Central bank currency losses: how to read the headlines
Every year some central bank reports a loss of several billion on its reserves, and every year the number is quoted as if it were a trading result. It almost never is.
Headlines of the form "central bank lost X billion on foreign exchange" circulate regularly, about the Swiss National Bank, the Bank of Japan, the Bank of Russia and others. They are usually accurate as arithmetic and misleading as economics.
What a reserve portfolio is
A central bank holds reserves in foreign assets — mostly government bonds of other countries, plus deposits and gold. The purpose is not profit. It is the capacity to defend the currency, service external obligations and absorb shocks.
Those assets are denominated in dollars, euros, yen, pounds and yuan. The accounts are kept in the domestic currency. So every reporting period the portfolio is translated back — and the exchange rates used have moved.
Revaluation is not trading
This is the crux. If a bank holds euro bonds and the euro falls against the domestic currency, the reported value of those bonds drops. Nothing was bought, nothing was sold, no decision was taken. The number changed because a rate changed.
A revaluation loss is unrealised and reverses when the rate does. Reporting it as a trading loss is like saying you lost money on your house because prices dipped in a year you did not sell.
Headlines rarely distinguish between three very different things: revaluation of holdings, realised results from actual interventions, and the carry earned on the bonds themselves — which is usually positive and usually omitted.
When intervention losses are real
Sometimes a central bank genuinely does lose money defending a peg or slowing a slide. It sells foreign currency to support its own, the pressure continues anyway, and it buys back later at a worse rate.
- Defending a level the market has decided is wrong is expensive, and the record of such defences succeeding is poor.
- Reserves are finite; a determined market is not. That asymmetry is the whole story of several famous currency crises.
- Even then the mandate is stability, not profit. A bank may knowingly accept a loss to avoid a disorderly move.
Why it matters to a trader — and why it mostly does not
Reserve data is published with a long lag and describes positions, not intentions. As a timing input it is useless.
What is worth watching is the direction and pace of change in reserves over months. A central bank steadily spending reserves is under pressure, and pressure that persists tends to resolve in the direction the market was pushing. That is a slow contextual signal, not a trade.
The practical lesson from these headlines is narrower and more useful: when a number is presented without saying whether it is realised or a revaluation, it has not been explained. That applies to central bank reporting and to trading results people show you on the internet.
Where to check rather than guess
- Central banks publish reserve composition and annual reports themselves. The primary source distinguishes revaluation from realised results; press coverage often does not.
- The IMF's COFER data shows the currency composition of global reserves in aggregate, which is more informative than any single bank's headline.
- Where a figure comes from a secondary source with no methodology stated, treat it as a claim rather than a fact.