For most developing economies the exchange rate is not a sideshow to monetary policy — it is at its centre, because of the high pass-through from the currency to inflation and the importance of the external sector. This module covers how and why central banks intervene in the FX market, the iron constraint of the trilemma, and the pervasive reality that almost no one really lets their currency float freely.
The trilemma
The impossible trinity
The Mundell-Fleming trilemma states that a country can have at most TWO of these three: • A fixed (or heavily managed) exchange rate • Free international capital movement • An independent monetary policy The logic: if capital can move freely and you fix the exchange rate, then your interest rate must equal the world rate (arbitrage forces it — any gap triggers capital flows that you must accommodate to hold the peg), so you lose monetary independence. To regain monetary independence with open capital, you must let the exchange rate float. To keep both a fixed rate AND monetary independence, you must restrict capital flows. There is no escape: every exchange-rate-and-monetary arrangement is a choice of which corner of the trilemma to give up. Understanding which two a country has chosen explains most of its monetary constraints.
FX intervention: sterilised and unsterilised
Intervention means the central bank buying or selling foreign exchange to influence the rate. The crucial distinction is whether the money-supply effect is offset:
- Unsterilised intervention — the bank sells FX (to support the currency) and lets the resulting fall in domestic money supply stand. This is powerful (it actually tightens monetary conditions) but it means the exchange-rate goal is driving monetary policy — you have, in effect, chosen the fixed-rate corner of the trilemma for that moment.
- Sterilised intervention — the bank sells FX but simultaneously offsets the money-supply effect with an open-market operation (buying domestic securities to put the reserves back), so the FX intervention does not change domestic monetary conditions. This tries to influence the exchange rate WITHOUT giving up monetary independence — squaring the trilemma, at least temporarily.
Does sterilised intervention work?
Sterilised intervention is the central bank's attempt to have it both ways, and its effectiveness is debated. With perfect capital mobility and perfect asset substitutability, it should not work at all (the offset means nothing real changed). In practice it can work through two channels: the portfolio-balance channel (changing the relative supply of domestic vs foreign assets shifts their relative price/yield, moving the exchange rate) and the signalling channel (the intervention signals the bank's intentions and view, moving expectations). The evidence: sterilised intervention has modest, temporary effects, larger in less-developed/segmented markets (where assets are imperfect substitutes and signalling matters more) — which is precisely the African context. So it is a real but limited tool, useful for smoothing volatility, not for resisting fundamental misalignment.
Reserves and the fear of floating
Central banks accumulate FX reserves to give themselves the ammunition to intervene, to insure against sudden stops (the International Macro course), and to bolster confidence. Holding reserves is costly (they earn low returns while the country may borrow at high rates — a negative carry), so the level held reveals how much a country values exchange-rate stability and insurance.
Fear of floating (Calvo-Reinhart)
Calvo and Reinhart (2002) made a famous observation: many countries that officially claim to float their currencies in fact intervene heavily to limit its movement — they exhibit 'fear of floating'. Why? Because high exchange-rate pass-through means a depreciation quickly becomes inflation; because firms and governments with foreign-currency debt suffer balance-sheet damage when the currency falls (currency mismatch, the 'original sin' of the Sovereign Debt course); and because volatile currencies deter trade and investment. So the de jure 'float' is a de facto managed float — the de jure/de facto gap again. This is the single most important fact about exchange-rate regimes in practice: declared floats are usually managed, and a central bank's true regime is revealed by how much its reserves and the exchange rate actually move, not by what it announces (Reinhart-Rogoff's 'natural classification', International Macro course).
Exercise
A central bank with a managed float and largely open capital account faces sharp depreciation pressure: the currency is falling fast, threatening to spike inflation (high pass-through) and damage firms with dollar debt. It is considering (a) selling reserves to defend the currency, (b) hiking the policy rate, or (c) letting the currency fall. (1) Use the trilemma to frame the bind it is in. (2) Compare selling reserves (sterilised vs unsterilised) as a response. (3) Explain how 'fear of floating' makes option (c) painful despite the official float. (4) What determines whether defending the currency is wise or futile here?