Terms of Trade

One ratio, one statement about real income

The terms of trade is a price ratio with a real-income interpretation. Take a country's export-price index P_x and its import-price index P_m, both rebased to 100 in a reference year, and divide:

ToT_t = (P_x,t / P_m,t) × 100

A ToT of 120 means export prices have risen 20 percent faster than import prices since the base year. Every barrel, tonne, or container the country sends out now commands 20 percent more of the foreign basket of goods it imports — without changing what it actually produces. A ToT of 80 says the opposite: same exports, 20 percent less imports back. The country runs harder for less.

This is what economists mean when they say a terms-of-trade improvement is "equivalent to a real income gain". Real GDP measures the volume of output; real gross domestic income (RGDI) corrects that volume for changes in the relative price of what the country sells against what it buys. The difference — the "trading gain" — is exactly the terms-of-trade effect, and it can be enormous. For a commodity exporter, it can dominate the underlying GDP growth rate by an order of magnitude over a five-year stretch.

Why one ratio carries so much weight

Two countries can have identical production possibilities and identical labour productivity and still end up with wildly different welfare trajectories — purely because of how their export and import prices move. That makes terms of trade a first-class macro variable, alongside output, inflation, and the exchange rate. Three reasons:

• It splits "more output" from "better trades". Standard GDP growth confounds the two. ToT separates them: produce the same, trade better, get richer.
• It is the right lens on commodity exporters. For Saudi Arabia, Australia, Chile, or Russia, ToT swings drive fiscal balances, real exchange rates, and political cycles. Domestic growth often follows the terms-of-trade impulse rather than the other way round.
• It anchors the analysis of currency moves. Whether a devaluation helps depends on demand elasticities (Marshall-Lerner) and timing (J-curve) — both formulated explicitly in terms-of-trade terms.

Three definitions you will see in the literature

"Terms of trade" is a family of related indices, and which one is reported matters. The three canonical variants:

Name	Definition	What it tells you
Net barter terms of trade	P_x / P_m	How much imports one unit of exports buys (price ratio only).
Income terms of trade	(P_x / P_m) × Q_x	Total import-purchasing power of all exports (includes volume).
Single-factoral terms of trade	(P_x / P_m) × (Z_x)	Productivity-adjusted: imports per unit of factor input embedded in exports.

The bare net barter ratio is by far the most-quoted form and what people usually mean by "terms of trade". Income terms of trade is more relevant when export volumes change a lot — it captures, say, Norway's combined gain from rising oil prices and rising output. Single-factoral is the one development economists reach for when arguing about whether productivity gains in commodity production are being captured locally or competed away in lower prices — the Prebisch-Singer concern in its purest form.

Prebisch-Singer: the commodity-trap hypothesis

The most influential single claim in terms-of-trade history is the Prebisch-Singer hypothesis, advanced independently in 1950 by Raúl Prebisch (at UN ECLA in Santiago) and Hans Singer (at UN DESA in New York). Their argument: the long-run terms of trade of primary-commodity exporters trends down relative to manufactured-goods exporters. The mechanism has four parts:

1. Asymmetric income elasticities. As world income rises, demand for manufactures (Engel's law in reverse) grows faster than demand for food, fibres, and basic raw materials. Income elasticity is roughly 1.5 for manufactures, 0.3-0.7 for food, 0.1-0.3 for industrial commodities. So commodity demand growth lags rich-world income growth.
2. Asymmetric pass-through of productivity. Productivity gains in industrial centres are captured by unionised labour and corporate margins — prices stick or rise. Productivity gains in commodity production, where labour is unorganised and producers are price-takers in atomistic global markets, are competed away into lower output prices.
3. Synthetic substitution. Many commodities face slow-burning replacement by synthetics or recycling — natural rubber by synthetic, copper by aluminium and fibre, jute by polypropylene.
4. Cobweb dynamics. Commodity supply lags price by years (mines, plantations, herds). Prices overshoot up and down; investment-induced supply gluts can hold prices below trend for a decade.

The empirical record from roughly 1900 to 1980 was broadly consistent with Prebisch-Singer: indices of the relative price of non-fuel commodities to manufactures fell on average by about 0.5 to 1.0 percent per year, with massive cyclical variation around the trend. The hypothesis was the intellectual underpinning of import-substitution industrialisation strategies across Latin America (Argentina, Brazil, Mexico) and parts of Asia and Africa from the 1950s to 1970s — the case for tariffing your way past the commodity trap into a domestic manufacturing base.

The 2002–2014 reversal

From 2002 to 2014, the long-run pessimism reversed sharply. China's commodity-intensive growth phase — building, by some measures, the urban infrastructure equivalent of half the U.S. housing stock over a decade — drove a "supercycle" in oil, iron ore, copper, soy, and coal. Indexed against manufactured-goods prices, commodity prices roughly doubled in real terms over twelve years. The terms of trade of commodity exporters surged: Brazil's improved by roughly 60 percent, Australia's by 75 percent against its 2000 base, Russia's hydrocarbon-driven index nearly tripled. Income effects were enormous — Australia's mining-boom-driven income gain alone is estimated at around 10–15 percent of GDP cumulative, sustained for a decade.

For about ten years, Prebisch-Singer looked dead. Then the cycle turned. From mid-2014, a combination of China's investment slowdown, U.S. shale supply, OPEC's market-share strategy, and slowing emerging-market growth crashed commodity prices. By 2016 Brent crude was under $30 (from $110), iron ore had fallen 70 percent from its 2011 peak, and copper had halved. The decade-long ToT bonanza for commodity exporters partially or fully reversed in under two years.

The contemporary verdict is closer to: Prebisch-Singer is real as a low-frequency trend but the cyclical variation is much larger than the trend, so over any sub-thirty-year window the cyclical signal dominates. Commodity exporters face structural headwinds and spectacular cyclical opportunities. Both are true.

Marshall-Lerner: when devaluation helps

The terms of trade is also the variable through which currency depreciation transmits to the trade balance — but with two important wrinkles. Start from the identity that the value of net exports (in domestic currency) is

NX = P_x · Q_x − e · P_m* · Q_m

where e is the nominal exchange rate (domestic per foreign), P_m* is foreign-currency import price, and quantities Q_x, Q_m depend on relative prices. A devaluation raises e: it makes domestic exports cheaper in foreign currency (Q_x rises) and foreign imports more expensive in domestic currency (Q_m falls). But it also raises the domestic-currency value of the existing import bill (the e·P_m* term).

Whether the trade balance improves depends on whether the quantity responses outweigh the price effect on the existing import bill. The classical condition — the Marshall-Lerner condition, after Alfred Marshall and Abba Lerner — is:

|ε_x| + |ε_m| > 1

where ε_x and ε_m are the price elasticities of export and import demand. If demand is elastic enough on both sides — sum greater than one — devaluation improves the trade balance. If imports are inelastic (you have to keep buying oil and food no matter what they cost), the condition can fail and devaluation worsens the balance.

Long-run estimates for diversified economies put the sum of elasticities comfortably above 1, usually in the 1.5–2.5 range — Marshall-Lerner holds for most countries over multi-year horizons. But short-run elasticities are much smaller because of contract stickiness, leading directly to the J-curve.

The J-curve: worse before better

The J-curve is the empirical observation that immediately after a devaluation the trade balance typically gets worse before it starts to improve, tracing a J-shape over time. The mechanism is contract lag combined with currency-of-invoice stickiness:

1. Devaluation hits at t = 0. Exchange rate e jumps up.
2. For roughly 0–6 months, import volumes are locked in by existing contracts and shipments in transit. Quantities Q_m haven't fallen, Q_x hasn't risen. But the domestic-currency value of those unchanged import volumes has risen sharply.
3. Result: trade balance worsens.
4. Over 6–18 months, importers find substitutes, foreign buyers expand orders at the new lower prices, contracts roll over. Quantities respond. Marshall-Lerner kicks in.
5. Trade balance turns the corner and starts to improve.

The shape of the path traced by net exports against time is a "J": down first, then up. Empirically the depth and duration of the J vary enormously by country and shock — Mexico's 1994-95 peso crisis traced a sharp 6-month J, the UK's post-Brexit pound depreciation in 2016 had a long shallow trough lasting nearly two years. Brazilian real, Indonesian rupiah, Turkish lira — every emerging-market currency crisis since the 1980s has produced an identifiable J-curve in current-account data.

Why commodity exporters live and die by their terms of trade

Three structural reasons make commodity-exporting economies extraordinarily exposed to terms-of-trade swings.

• Concentration on the export side. A country with 60 percent of exports in oil sees its export price index move almost 1-for-1 with oil. A diversified manufacturer with 4 percent in each of 25 product categories needs all 25 to move together to get the same swing — which essentially never happens.
• Homogeneity and globally-priced. Commodities are sold into world markets at world prices. A Saudi barrel and a Russian barrel sell for the same dollar number (give or take quality differentials). There is no domestic pricing power.
• Inelastic short-run supply. Mines, oil fields, plantations take years to bring on or shut down. Output cannot quickly respond to price, so all the adjustment falls on price during demand swings.

The result is that commodity-exporter macros are pro-cyclical with world commodity prices. Their fiscal balances surge on the way up (resource taxes and royalties balloon), their real exchange rates appreciate (Dutch disease), wage demands accelerate. On the way down, fiscal deficits explode, foreign-exchange reserves drain, currencies depreciate, and import-dependent inflation surges. The amplitude is roughly an order of magnitude larger than for diversified manufacturers facing the same world demand shocks.

Case study: Saudi Arabia 2014–2016

The cleanest 21st-century example of a terms-of-trade collapse in a commodity exporter is Saudi Arabia 2014–2016. Brent crude fell from $112/barrel in June 2014 to $26/barrel in January 2016 — a 77 percent collapse over 19 months. Oil was roughly 90 percent of Saudi merchandise exports. The kingdom's net barter terms of trade halved.

Indicator	2013 (peak)	2016 (trough)	Δ
Brent crude average	$108.7 / bbl	$43.7 / bbl	−60 %
Oil revenue (SAR bn)	1 037	329	−68 %
Fiscal balance (% of GDP)	+5.6 %	−12.9 %	−18.5 pp
Current account (% of GDP)	+18.1 %	−3.7 %	−22 pp
FX reserves (USD bn)	737 (Aug 2014)	528 (May 2017)	−$209 bn

Two facts about the response worth noting. First, Saudi Arabia kept the riyal pegged to the dollar — no Marshall-Lerner-style adjustment via exchange rate was attempted. The entire shock was absorbed in the fiscal balance and reserves. Second, the political-economy response was strategic: Vision 2030 (April 2016) committed the country to long-run diversification, the partial Aramco IPO (December 2019) was structured to fund it, and November 2016's OPEC+ production-cut agreement brought Russia formally into the price-stabilisation cartel for the first time. All three are direct descendants of the 2014–2016 terms-of-trade shock.

Pricing power, monopoly leverage, and rare earths

Terms of trade are not just exogenous — countries try to manipulate them. Three twenty-first-century cases:

• OPEC and OPEC+. By coordinating production cuts, OPEC tries to keep oil prices above the level a competitive market would set. When it works (1973–74, 1979–80, 2017–22), it boosts members' terms of trade by tens of percentage points. When it fails (1986, 2014–16, 2020), the bust is brutal. The 2016 enlargement to OPEC+ (adding Russia) was a bid for more durable pricing power.
• Chinese rare-earth dominance. China supplies roughly 60–70 percent of mined rare-earth elements and over 85 percent of processed output, and has exercised that position three times: a 2010 export-quota cut against Japan after the Senkaku/Diaoyu incident, periodic restrictions through the 2010s, and explicit export-licence regimes from 2023 onward. The terms-of-trade impact on rare-earth-consuming countries — Japan, the EU, the U.S. — has been small in dollar terms (rare earths are a tiny fraction of total imports) but the strategic salience is outsized because of magnet, motor, and battery supply-chain implications.
• The U.S. dollar's exorbitant privilege. The largest pricing-power case in modern macro is not a cartel but a reserve currency. Because most U.S. imports are invoiced in dollars and most foreign reserves are held in dollar assets, the United States can run persistent current-account deficits financed by issuing its own currency. A dollar depreciation transfers wealth toward the U.S. (whose foreign assets are denominated in foreign currency, but whose liabilities to foreigners are in dollars) — almost the inverse of how depreciation hurts other countries. Valéry Giscard d'Estaing in the 1960s called it "exorbitant privilege"; Pierre-Olivier Gourinchas and Hélène Rey quantified it in 2007 at roughly 2–3 percent of U.S. GDP per year in pure financial-yield terms, on top of the trade-channel asymmetry.

How statistical agencies actually measure it

Real-world terms-of-trade indices are constructed from matched export and import price (or unit-value) indices, typically following one of two recipes:

• Laspeyres unit-value index. Holds the basket of goods constant at base-year quantities; the index moves only with price changes. The OECD's main ToT series follows this convention.
• Paasche unit-value index. Uses current-period quantities. Used in several national accounts (Australia's ABS, Statistics Canada).

Three known measurement issues worth flagging:

• Composition vs price. Unit-value indices mix true price changes with shifts in the composition of trade. If a country shifts from cheap-cotton-shirts to luxury-cashmere exports, its export "unit value" rises even if no underlying price has moved. Pure price indices (where available) are preferred but harder to construct.
• Services trade. Goods trade is well-indexed; services trade (finance, software, royalties, tourism) is much harder. As services rise to 20–30 percent of total trade for most rich economies, goods-only ToT increasingly miss the real picture.
• Global value chains. A 'Mexican car export' might be 70 percent imported inputs by value. Headline ToT swings overstate income effects because most of the import-content costs already get netted. The OECD now publishes value-added-adjusted ToT for the largest economies, which can differ from headline ToT by 10–20 percent in either direction.

Worked example: an Australian household's mining-boom dividend

Consider a stylised Australian terms-of-trade gain from 2003 to 2011. Iron-ore export prices (FOB, USD per tonne) rose from $14 to $168 — a 12× increase. Coal prices roughly tripled. Manufactured-goods import prices were essentially flat in nominal terms over the period. Australia's ToT index (ABS, base 2002 = 100) rose from 100 to roughly 175.

ToT_2011 / ToT_2003 ≈ 1.75
Implied real income gain ≈ (ToT − 1) × (Exports / GDP)
                        ≈ 0.75 × 0.21
                        ≈ 16 % of GDP cumulative

For a country with ~$1.5 trillion GDP at the time, that is on the order of $240 billion in cumulative real income windfall over the boom decade — without producing a single additional tonne. Distributing that across roughly 9 million households gives an order-of-magnitude per-household figure of $25 000–30 000 of real-income gain, mostly captured by the federal government (resource taxes), state governments (royalties), mining-company shareholders, and through real-wage gains. Critics — Saul Eslake, Ross Garnaut, John Quiggin — have argued substantial parts of the windfall were dissipated in tax cuts and middle-class transfers rather than saved into a sovereign-wealth fund as Norway did with its oil revenue.

The link to the real exchange rate

Terms of trade are tightly linked to a related variable, the real exchange rate. The real exchange rate is the relative price of foreign goods to domestic goods, expressed in a common currency. Improvements in the terms of trade — especially for commodity exporters — tend to drive real-exchange-rate appreciation through the so-called "commodity-currency" mechanism: higher commodity prices boost export receipts, capital flows in, the nominal exchange rate strengthens, domestic non-tradable-sector wages rise faster than tradable-sector wages, and the real exchange rate appreciates.

For the Australian, Canadian, Norwegian, Brazilian, Russian, and Chilean currencies, the correlation between commodity terms of trade and the real exchange rate over 1980–2020 is roughly +0.7 to +0.9. These are quintessential commodity currencies. The mechanism is at the heart of the Dutch-disease literature: a terms-of-trade-driven real appreciation hollows out manufacturing exports and concentrates resources further into the booming commodity sector — a feedback loop that can persist for decades.

Common pitfalls and misreadings

• Confusing terms of trade with trade balance. A country can have improving terms of trade and a worsening trade balance simultaneously (if quantities fall faster than prices rise) and vice versa. They are different variables capturing different things — price ratios vs net flows.
• Treating an improvement as unambiguously good. Even a terms-of-trade gain has distributional consequences: it appreciates the real exchange rate, hurts tradable-sector employment (Dutch disease), and concentrates wealth toward resource owners. The net welfare effect depends on how the gain is shared.
• Quoting nominal vs real series interchangeably. Net barter ToT is in relative price units — it is already real. Income ToT mixes in volume changes. Always check which one a chart shows; the two can move opposite directions in a given quarter.
• Forgetting the invoicing currency. For commodity exporters whose exports are invoiced in dollars but imports in mixed currencies, dollar-strength shocks alter terms of trade independently of underlying commodity prices.
• Applying Marshall-Lerner without J-curve timing. A devaluation can satisfy Marshall-Lerner in the long run while failing it in the short run. Policy debates about whether a depreciation is "working" depend heavily on how much time has elapsed — six months is too soon to call, two years is usually enough.
• Reading Prebisch-Singer as a settled finding. Empirically the secular decline holds over century-scale data but is dominated by cyclical variation at any sub-thirty-year horizon. Strong claims that commodity exporters are doomed look very different in 2011 vs 2016.

The big picture

Terms of trade is the variable that makes "real income" different from "real output". A country can produce the same physical basket of goods two years running and end up richer if global prices have moved its way — or poorer if they have moved against it. That is one of the deep insights of open-economy macroeconomics, and it sits behind everything from why Saudi Arabia announced Vision 2030 in 2016 to why the Australian dollar moves with iron-ore futures to why import-substitution industrialisation was the development orthodoxy of the mid-twentieth century. One ratio. P_x over P_m. The single most underrated number in international economics.