Writing Startup Finance
Revenue in One Currency, Costs in Another: Currency Exposure at a Startup
A startup that sells in dollars and pays its team in euros or lira has a currency position, whether or not anyone chose it. When the rate moves, burn moves with it, and the board pack calls it higher costs. Seeing the position clearly, deciding which part of it to keep and modelling it honestly takes an afternoon; skipping it shows up as months of runway nobody spent.
A founder once showed me a month in which burn had jumped by almost a fifth. Nobody had been hired, no contract had been signed, no campaign had launched. Same team, same work, same salaries. The only thing that had changed was the exchange rate between the currency the company billed in and the currency it paid its people in. The board pack explained the jump as "higher operating costs", which was true in the reporting currency and false in every other sense.
Most early-stage companies are exposed to currency movements long before anyone in the company thinks of it as a finance question. A startup that sells to American customers in dollars and employs an engineering team in Europe or Turkey has a currency position, whether or not it ever chose one. This post is about seeing that position clearly, deciding which part of it to keep, and making sure the runway model and the board pack say what is actually happening.
A startup's currency position is set by accident
Nobody sits down at incorporation and decides how much euro or lira risk to carry. The position falls out of three decisions made for other reasons: where the customers are and what currency the price list uses, where the team is and what currency the contracts use, and where the fundraising happened and what currency the money arrived in. Each is sensible alone. Together they decide how much of the company's runway depends on a rate nobody in the company controls.
The exposure has three layers, each managed differently.
The first is structural exposure: the mismatch between the currency of revenue and the currency of costs. If revenue is in dollars and costs are in euros, a stronger euro makes every month less profitable until something changes. This is the layer that moves runway.
The second is cash exposure: the currency the bank balance sits in, compared with the currency it will be spent in. A company that raised dollars and holds them in dollars, but will spend most of them on euro salaries, is carrying an open position the size of its runway.
The third is transaction exposure: the gap between the day an invoice is issued or received and the day it is paid. A customer invoice in dollars, on sixty-day terms, is worth a slightly different amount in euros on the day the cash lands. It is usually the smallest layer, and the one that gets the most attention because it has its own line in the accounts.
What a ten percent move does to runway
Take a company with $100,000 of monthly revenue, all billed in dollars, and €180,000 of monthly costs, almost all of it payroll and contractors in Europe. It reports in dollars and holds $1.5 million in a dollar account.
At a rate of 1.08 dollars per euro, monthly costs are $194,400 and net burn is $94,400. Cash covers a little under sixteen months.
Now the euro strengthens to 1.18, a move of about nine percent, which is not an unusual year for that pair. Costs become $212,400, net burn becomes $112,400, and runway falls to about thirteen months. The company has lost more than two and a half months of runway without spending a single extra euro. Burn rose by nineteen percent while costs, in the currency they are actually paid in, did not move at all.
The effect is leveraged because burn is a small number sitting between two large ones. A nine percent change in costs became a nineteen percent change in burn because revenue covered about half of costs. The closer a company gets to breakeven, the more violent this leverage becomes: a business that is nearly profitable in one currency can swing to a meaningful loss on a rate move alone, which is exactly the point at which founders start telling investors they are close to default alive.
Currency risk does not show up as a separate line. It shows up as a burn rate that changed for no operational reason, and a runway that is shorter than the plan everyone approved.
The first hedge is where the cash sits
Before anyone talks about forwards or options, the cheapest and most effective decision is the currency of the bank balance. Cash should be held, broadly, in the currency it will be spent in. If the next twelve months of net outflows are mostly in euros, a large share of the runway should be in euros, converted at a time and in amounts the company chooses rather than one salary run at a time at whatever the rate happens to be that week.
A workable rule is to estimate net outflows by currency for the next twelve to eighteen months, using the same model that drives runway, and hold roughly that mix. Revenue that arrives in dollars and is spent in dollars nets out; only the shortfall in each currency needs funding. The rest of the balance can stay in the currency it was raised in, which is usually the currency the next round will be priced in.
The caveat: more currencies usually means more accounts, and the controls in the minimum viable control environment must cover every one of them.
The price list is a currency decision
The currency a company invoices in decides who carries the exchange risk on every sale. Pricing in dollars to a European customer moves the risk onto the customer, which they will notice when the rate moves against them and the renewal comes up. Pricing in the customer's currency keeps the customer comfortable and moves the risk back onto the company.
Neither is right in general. The useful question is whether the pricing currency can be aligned with the cost base. A company whose team is in Europe and whose customers are partly in Europe gains a natural hedge by selling to those customers in euros: each euro of revenue cancels a euro of cost. A company that sells entirely in dollars and pays entirely in euros has no natural hedge at all, and every pricing decision should be made knowing that.
For consumer apps, the store sets the local prices and the developer receives proceeds in a currency of its choosing, with conversion and timing differences that I walked through in App Store Proceeds Are Not Revenue. The exposure is still there; it is just converted for you, at a rate you do not pick.
Build the model in native currency
The most common modelling mistake is converting everything to the reporting currency on the way into the model and then forecasting in that currency. Once salaries are entered as dollars, there is no way to ask what happens if the euro moves, because the euro is no longer in the model.
The fix is worth the extra columns. Every line carries its native currency, conversion happens in one row of assumed monthly rates, and changing one rate cell flows through burn, runway and cash by currency at once.
A currency sensitivity then takes two minutes. Where more than about a fifth of costs or revenue is in a second currency, I would put it in every board pack: runway at the plan rate, and with the rate moved ten percent against the company. It belongs next to the scenarios described in Runway Isn't a Number, It's a Model, because it is one of the few scenarios where the size of the shock is not a matter of opinion.
Plan rates also need a policy. The simplest is to set the rate once a year, at budget time, and hold it fixed so that the variance between plan and actual separates operating performance from currency. Updating it monthly hides exactly what the comparison should show. When the forecast is scored against actuals, as in Forecast Error Is Data, the currency effect should be its own line, so that a good forecast is not blamed for a rate it never claimed to predict.
Report growth and burn in constant currency
When revenue or costs span currencies, the reported growth rate mixes two things: how much the business sold, and how the rates moved. A company billing partly in euros can show twenty percent revenue growth in a year when it sold only twelve percent more, simply because the euro strengthened. It can equally show flat growth in a year of real progress.
The board needs both numbers. Report revenue growth and burn at actual rates, because that is what the bank sees, and alongside it at constant rates, using last year's or the plan rate, because that is what the team did. The difference is the currency effect, and it should be stated in one sentence rather than buried in the commentary.
In the monthly close, the realised gains and losses on settled invoices and the revaluation of open balances should be booked consistently, using one rate source and one timing convention. The five-day close has room for this as one step on the second or third day: pull the month-end rates, revalue the foreign-currency balances, and post the difference to a single FX line that someone actually reviews.
When financial instruments make sense
Banks will happily sell a startup forwards, options and structured products. Most early-stage companies need none of them, and the exotic ones carry risks harder to understand than the exposure they remove.
The one instrument that often does make sense is a simple forward for a known, dated, material payment: a large annual licence, a deposit on an office, a lump of salaries in a month when the company knows exactly how much it will owe. A forward fixes the rate for that amount on that date. It removes uncertainty, not cost, and it is only useful when both the amount and the date are reasonably certain. Hedging forecast revenue that has not yet been sold is a different matter: if the sale slips, the company is left holding a contract for money it does not have.
Avoid anything with a margin requirement: it can pull collateral out of the company exactly when the rate has moved sharply, which matters more to a company counting months of runway than the hedge does. And take the decision through the board, with the bank and the accountant, not in a call with a salesperson.
High-inflation cost bases are a special case
Many startups run their team in a country whose currency loses value quickly, while billing in dollars or euros. On paper this looks like a gift: the local currency weakens, costs fall in the reporting currency, and burn improves. For a while, it does.
The gift is temporary. Salaries in a high-inflation economy are renegotiated more often and by more, and the good people leave if they are not. The usual pattern is a slow improvement in burn for several months followed by a sharp step up when salaries are reset, sometimes for the whole team at once. A runway model that extrapolates the improving months will be wrong at precisely the point the reset arrives.
The honest treatment is to model salary resets explicitly, on the calendar the team actually expects, and to describe any burn improvement from currency as currency, not efficiency. The hiring plan should carry local-currency salaries and an assumed reset schedule rather than a single converted figure. This is also where the cash rule needs judgment: holding a few months of local payroll in the weakening currency is matching, holding a year of it is a loss in waiting.
Currency is a decision, even when nobody made it
A startup does not need a treasury function. It needs to know its position, choose the part of it it is willing to carry, and stop letting a rate movement masquerade as an operating result. In practice that means holding cash in the currency it will be spent in, choosing pricing currencies with the cost base in mind, building the model in native currency, reporting constant-currency numbers alongside actuals, and writing down a short policy before the first large move rather than after it.
None of this is complicated, and most of it takes an afternoon. The cost of skipping it is usually discovered in a board meeting, as a few months of runway that went missing without anyone spending them. If your revenue and costs live in different currencies and you would like the model, the policy and the board view set up properly, this is a natural part of a fractional CFO engagement.