Purchasing power parity and interest parity look like four formulas to memorise. They are one idea, that arbitrage removes easy profit, applied to goods, to interest rates with a forward contract, and to interest rates without one. This page derives each, gives you a lab for each, and walks through the numeric traps. 23 past-paper questions at the end.
An arbitrage is a profit with no risk and no money of your own: buy something where it is cheap and sell it where it is dear. If one were available, everyone would do it, and their trading would erase it. All four parity conditions are what is left once that has happened. Each applies the idea to a different kind of price.
| Condition | What arbitrage equalises | Says | Section |
|---|---|---|---|
| Absolute PPP | The price of a basket in two countries | S = P ÷ P* | Absolute PPP |
| Relative PPP | The same, as changes | %ΔS ≈ π − π* | Relative PPP |
| Covered interest parity | Returns at home and abroad, with a forward contract | F ÷ S = (1 + i) ÷ (1 + i*) | Covered parity |
| Uncovered interest parity | The same, using the spot rate you expect | Expected %ΔS ≈ i − i* | Uncovered parity |
Throughout, S is the spot exchange rate in rupees per dollar (the home currency per unit of foreign currency), F the forward rate, i the Indian interest rate, i* the American one, P and P* the Indian and American price levels, and π and π* the two inflation rates. If you have not met the quoting conventions, the page on exchange rates takes five minutes.
The law of one price says an identical, tradable good costs the same everywhere when measured in one currency. If a tonne of steel were cheaper in Mumbai than in Rotterdam, traders would buy in Mumbai and sell in Rotterdam until prices matched. Absolute purchasing power parity (PPP) applies that to a whole basket:
The best-known test is the Big Mac, and JAM 2022 gives its data. A burger costs 178 rupees in India and $5.58 in the US. At the market rate of ₹69.69 per dollar the Indian one costs $2.55, which is 54% cheaper. The exchange rate that would equalise the two is 178 ÷ 5.58 = ₹31.90 per dollar. Click through the table.
The data are those printed in JAM 2022 Q50 (January 2019). Pick a country to compare its burger with the American one at the market exchange rate and at the PPP rate.
A burger costs 178 in local currency. At the market rate (69.69 per dollar) that is $2.55, which is 54% cheaper than in the US. The PPP rate that would make it cost $5.58 is 31.9 per dollar, against the market's 69.69.
Only tradable goods can be arbitraged. A haircut or a flat cannot be shipped abroad, and these are most of what households buy. Transport costs, tariffs, taxes and different brands add further wedges. This leaves a pattern known as the Balassa-Samuelson effect: in poorer countries wages in tradables are low, and since the same wages apply to the non-tradable sectors, services are cheap, so overall price levels are low compared with the market exchange rate. That is why a rupee buys more in India than its dollar value suggests, and why economists compare incomes at PPP rather than market rates.
Absolute PPP asks for price levels, which are hard to compare. Relative PPP asks only for changes: the exchange rate moves by the difference in the two inflation rates. It can hold when levels differ permanently because of transport costs, since those costs cancel out of a change. The derivation is three steps, and the lab inside it draws the path of the rate year by year.
Absolute PPP says a basket costs the same in both countries when measured in one currency: S = P ÷ P*, where S is rupees per dollar, P the rupee price of the basket in India and P* its dollar price in the US. Suppose it holds today and a year from now:
Prices have grown by π in India and π* in the US, so P1 = P0(1 + π) and P1* = P0*(1 + π*). The starting prices cancel and only growth rates are left:
This is relative PPP, exact. It needs no assumption about the levels, only that the same relation holds at both dates.
CUET 2023: India 10%, US 4%. Exact: 1.10 ÷ 1.04 − 1 = 5.77%. Shortcut: 10 − 4 = 6%. The question’s options are 6%, 3%, 14% and 2.5%, so the shortcut is what it wants. When options are close together, use the exact form.
That one result answers a group of questions. CUET 2019: which causes depreciation, according to PPP? An increase in the nation’s own inflation (the other options are the opposite or unrelated). Hyderabad 2015 and 2018 test the same direction: if A’s inflation exceeds B’s, A’s currency weakens. CUET 2024 asks what holding PPP implies, and the true statements are the three in which the exchange rate equals the ratio of price levels, the law of one price holds, and price levels are equal in a common currency.
Market exchange rates exaggerate how rich rich countries are, so economists convert incomes at PPP. Two JNU questions test the method, and both can be done in one line once you see that the PPP rate comes from the basket.
The spot rate is for exchange now (two business days is standard). A forward contract fixes today the rate at which currency will be exchanged on a later date, say in three months. The forward premium on the foreign currency is the percentage by which the forward rate exceeds the spot rate, and a negative premium is a forward discount:
GATE 2024: the forward rate is 20 and the spot rate 8, so the premium is (20 − 8) ÷ 8 = 1.5, which is 150%. With S = 80 and a three-month F = 81, it is 1.25% for the quarter and 5% annualised.
Four words the papers like to match to their meanings:
Suppose you hold rupees for three months. You can lend them in India, or sell them for dollars, lend in the US, and lock in today, with a forward contract, the rate at which to convert back. Both routes are riskless. If one pays more, money floods into it until prices change to remove the difference. The derivation walks through it in four steps.
Take ₹1 and a period of, say, three months. Route 1: lend it at home at the period rate i. Route 2: sell the rupee for dollars at the spot rate S, lend the dollars at the American rate i*, and fix today, with a forward contract, the rate F at which those dollars will be turned back into rupees.
Both routes are riskless, because the forward contract removes all exchange-rate uncertainty. Both start with ₹1 and end with rupees.
Route 1 ends with 1 + i rupees. Route 2 takes three steps:
Put in UGC NET Dec 2023’s numbers: S = 80, F = 81, and quarterly rates i = 0.75% and i* = 0.25% (3% and 1% a year, divided by four). Route 1 gives 1.0075. Route 2 gives 1.0025 × 81 ÷ 80 = 1.0150. Route 2 wins by 0.75% of the sum invested every three months.
Anyone can borrow rupees at i, take route 2 and keep the difference, with no risk and no money of their own. So they do, in huge amounts, and the trades move four prices. Buying dollars spot pushes S up (the rupee weakens on the spot market). Selling dollars forward pushes F down. Money leaving India pushes Indian interest rates up, and money arriving in America pushes American rates down. All four moves narrow the gap, and they stop only when the two routes pay the same.
The left side is the forward premium on the dollar. For small rates the denominator 1 + i* is close to 1, which gives the exam form: forward premium ≈ i − i*. In the example, the no-profit forward rate is 80 × 1.0075 ÷ 1.0025 = ₹80.40, well below the market’s ₹81, which is why the arbitrage pays.
Now play with it. The lab starts with the UGC NET example. Move the forward rate and see where money flows, then use the presets for the other covered question, and the third for the uncovered case.
Lending in the US pays more in rupees than lending at home (6.01% against 3% a year), so money leaves India. Arbitrageurs buy dollars spot (the rupee weakens in the spot market), sell dollars forward (pulling the forward rate down), Indian interest rates are pushed up and American ones down, until the gap closes.
Covered parity needs F = S(1 + i)/(1 + i*) = 80.399. The market forward rate is 81. The shortcut is the premium equals the interest differential: 5% against 2%.
Skip the forward contract and use the rate you expect. Now the foreign return is risky, so the condition holds only if investors do not care about that risk (otherwise a risk premium enters, as in the DSE question). Under that assumption, the currency with the higher interest rate must be expected to depreciate, by about the interest differential:
Why? If India paid more and the rupee were not expected to weaken, everyone would hold rupees, pushing the rate up now until holding rupees looks no better than holding dollars. JNU 2021 puts it in words: if the interest rate is higher in X than in Y, the currency of Y is expected to appreciate relative to X.
It seems to contradict an earlier statement. Raising an interest rate attracts capital and makes the currency appreciate at once (Hyderabad 2019: rupee appreciates against the yen when Indian real interest rates rise relative to Japan’s). Yet uncovered parity says the currency with the higher rate is expected to depreciate afterwards. Both are true, one after the other: the currency jumps up when the news arrives, then drifts down, and the drift is what offsets the higher rate for investors. A question that says “expected to appreciate” after a rate rise (Hyderabad 2018) contradicts JNU 2021’s key, so we have left it out of the practice set.
Use the exact form unless the question asks for the shortcut.
Graded by each paper’s own difficulty. Questions with a key that conflicts with another paper’s (Hyderabad 2018 on the rate rise) or an ill-defined formula (UGC NET 2026) are left out; the reasons are listed at the top of the source file.
Definitions, directions and one-step sums. Most CUET, UGC NET and Hyderabad questions sit here.
Classification traps, multi-step numericals and statement questions, up to the hardest JNU, JAM and GATE items.
The Delhi School entrance asks this topic harder than any other exam. These questions are kept apart from the ordinary sets.
These are questions from the TestStop bank. A free account answers 10 a day, each with its worked solution; All-Access opens every set with no daily limit.
Absolute PPP says the exchange rate equals the ratio of price levels, S = P/P*, so a basket costs the same in both countries when measured in one currency. Relative PPP is the weaker version about changes: the exchange rate changes by the difference in inflation rates, so the home currency depreciates by about π − π*. Relative PPP can hold even when absolute PPP does not.
Covered interest parity says that, once the exchange-rate risk is removed with a forward contract, lending at home and lending abroad must earn the same return. That requires the forward premium on the foreign currency to equal the interest differential: F/S = (1 + i)/(1 + i*), approximately (F − S)/S = i − i*. If it fails, riskless arbitrage profits exist.
Covered parity uses the forward rate fixed today, so there is no risk and the condition is enforced by arbitrage. Uncovered parity uses the spot rate you merely expect, so it involves risk and holds only if investors are indifferent to that risk. Uncovered parity says the currency with the higher interest rate is expected to depreciate by about the interest differential.
The forward premium on the foreign currency is (F − S)/S, with both rates in home currency per unit of foreign currency. With a spot rate of 80 and a three-month forward rate of 81 it is 1/80 = 1.25% for the quarter, or 5% a year when annualised by multiplying by four. GATE 2024 gives a forward rate of 20 against a spot of 8, so the premium is 12/8 = 1.5, or 150%.
Because many goods and most services are not traded, transport costs and trade barriers separate prices, firms price differently in different markets, and price indices contain different baskets. Poorer countries also tend to have lower price levels for services, which is the Balassa-Samuelson effect, so market exchange rates understate their purchasing power.
Convert one income into the other currency at the PPP rate, the rate that makes a standard basket cost the same, instead of the market rate. If a burger costs ₹120 in India and £3 in the UK, the PPP rate is ₹40 per pound, so a £5,000 salary is worth ₹2,00,000 in Indian purchasing power, less than a ₹2,50,000 Indian salary even though the market rate of ₹100 per pound says the opposite.