Exchange rates

A depreciation makes a country's exports cheaper and imports dearer. Whether this improves the current account, and the timing of the effect, are explained by the Marshall-Lerner condition and the J-curve.

The Marshall-Lerner condition states that a depreciation improves the current account only if the combined price elasticity of demand for exports and imports is greater than one (PEDx + PEDm > 1). The logic: a depreciation lowers export prices and raises import prices, but whether the balance improves depends on how much quantities respond. If demand is sufficiently elastic (sum > 1), the rise in export volumes and fall in import volumes are large enough to outweigh the price effects, so export revenue rises and import spending falls — the current account improves. But if demand is inelastic (sum < 1), quantities respond little, so the higher cost of imports (which are still bought in similar quantities) worsens the current account. So a depreciation is only an effective cure for a deficit if demand is elastic enough.

The J-curve describes the timing. Even when the Marshall-Lerner condition holds in the long run, demand is typically inelastic in the short run — contracts are fixed, and consumers and firms are slow to change suppliers and habits. So immediately after the depreciation, import volumes barely fall while imports cost more, and export volumes barely rise: the current account worsens first (the dip of the 'J'). Over the long run, demand becomes more elastic as buyers switch to the cheaper exports and away from dearer imports, so export volumes rise and import volumes fall, and the current account improves (the upswing of the 'J'), ending up better than before.

Together these show that a depreciation will only improve the current account if demand is sufficiently elastic (Marshall-Lerner), and even then only after a time lag during which the balance worsens (the J-curve). This is a key reason a depreciation is an unreliable and potentially painful tool for correcting a deficit, and why supply-side policies that raise competitiveness may be preferable.

A fixed exchange rate is pegged and defended by the central bank; a floating rate is set by market forces; a managed system lies between. Whether fixed is 'better' depends on the country's circumstances.

The case for a fixed system. A fixed rate provides certainty and stability for trade and investment — firms can plan exports/imports and FDI without exchange-rate risk, encouraging international business. It imposes discipline on inflation, because the government cannot let the currency fall to mask high inflation, and it can build credibility (e.g. pegging to a stable currency). For small, trade-dependent or developing economies, this stability can be very valuable.

The case against a fixed system / for floating. Maintaining a fixed rate requires the central bank to hold large foreign-currency reserves to defend the peg and to use interest rates to support the currency — which removes monetary-policy independence (rates must defend the peg, not manage the domestic economy). Fixed rates are vulnerable to speculative attacks if markets doubt the peg, which can force a costly devaluation. A floating rate, by contrast, adjusts automatically to correct imbalances (a deficit causes depreciation, restoring competitiveness), frees monetary policy for domestic objectives, and needs no reserves to defend a rate. Its drawback is volatility, which creates uncertainty and can be driven by speculation.

Marshall-Lerner / adjustment. A floating rate's automatic adjustment only corrects a deficit if the Marshall-Lerner condition holds (elastic demand), and the J-curve means adjustment is slow and initially painful — so floating is not a perfect self-corrector. A fixed rate, meanwhile, forces the domestic economy (via deflation/contraction) to adjust to imbalances, which can be costly in unemployment.

Judgement. Neither system is universally better — it depends on the country's circumstances and priorities. A fixed system suits a country that values stability and inflation discipline above all (e.g. a small, open, trade-dependent or high-inflation economy), and is willing to sacrifice monetary independence and hold reserves. A floating system suits a country that values monetary-policy independence and automatic adjustment and can tolerate volatility (e.g. a large, diversified economy). Many countries adopt a managed system to get some of both. The most defensible conclusion is that a fixed rate is not inherently better: it offers stability at the cost of flexibility and independence, so the right choice depends on whether the economy prioritises stability (favouring fixed) or flexibility and policy independence (favouring floating).

This pattern — a current-account deficit worsening first and then improving after a devaluation — is the J-curve effect, and reflects how demand elasticities change over time.

Immediately after the devaluation (the first six months), demand for exports and imports is inelastic: trade contracts are already fixed, and consumers and firms are slow to change their suppliers and buying habits. So import volumes barely fall, but imports are now 20% dearer, costing more in total; meanwhile export volumes barely rise. The result is that import spending rises relative to export revenue, so the current-account deficit worsens — the dip of the 'J'.

Over time, demand becomes more elastic as buyers respond: foreign customers switch toward the now-cheaper exports, and domestic consumers switch away from the dearer imports. Export volumes rise and import volumes fall, so export revenue increases and import spending decreases, and the current account improves — the upswing of the 'J'.

So the devaluation does eventually improve the current account (implying the Marshall-Lerner condition holds in the long run, with combined elasticities greater than one), but only after a lag during which the deficit gets worse. This is why a devaluation can be painful in the short run and why governments must be patient — or use supply-side measures instead if demand is too inelastic.

A depreciation lowers the external value of a currency, making exports cheaper and imports dearer — apparently improving competitiveness. But its effectiveness is conditional.

The case that it improves competitiveness. A depreciation immediately makes a country's exports cheaper in foreign currency and imports dearer at home, so domestic goods become more price-competitive both abroad and in the home market. This can raise net exports, output and employment, and help correct a current-account deficit. For an economy whose goods are price-sensitive, this can be an effective short-cut to competitiveness.

The limitations. First, the improvement in the current account depends on the Marshall-Lerner condition (PEDx + PEDm > 1); if demand is inelastic, a depreciation may not improve the balance and could worsen it. Second, the J-curve means even a successful depreciation worsens the current account before it improves. Third, the gain in competitiveness is only based on price and may be temporary: dearer imports raise the cost of imported raw materials and goods, causing cost-push inflation that erodes the very competitiveness the depreciation created. Trading partners may also retaliate or let their own currencies fall ('currency war'). Crucially, a depreciation does nothing to improve underlying competitiveness — productivity, quality and innovation — so it treats the symptom, not the cause.

Judgement. A depreciation can improve price competitiveness in the short run, and may help correct a deficit if demand is sufficiently elastic (Marshall-Lerner) and the economy can bear the J-curve and inflation. But it is not a reliable or lasting route to competitiveness, because it can fail (inelastic demand), worsens things first (J-curve), and is undermined by the resulting inflation. The most defensible conclusion is that genuine, lasting competitiveness comes from supply-side improvements (productivity, quality, innovation) rather than a depreciation, which at best offers a temporary, conditional and partly self-defeating boost. So a depreciation is an unreliable tool for improving competitiveness.