ECO 7 Capital Flows and the FX Market

Currencies trade against one another in the foreign exchange (FX) market, which stands by a wide margin as the largest trading venue in the world. Current estimates put daily turnover at roughly USD6.6 trillion for 2019, which is roughly 10 to 15 times what changes hands daily in the world’s bond markets and around 50 times the daily figure for equities. The market runs 24 hours a day on each business day, linking participants in every time zone through electronic networks, from multibillion-dollar investment funds down to individuals trading for their own account. International trade and cross-border capital flows would both be impossible without the currency trading that supports them.

Because production, trade, and investment are increasingly transnational, the FX market matters even to investors who never think they hold foreign assets. Large companies depend heavily on foreign operations (on some estimates roughly 30 percent of what the S&P 500 Index earns comes from abroad), nearly every firm faces some foreign competition, and the pricing of domestic equities, bonds, and real estate depends partly on demand from foreign investors. As investors let go of their home bias, the exchange rate, meaning the domestic-currency value placed on assets held in a foreign currency, weighs more heavily on portfolio returns.

Currencies versus exchange rates

It is important to separate an individual currency from an exchange rate. A person can hold a single currency, for example a EUR100 million deposit. An exchange rate, by contrast, always ties two currencies together: it prices one currency against a second. Standardized three-letter codes agreed through the International Organization for Standardization (ISO) name individual currencies, for example USD for the US dollar, EUR for the euro, GBP for the British pound, JPY for the Japanese yen, and CHF for the Swiss franc.

An exchange rate reports how many units of the price currency are needed to purchase a single unit of the base currency, so it doubles as the cost of a single base-currency unit expressed in the price currency. This lesson uses the A/B format, where the figure shows how much of A trades for one unit of B. Take USD/EUR = 1.1700: one euro trades for 1.1700 US dollars, which makes the euro the base and the US dollar the price currency. If that figure falls, each euro fetches fewer dollars, and the reading is that the US dollar is strengthening while the euro is weakening.

Nominal rates, real rates, and purchasing power parity

The rates quoted and traded in the market are nominal exchange rates. Real exchange rates, in contrast, are indexes built by economists and analysts to compare the relative purchasing power of two currencies. Building one means adjusting the nominal rate for the price levels in each country, which is why they carry the word real. Real exchange rates are never quoted or traded in the market; they exist only as analytical indexes that describe the international competitiveness of an economy.

The intuition behind the adjustment is purchasing power parity (PPP). In a world of identical goods with no trade barriers or frictions, no one would pay more in real terms at home for a good that could be imported more cheaply, so relative purchasing power across countries would equalize. PPP therefore asserts that nominal rates adjust until identical baskets of goods cost the same everywhere. In practice the conditions never hold: baskets differ across countries, many goods are not traded internationally, shipping costs and import taxes intervene, and capital flows matter at least as much as trade flows in setting rates. Nominal rates therefore deviate from PPP persistently, and relative purchasing power shows little tendency to equalize over the long run. The Economist magazine’s Big Mac index, which compares the price of a standardized hamburger across countries, is a simple illustration that such price gaps are wide and typical.

Consider someone who wants to buy goods from a foreign country. That person can buy fewer foreign goods if the nominal spot rate for the foreign currency appreciates or if the foreign price level rises, and can buy more if domestic income rises. Assuming income changes in proportion to the domestic price level, the real exchange rate a buyer faces rises with the nominal exchange rate (quoted as domestic currency per unit of foreign currency) and with the foreign price level, and falls with the domestic price level. A higher real exchange rate leaves the buyer able to purchase fewer foreign goods and worse off in relative purchasing power.

An equivalent view is that the real exchange rate expresses relative price levels at home and abroad. Writing the foreign price level in domestic currency terms:

Dividing by the domestic price level P(d) gives the real exchange rate as the ratio of the two price levels:

For a British consumer buying Eurozone goods, the domestic currency (the pound) is the price currency, so the real exchange rate is defined in GBP/EUR terms. It rises with the nominal GBP/EUR rate and the Eurozone price level, and falls with the UK price level:

To track how the real rate moves, combine the growth rates of the three inputs. The exact expression multiplies the gross changes; a convenient approximation simply adds the nominal change and the foreign inflation rate and subtracts the domestic inflation rate:

The same logic explains why inflation gaps matter even when the nominal rate is stable. During 2018 the nominal INR/USD rate climbed roughly 6.7 percent, meaning the US dollar gained on the rupee. Inflation over that year ran at about 2.5 percent in the United States and about 4.7 percent in India. Combining these, the real INR/USD rate changed by about 6.7% + 2.5% − 4.7%, or roughly 4.5 percent. The weaker rupee together with higher Indian inflation raised the real exchange rate India faced, cutting Indian purchasing power in US dollar terms. Because PPP is a poor guide to future nominal rates, real exchange rates likewise have a weak record as predictors of future nominal rate moves.

The universe of FX participants ranges from multibillion-dollar funds to a tourist changing money at an airport kiosk. A useful first split is between the sell side and the buy side. The sell side is made up of large FX dealing banks, among them Deutsche Bank, UBS, and Citigroup, which quote prices and sell FX products. The buy side is the clients who buy those products from the sell-side banks.

Buy-side categories

• Corporate accounts: firms of all sizes trading FX for cross-border purchases and sales, and for investment-related flows like overseas mergers and acquisitions, buying foreign assets, and borrowing in foreign currency.
• Real money accounts: funds run by insurers, mutual funds, pension funds, endowments, and exchange-traded funds. They are called real money because they are usually restricted in their use of leverage and derivatives.
• Leveraged accounts: the professional trading community, including hedge funds, proprietary trading shops, commodity trading advisers, high-frequency algorithmic traders, and bank proprietary desks. They account for a large and growing share of daily turnover, with horizons ranging from long-term macro views down to milliseconds.
• Retail accounts: individuals trading their own accounts, smaller active traders, and households shifting savings into foreign currencies. Retail activity is estimated at as much as 10 percent of spot transactions in some pairs, and it is growing.
• Governments: public entities with FX needs from small (consulates) to large (military purchases), sometimes purely transactional and sometimes tied to public policy. Some public pension and insurance schemes, such as the Caisse de depot et placement du Quebec (nearly CAD420 billion of assets at the end of 2021), behave much like investment funds.
• Central banks: they sometimes intervene to influence the level or trend of the domestic rate, for instance when the currency looks too weak after a speculative attack or too strong for export competitiveness. They also manage FX reserves, behaving much like cautious real money funds. Global FX reserves stood close to USD13 trillion by the end of 2021, of which almost 60 percent of allocated reserves sat in US dollars and a little over 20 percent in euros.
• Sovereign wealth funds: government investment vehicles funded from current account surpluses, closer to real money accounts but sometimes using derivatives or aggressive strategies. Their flows can move most major pairs.

Sell-side structure

A large and growing share of daily turnover runs through the biggest money center banks, such as Deutsche Bank, Citigroup, UBS, and HSBC. Competing in FX requires heavy fixed-cost investment in technology and a broad global client base, so only the largest first-tier banks can quote competitively across the full product range, and much of their business is crossed internally by matching buyers and sellers within their own client base. Second- and third-tier banks, often regional or emerging-market institutions, lack that scale and frequently outsource FX to the tier-one banks or rely on their liquidity.

What the market trades

Besides spot transactions, the market includes forwards (agreed today for settlement at a future date, useful for managing FX risk) and FX swaps (a spot and a forward combined). Swaps make up the largest daily volume. Drawing on the Bank for International Settlements (BIS) 2016 Triennial Survey, average daily turnover in the traditional market (spot, outright forwards, and FX swaps) was about USD5.1 trillion. The tables below add exchange-traded derivatives to that estimate; the spot and outright forward figures exclude transactions that are part of a swap.

An FX swap differs from a currency swap; the latter usually spans several periods and payment exchanges. Columns may not sum to 100 percent because of rounding.

Financial client flows exceeded interbank volume for the first time in 2010. Only a minority of daily flow comes from corporations and individuals buying foreign goods and services.

Each of the most active pairs includes the US dollar. The largest share of trading is in London, then New York, so the market is busiest between about 8:00 a.m. and 11:30 a.m. New York time; Tokyo is the third-largest hub.

An exchange rate can be written two ways that are simply the inverse of each other: how many units of B one unit of A buys, or how many units of A one unit of B buys. In the convention A/B there are a number of units of currency A (the price currency, also called the quote currency) per one unit of currency B (the base currency), and the base currency is always set to a quantity of one.

Direct and indirect quotes

In a direct quote the home country’s currency sits in the price slot and the foreign currency is the base. A trader based in Paris treats the euro as the home currency, so a direct quote against the pound is EUR/GBP: a quote of EUR/GBP = 1.1211 means GBP1 costs EUR1.1211. For that same trader an indirect quote puts the domestic euro as the base currency, so GBP/EUR = 0.8920 means EUR1 costs GBP0.8920. The two are reciprocals. Whether a quote is direct or indirect depends on where the trader sits, so the professional market avoids those labels and relies on agreed conventions instead.

Three-letter codes always refer to a rate against the US dollar; six-letter codes are cross-rates that need not involve the US dollar.

Three-letter codes name a major rate against the US dollar, so EUR alone means the euro-US dollar rate quoted as USD/EUR. Six-letter codes name cross-rates, which are secondary rates that need not involve the dollar. When a code or name lists both currencies, the base currency is named first, the reverse of the actual ratio: the code GBPJPY (sterling-yen) is quoted as yen per pound, JPY/GBP. A hierarchy sets which currency is the base: the euro is the base in any pair that includes it, then the pound in any remaining pair, then the US dollar. The Australian and New Zealand dollars are exceptions, serving as the base against the US dollar (USD/AUD, USD/NZD).

Two-sided prices

When a client asks for a quote, a bank shows a two-sided price: a bid, at which it will buy the base currency, and an offer, at which it will sell it, each expressed in the price currency per single unit of the base currency. For euro-Swiss (CHF/EUR), a quote might read 1.1583 to 1.1585. The client receives CHF1.1583 for selling EUR1 to the bank and pays CHF1.1585 to buy EUR1. The bid is always below the offer: the bank buys the base low and sells it high, and the bid/offer spread is its margin. Most major spot rates are quoted to four decimal places, with the yen an exception at two decimal places because yen amounts per unit are large. For simplicity, the rest of this lesson treats each rate as a single number with no spread.

Percentage changes and the base-currency trap

Any move in a rate can be written as a percentage gain of one currency on the other, but the figure hinges on which currency is the price currency. Suppose USD/EUR rises from 1.1500 to 1.2000. The percentage change is computed on the base currency (the euro):

That works out to a 4.35 percent gain by the euro on the US dollar, since USD/EUR carries the US dollar as the price currency. The dollar’s depreciation against the euro is not the same number. Inverting the quote so the euro becomes the price currency:

The euro’s 4.35 percent appreciation corresponds to a 4.17 percent depreciation of the US dollar, not 4.35 percent. Mathematically these two percentages are always different, because each is measured against a different base. The safe habit is to fix which currency is the base, compute the change on that base, and never assume the reverse move equals the same percentage.

Sharp swings in exchange rates breed uncertainty that eats into real activity, muddies investment, and raises hedging costs. How much a currency swings depends partly on the policy framework around it. Virtually every rate is managed to some degree by a central bank, and the framework each central bank adopts is its exchange rate regime.

Why there is no ideal regime

An ideal regime would have three properties: exchange rates credibly fixed between all currencies, currencies fully convertible for any purpose and amount, and each country able to run fully independent monetary policy. These three cannot hold together. If rates were credibly fixed and currencies fully convertible, there would in effect be a single world currency, and adjusting one currency relative to another to steer interest rates or inflation would be futile. So independent monetary policy is impossible once rates are credibly fixed and currencies fully convertible.

Under perfect capital mobility the tension is stark. With a credibly fixed rate, trying to cut domestic interest rates below foreign levels would trigger a potentially unlimited capital outflow chasing that higher yield, and the central bank would then be forced to sell foreign currency and buying its own to defend the peg, bleeding reserves and contracting the money supply until rates were dragged back up, undoing the policy. With a floating rate the story reverses: a rate cut makes the currency less attractive, and the resulting depreciation shifts demand toward domestic goods, reinforcing the stimulus. The broad lesson: the more freely a rate floats and the more that convertibility is reined in, the more room a central bank has to act on its own, though at the cost of exchange rate risk and a less efficient spread of capital.

A short history

For most of the 19th century up to World War I, the pound and the US dollar both ran on the classical gold standard, each fixed against gold, which served as the numeraire. It self-adjusted through the price-specie-flow mechanism: a trade surplus brought in gold, expanded the money supply, raised prices, and cut exports, while a deficit did the reverse. In the 1930s, protectionism, deflation, and hyperinflation brought chaos, world trade collapsed by over 50 percent, and the gold standard was dropped. Late in World War II, Keynes and White designed the Bretton Woods system of fixed parities with periodic realignments, adopted by 44 countries in 1944. Chronic inflation broke it by the early 1970s, and most industrial countries moved to flexible rates under the Smithsonian Agreements, an outcome Milton Friedman had argued for in the 1950s. Rates then moved far more than expected, revealing the FX market as a highly liquid, forward-looking asset market driven by investment and speculation. In 1979 the European Economic Community adopted the limited-flexibility Exchange Rate Mechanism (ERM); a speculative attack drove the United Kingdom out of it in September 1992, and in 1999 most of Western Europe (without Switzerland or the United Kingdom) launched the euro, gaining price transparency but giving up national monetary policy.

A taxonomy of regimes

In its 30 April 2008 review, the International Monetary Fund (IMF) sorted regimes into eight categories, from the least to the most flexible. The classifications are somewhat arbitrary and change over time, but they show how much law and policy, not just market forces, shape FX pricing.

The IMF taxonomy lists eight regime types; dollarization and monetary union are the two arrangements with no separate legal tender.

A few distinctions are worth holding onto. A currency board differs from simple dollarization in that the authority keeps the profit, called seigniorage, from paying little on its monetary base while earning a market return on reserve assets; under dollarization that seigniorage goes to the country whose currency is used. Hong Kong SAR has run a currency board since 1983 at a parity of HKD/USD = 7.80, with the rate held in a narrow convertibility zone by buying and selling US dollars just off parity. A simple fixed parity differs from a currency board in two ways: there is no legislative commitment to the peg, and the reserve target is discretionary, so the central bank keeps traditional functions such as lender of last resort but market participants know the parity can be adjusted or abandoned. Even nominally floating currencies see occasional intervention, as in the 1985 Plaza Accord, in which Japan and Germany appreciated their currencies against the US dollar. The most heavily traded currencies (the US dollar, the euro, the yen, the pound, the Swiss franc, plus the Canadian and Australian dollars) are generally regarded as free floating, drawing only occasional intervention.

A household that outspends its income has to borrow or sell what it owns, and a country importing more than it exports is in the same bind: it must borrow from foreigners or sell them assets to cover the gap. A country that exports more than it imports does the opposite, lending to foreigners or buying their assets. A trade deficit or surplus is therefore matched exactly by an offsetting capital account surplus or deficit. Whatever shifts the trade balance leaves an equal and opposite mark on the capital account: the effect of exchange rates on trade is mirrored in capital flows, and one cannot move without the other.

A standard macroeconomic identity links the trade balance to saving and spending decisions:

A trade surplus (X greater than M) must surface as a fiscal surplus (T greater than G), as private saving that outruns investment (S greater than I), or as both. Reading a fiscal surplus as government saving, a trade surplus tells us the nation saves beyond what its own investment needs and parks the extra in claims on other economies. A trade deficit says the opposite: saving falls short of investment, so the country runs down its net claims abroad. The identity does not say which assets change hands or in what currency; it only demands that asset prices and exchange rates settle where investors are content to hold every financial asset outstanding.

When investors expect a rate to move, they try to sell the currency expected to depreciate and buy the one expected to appreciate. That latent flow of capital must be either offset by a move in the trade balance or damped by shifts in asset prices and exchange rates. Because spending and saving decisions and goods prices move slowly, while financial decisions and asset prices move fast, most of the adjustment happens in financial markets. In a fixed-rate regime the central bank offsets private capital flows to hold the peg, and adjustment shows up in other prices, typically interest rates. In a floating regime the main adjustment is a rapid rate move that dampens expectations of further movement. The upshot is that capital flows, actual and potential, are the primary determinant of exchange rates in the short to intermediate term, while trade flows grow more important over the long term as spending, saving, and goods prices adjust.

Governments curb the movement of capital in and out for many reasons, from jobs and regional development to strategic or defense aims. A number of countries insist on sign-off before foreigners may invest at home or citizens may invest overseas. Inward controls can cap how much may be invested and restrict ownership of strategic industries such as defense and telecommunications. Outward controls can restrict repatriation of capital, interest, profits, royalties, and license fees, and can limit citizens investing abroad, especially where foreign currency is scarce.

Why open capital flows are usually beneficial

Economists generally favor free movement of capital because it directs funds to where they earn the highest return. Inflows let a country invest in productive capacity faster than domestic savings alone would allow, raising the achievable growth rate. Long-term investment by foreign firms can bring not only capital but new technology, skills, and management practices, plus spillover benefits for local suppliers who receive training and closer working relationships. Greater competition can push domestic firms to become more efficient, although some firms unable to compete may be forced out of the market.

When governments restrict capital

In a macroeconomic crisis, mobile capital can turn into capital flight, especially where inflows were mostly short-term portfolio flows into liquid stocks and bonds rather than foreign direct investment. Capital restrictions are then often paired with fixed exchange rate targets, because under perfect capital mobility a government cannot meet both domestic and external policy goals with monetary and fiscal tools alone. Limiting capital flow gives the authorities a separate handle on the external balance while other tools address domestic objectives.

Modern capital controls trace to World War I, when belligerents restricted capital outflows to finance the war effort, keep capital at home, tax wealth, and hold down government borrowing costs. Since then, mostly developing countries have used outflow controls to raise revenue or to allocate credit domestically without risking flight. Put broadly, a capital restriction is any measure that caps or reroutes the flow of capital. It can take the form of taxes on international investment returns or transactions, mandatory reserve requirements (for example forcing a foreign depositor to place a share of the inflow with the central bank at zero interest for a minimum period), quantity limits such as ceilings or authorization requirements on foreign borrowing, or administrative approval of certain asset transactions.

Restrictions carry costs. Effective enforcement can require heavy administration, especially as rules are broadened to close loopholes; controls can postpone needed policy adjustments and slow private-sector adaptation; and, most damaging, they can create negative market perceptions that make foreign funding more costly and harder to obtain. An IMF study of effectiveness found that inflow controls needed comprehensive coverage and forceful enforcement to work, that their effects were hard to separate from other policies, and that outflow controls during crises produced mixed results, offering only temporary relief to some countries while shielding others long enough to restructure.