How Exchange Rates Work: A Ratio, Not a Price

A share price can rise because the company did well. A ratio can move because the numerator changed, because the denominator changed, or because both changed by different amounts — and it cannot tell you which. The previous article established that every position is two positions; this one establishes why every rate is two stories, and what follows for anyone trying to interpret one.

Nominal, real, and effective — three different questions

The number quoted on a screen is the nominal exchange rate: units of one currency per unit of another, and nothing more. It answers "how many marks does a dollar buy today." It does not answer the question most people actually have, which is about purchasing power. The real exchange rate adjusts the nominal rate for the difference in price levels between the two economies, and it answers a different and more useful question: "has a dollar become more or less able to buy things in Meridia?" The two can move in opposite directions. If the dollar buys 4% more marks but Meridian prices have risen 6% over the same period, then a dollar buys more marks and yet each mark buys less — in real terms the dollar's purchasing power in Meridia has fallen by about 2%, so the nominal rate strengthened while the real rate weakened. Anyone reasoning about a rate for a practical purpose — a foreign purchase, a relocation, income in another currency — is asking a real-rate question and reading a nominal-rate number. That mismatch is the commonest confusion about exchange rates and it is worth naming plainly. The effective exchange rate answers a third question: how a currency has moved against a basket of trading partners, weighted by trade volumes. This matters because a currency can strengthen against one counterpart while weakening on average — so a headline about the dollar rising against the mark says nothing reliable about the dollar generally. Effective rates are the closest thing to a single answer about one currency in isolation, which is why they are the exception to the pairs rule the last article set out. Three practical notes on the numbers themselves. The rate in a news report is an indicative mid-market figure — halfway between what buyers and sellers were quoting — and nobody transacts at it, which is why the rate a bank or app offers you is always worse than the one you read. That difference is a spread, and on retail currency conversion it is frequently the largest cost of the transaction while being presented as though the service were free. Rates differ slightly between sources at any instant, because there is no central exchange. And a rate quoted "to four decimals" is not more precise in any meaningful sense than the liquidity underneath it supports.

What determines a rate — and the honest limits of that question

Two frameworks are worth understanding, both taken further in the drivers article. Purchasing power parity holds that in the long run rates should adjust so that identical goods cost the same across countries, because otherwise arbitrage would be available. As a long-run anchor the idea has real content: currencies of high-inflation economies do tend to depreciate against those of low-inflation ones over long periods, which is close to an arithmetic consequence of what inflation is. As a guide to anything shorter, its record is poor — deviations from parity persist for years, because most of what people buy is not tradeable, transport and tariffs and taxes intervene, and capital flows swamp trade flows in the short run. Interest-rate parity holds that the difference between spot and forward rates should reflect the interest-rate differential between the two currencies, because otherwise a riskless profit would exist by borrowing in one currency and lending in the other. This one holds tightly in practice for the covered version — where the future rate is contracted in advance with a forward — because it is enforced by arbitrage. The uncovered version, which would imply that high-interest currencies depreciate enough to cancel their rate advantage, does not reliably hold, and the gap between the two is precisely what the carry trade attempts to harvest and why that trade has its documented crash pattern. Now the honest limit, stated because omitting it would misrepresent the state of knowledge. Short-horizon exchange-rate forecasting has a poor documented record, and this is not a claim about retail participants — it is a long-standing finding in the academic literature, dating from the early 1980s and repeatedly re-examined since, that structural models struggle to beat a simple no-change assumption over short and medium horizons, and institutions with substantial research resources do not reliably forecast rates either. That finding is not universally interpreted as meaning rates are unforecastable in principle, and there is continuing research on longer horizons and on particular conditions. But the practical implication is worth stating: a framework that explains why a rate moved is not the same as a framework that predicts where it will go, and retail FX marketing routinely presents the first as though it were the second. This portal explains the mechanisms so a reader can interpret what they observe. It presents no rate as cheap, dear, or due to move, because it cannot and neither can anyone else with the reliability that a leveraged position would require.