What is purchasing power parity and why does it fail short-term?

The short answer

PPP starts with a simple intuition: if a Big Mac costs $5 in New York and €4 in Paris, the EUR/USD rate should be about 1.25 — otherwise an arbitrageur could buy burgers cheap and sell them expensive. Extended to baskets of goods, PPP becomes a theoretical anchor for what exchange rates should be in the long run.

The Big Mac Index, published by The Economist since 1986, illustrates how often markets diverge from this anchor. The euro can spend years 20% above or below its PPP value against the dollar, and the Japanese yen has been chronically undervalued by 30-40% on PPP measures since 2022.

The deeper finding from the academic literature is that PPP does work — but on horizons of 5 to 10 years, not weeks or months. The puzzle Obstfeld and Rogoff identified in 2000 is why convergence is so slow when economic theory suggests arbitrage should move much faster.

→ New to FX dynamics? FX hub

What the data shows

Empirical estimates from the OECD, the BIS and academic research converge on a striking pattern (OECD PPP database, BIS effective exchange rates, 2000-2024):

• Real exchange rate deviations from PPP can persist for 3 to 5 years before significant reversal
• The half-life of PPP deviations — the time for half the gap to close — is typically estimated at 3-5 years for major currencies
• The yen has traded 30-40% below its OECD PPP estimate against the dollar since 2022
• The Swiss franc has consistently traded above PPP since the 2008 crisis, reflecting its safe-haven premium
• Among emerging market currencies, deviations of 50% or more from PPP are common and can last a decade

One notable exception: hyperinflation episodes show near-immediate PPP adjustment because the inflation differential dominates everything else. Argentina and Turkey provide recurrent case studies.

→ Dataset: US Dollar Index (broad trade-weighted)

Why it happens — the macro mechanism

The slow convergence to PPP — the heart of what economists call the PPP puzzle — has three reinforcing structural causes.

Non-tradable goods. A large share of any consumption basket is non-tradable: housing, haircuts, healthcare, restaurant meals. These cannot be arbitraged across borders, so their prices reflect local wages and rents rather than global market prices. The Balassa-Samuelson effect predicts that richer countries have systematically higher non-tradable prices, creating persistent PPP deviations.

Pricing-to-market and sticky prices. Even tradable goods are not perfectly arbitraged. Multinationals price-discriminate by country, keeping local prices stable in local currency rather than passing exchange rate moves through. A car can cost very different amounts in different countries even when shipping is cheap.

Conventional textbooks treat PPP failures as short-term noise around a long-term equilibrium. The empirical reality is that PPP deviations can be both very large and very persistent — large enough that PPP is essentially useless as a short-term forecasting tool, but real enough that it eventually anchors valuations on multi-year horizons.

Capital flows dominate trade flows. FX markets trade roughly 9.6 trillion USD per day (BIS, 2025), while world trade in goods and services is closer to 70-80 billion USD per day. Even if PPP holds for trade flows, it cannot hold daily when capital flows of 100 times that size dominate price formation.

Synthesis by regime: under capital controls (Bretton Woods era and emerging markets pre-1990), PPP worked relatively well because capital flows were constrained and trade flows dominated; under free convertibility (post-1990 G10), capital flows overwhelm trade flows in price formation, allowing exchange rates to drift far from PPP for years; the regime parameter is the depth and freedom of cross-border capital movement.

PPP is a long-run gravitational pull, not a daily anchor — and gravity takes years to win against capital flows.

→ Framework: FX markets and monetary regimes

What it means for different economic actors

Savers considering geographic diversification should think in PPP terms over decade-long horizons, not in current exchange rates. A currency that is 30% undervalued on PPP today may compound an additional return through eventual mean reversion — but this can take years to materialise.

Investors in international equities have an embedded PPP exposure. Buying Japanese stocks at JPY 150/USD when PPP is closer to 100 means a portion of the expected return comes from FX mean reversion, not corporate fundamentals.

Multinational corporations face the operational version of this puzzle: where to locate production when currencies are far from PPP. The answer is not always to move toward the cheaper country — non-tradable input costs (rent, regulation) often offset apparent labour cost advantages.

A common error is using current PPP measures to time short-term currency trades. The literature consistently shows that PPP signals work over multi-year horizons but provide essentially no edge over 3-12 months.

Practical observation

Go deeper

Related questions

• How do currency exchange rates affect trade balances?
• What makes the dollar the global reserve currency?
• How does the dollar index (DXY) work?
• What causes the Japanese yen to act as a safe haven?

Frequently asked questions

Is the Big Mac Index a serious economic measure?

It started as a tongue-in-cheek illustration but has become a genuinely useful first-pass tool. The Big Mac is one of the few products with near-identical specifications across 70+ countries, making cross-country comparisons cleaner than abstract baskets. The IMF and several central banks reference it in research notes. Its limits are obvious — one product cannot capture a full consumption basket — but its directional signals correlate well with broader OECD PPP measures.

Why does PPP convergence take so long?

This is the heart of the PPP puzzle. Standard models predict that arbitrage should close PPP gaps within months. Empirically, half-lives of 3-5 years require either very high transaction costs (implausible for major currencies) or persistent shocks to real exchange rates (which the data does not strongly support). The leading candidate explanation combines pricing-to-market behaviour, sticky local-currency contracts, and the dominance of capital flows over trade flows in setting daily prices.

Does PPP apply to digital goods and services?

Imperfectly. Digital subscriptions like Netflix or Spotify show large PPP-violating price differences across countries — Netflix in India costs a fraction of Netflix in Switzerland for similar content. This persists because location-based pricing locks consumers in by national billing accounts, and VPN arbitrage is restricted by the platforms. Digital tradables behave somewhere between physical tradables (which arbitrage faster) and pure non-tradables (which barely arbitrage at all).

Last updated — 19 May 2026