foreign invoice constant to isolate the exchange-rate component. Supplier repricing, dollar invoicing and existing currency arrangements can produce very different outcomes.

Reconstruct the 11.68% increase

The yen series is quoted in yen per dollar. If a dollar buys fewer yen at settlement, an unchanged yen invoice requires more dollars. The relevant multiplier is the starting exchange rate divided by the ending exchange rate. Taking a percentage change in the quoted yen-per-dollar number without reversing that relationship gives the wrong answer for the invoice.

Applied to July 8 to August 5, 2024, that calculation produces an 11.68% increase. A hypothetical balance worth $100,000 at the first observed rate would therefore have cost about $111,685 at the second, before conversion charges. This is a scale illustration, not a reported equipment transaction or an estimate of a buyer's realized loss.

The episode is especially useful because it does not require a story about where the yen went next. The outstanding balance had an exposure during that interval. A later reversal would not retroactively reduce a payment already made.

One dramatic episode is not the whole distribution

The archive contains 1,294 daily observations for each of five currencies. There are 1,274 ways to compare an observation with the one 20 observations later. For yen invoices, the 95th percentile increase across those windows was 4.74%; the 5th percentile was a 4.87% decline.

The same fixed-foreign-invoice calculation produced 95th percentile increases of 3.49% for euro invoices, 1.73% for yuan, 3.99% for Mexican pesos and 2.61% for Canadian dollars. These are comparable horizons within the saved sample. They do not imply that every transaction is equally exposed, or that one sourcing country is intrinsically safer than another.

The distribution also cannot be converted into a promise that a particular increase has a five-percent chance of being exceeded. The windows overlap, the exchange-rate environment changes, and the record is short relative to a currency's full history. It is historical stress evidence, not a loss-probability model.

Try to make the result disappear

Overlapping windows can make one volatile episode appear repeatedly. To check that dependence, we retained only every twentieth starting observation. The resulting 64 nonoverlapping intervals had a 4.42% 95th percentile increase. Their largest increase was 6.24%, between July 19 and August 16, 2024.

The smaller maximum is an honest reminder that the precise worst case depends on the chosen start date. It does not remove the underlying short-horizon exposure. Restricting the analysis to windows starting in 2024 or later also leaves the finding intact: among 652 overlapping windows, the 95th percentile increase was 4.67%, and the maximum remained 11.68%.

These checks support a narrower conclusion than a currency forecast. The long favorable trend did not prevent substantial increases during plausible payment windows. They do not tell a buyer whether tomorrow is a good day to convert currency.

The commercial terms can overwhelm the chart

A purchase contract may eliminate the buyer's direct yen exposure by stating a fixed dollar amount. That does not prove currency risk has vanished from the economics. The supplier may price the uncertainty into the quote, restrict validity, or seek adjustments elsewhere. Conversely, a yen quote might be attractive enough that accepting some exposure is a deliberate decision.

The useful comparison therefore includes the currency of each payment, the date its amount becomes fixed, the remaining balance and any conversion costs. Comparing only today's converted machine prices misses those differences. Two nominally similar quotes can allocate exchange-rate changes to different parties and over different periods.

A manufacturer also needs to separate operational value from financing and currency choices. Machine capability, expected utilization, installation and service still determine whether the equipment earns its cost. A favorable currency conversion can improve that case; it cannot supply missing production benefits.

Put the payment schedule beside the capital request

The practical improvement is small: show the unpaid foreign-currency balances and their dates alongside the project economics. Then calculate how a specified adverse exchange-rate move would change each remaining dollar payment. The historical 20-observation distribution provides transparent context for choosing scenarios, without pretending it identifies an optimal contingency.

The capital equipment payback calculator can help examine the project after those scenarios are translated into purchase costs. The yen history supplies the rate context. Neither replaces the actual agreement.

The central lesson is about the time window. Years of currency relief and several weeks of expensive settlement can coexist. A purchasing analysis becomes more useful when it preserves both facts instead of allowing the larger chart to erase the smaller obligation.

Sources and calculation

Calculations use the saved Federal Reserve/FRED yen series and corresponding EUR, CNY, MXN and CAD histories, July 6, 2021 to September 4, 2026. Returns span 20 observation intervals, not a fixed number of calendar days. Quantiles use linear interpolation; no missing day is filled. Dollar examples assume a fixed foreign invoice and exclude fees.

Sources and evidence

Evidence period: July 6, 2021 to September 4, 2026; 2024 episode retrospective. The frozen evidence record lists the source files and verified hashes available September 9, 2026. Source revision: 5e4fb7726c3d40060c0151c086c903baae856cab. Later live-data updates do not alter the historical evidence in this article.

fred.stlouisfed.org/series/DEXJPUS

fred.stlouisfed.org/series/DEXUSEU

fred.stlouisfed.org/series/DEXCHUS

fred.stlouisfed.org/series/DEXMXUS

fred.stlouisfed.org/series/DEXCAUS

Published 2026-09-29.