The Balance-Sheet Cost of FX Intervention

When a central bank buys foreign currency to weaken its own currency or builds reserves, it incurs costs that appear on its balance sheet. These costs include losses from holding reserves that fall in value, the gap between low interest rates earned on reserves and higher rates paid to sterilize the purchase, and mismatches between the currency composition of assets and liabilities. Unlike the short-term price effect of intervention, these balance-sheet costs accumulate over time and can eventually undermine the central bank’s credibility or force taxpayer recapitalization.

The basic mechanism: buying foreign currency creates immediate exposure

When a central bank intervenes in currency markets, it typically buys foreign currency (often dollars) and sells its own. This transaction appears on the balance sheet as:

Assets: Foreign currency reserves (denominated in dollars or euros)
Liabilities: Additional domestic money base (the cash created to fund the purchase)

If the central bank intervenes to buy $1 billion of US dollars when its own currency is worth 100 units per dollar, it records:

• Asset: $1 billion
• Liability: 100 billion units of domestic currency

The moment the domestic currency strengthens—say to 95 units per dollar—the same $1 billion is worth 95 billion units instead of 100 billion. The central bank has a 5 billion unit valuation loss on its foreign-currency assets, with no offsetting gain because it printed the home currency at cost.

How valuation losses accumulate

Valuation losses grow whenever the currency strengthens after a central bank has accumulated reserves. A central bank that spent years building a $300 billion reserve cushion during periods of currency weakness now holds assets that decline in value if the currency later strengthens.

Consider a central bank in a carry-trade hub that intervened to cap currency appreciation for a decade, buying dollars and other foreign currency as hot money flowed in. If the currency finally does strengthen—perhaps due to genuine economic improvement or a shift in capital flows—all those accumulated reserves lose value. A 10% appreciation over the prior intervention period translates to a 10% loss on reserves, potentially billions of dollars.

The loss is realized (converted from unrealized to realized) when the central bank sells those reserves, or when it’s forced to acknowledge the loss. For central banks that hold reserves to maturity, the loss may remain unrealized on the balance sheet for years, masked by historical-cost accounting. But the economic loss is real.

Sterilization and the interest-rate cost

A central bank that buys foreign currency and simply lets the money base expand risks inflation. To prevent this, it sterilizes the intervention by issuing domestic bonds or raising reserve requirements to absorb the newly printed cash.

Sterilization has a cost. If the central bank:

• Issues 100 billion units of bonds at 4% interest (to sterilize the cash created)
• And earns only 0.5% on its $1 billion in US Treasury reserves

It pays 4 billion units in interest annually while earning only 5 million units on its reserves (assuming simple numbers). The sterilization cost is the gap: roughly 3.995 billion units per year.

Over a decade of intervention, this differential compounds. If a central bank spends years sterilizing large interventions, it can accumulate billions in cumulative interest losses. These losses shrink the central bank’s net worth on the balance sheet and, if large enough, force the government to recapitalize it.

Currency mismatch and balance-sheet risk

Most central banks hold reserves in major currencies—US dollars, euros, yen. But their liabilities (the base money they’ve created) are in their domestic currency. This currency mismatch creates balance-sheet risk.

If the domestic currency weakens sharply against the dollar—the very opposite of what the intervention was designed to achieve—the dollar-denominated assets on the balance sheet become more valuable (in domestic-currency terms), but the domestic liabilities remain unchanged. Conversely, if the domestic currency strengthens, assets lose value while liabilities shrink. The central bank’s capital is squeezed.

A central bank that has intervened to resist currency strength, accumulated large dollar reserves, and then faces unexpected currency weakness will see a gain on reserves (good), but this often signals that its intervention failed or that the economic shock that weakened the currency is severe. The balance-sheet gain is cold comfort in a recession.

Who pays for the losses?

Central bank losses don’t trigger immediate cash payments (if the losses are unrealized). But they do erode the bank’s capital and ability to absorb future shocks. Eventually, if losses mount:

• The central bank’s capital ratio falls, raising questions about whether it can remain independent.
• The government may be pressured to recapitalize the central bank by injecting fiscal resources.
• If the central bank cannot recapitalize, it may be forced to accept higher inflation or lose credibility.

In practice, taxpayers bear the cost. Large intervention losses at the Bank of Japan, for example, have been absorbed by the government. Similarly, central banks in emerging markets that pursued sustained currency interventions during the 2000s–2010s accumulated massive losses, some of which necessitated government transfers decades later.

Real-world examples of intervention costs

Switzerland (1970s–1990s): The Swiss National Bank accumulated substantial losses from its attempts to contain the franc’s appreciation. These losses, although eventually erased, required decades of patience and questioned the bank’s independence at times.

Japan (2003–2004): The Bank of Japan intervened heavily to weaken the yen, buying over $200 billion in dollars and other currencies. As the yen later strengthened, unrealized losses mounted. By 2007, the cumulative losses from interventions in the prior 15 years were estimated at billions of dollars.

Emerging markets (2005–2008 and post-2020): Central banks in Brazil, Russia, India, China, and Korea accumulated large reserve buffers partly through intervention. When their currencies weakened during crises, the dollar and euro value of those reserves rose (a gain), but the deterioration signaled deeper problems. Some of these central banks later faced questions about whether their intervention losses contributed to fiscal imbalances.

The inflation cost: indirect balance-sheet burden

If a central bank fails to sterilize intervention adequately, the newly printed money inflates the money supply. This does not appear as a line-item balance-sheet loss, but it is a real economic cost: currency depreciation and loss of purchasing power. In effect, the inflation is a tax on money holders and savers—an indirect way to finance the central bank’s losses.

Many emerging-market central banks have faced this dilemma: sterilize and accumulate interest-rate losses, or skip sterilization and accept inflation. Some choose a middle path, accepting moderate inflation to keep sterilization costs manageable.

Limits on intervention endurance

The balance-sheet cost of FX intervention imposes a hard constraint on how long a central bank can sustain it. A central bank that has spent years accumulating losses cannot intervene indefinitely. Eventually, it runs low on capital, the government becomes unwilling to recapitalize it, or the central bank must accept failure.

This explains why many central banks, even those facing currency pressures, avoid large-scale sustained intervention. The cost is too high. Instead, they use intervention sparingly—during acute crises or to smooth intra-day volatility—while relying on interest-rate policy, monetary policy, and expectations management to influence currency direction long-term.

Countries with large, stable reserve positions (like Switzerland, Singapore, or the Eurozone) can tolerate larger losses without losing credibility. Countries with shaky fiscal positions or weak capital bases face harder constraints.

See also