International trade is largely carried out by exchanging the value of goods and services in the terms of one currency for that of another. Such trade does not make use of a barter trade system and as such we need to account for the fact that virtually every country (or group of countries that form a monetary union) has its own monetary unit or currency. In addition, there are also a number of international transactions that are purely financial in nature, such as trading activity, which may involve the exchange of different currencies.

Hence, the need for a foreign exchange market, where the various national currencies can be exchanged (bought and sold) for one another. The foreign exchange market, like any other concept of market used in economic theory, is not a precise physical place. It is actually formed by banks, brokers and other authorized representatives to whom economic agents apply to buy and sell the various currencies. These representatives of the market are linked by telephone, telex, computer, or by other means and thus the market has an international rather than national dimension.

Over time, the globalization process has led to spectacular growth in the volumes traded on the foreign exchange market.

1 Organisation of the Foreign Exchange Market

In this section, we describe the institutional structure that allows corporations, banks, international investors, and tourists to convert one currency into another.

The foreign exchange market consists of a number of different aspects that includes the interbank market, which comprises of the wholesale part of the foreign exchange market where banks manage inventories of currencies. This is a very large, diverse, over-the-counter (OTC) market, not a physical trading place where buyers and sellers gather to agree on a price to exchange currencies. Traders, who are employees of financial institutions in the major financial cities around the world, deal with each other via computer or over the phone, with back-office confirmations of transactions occurring at a later point in time.

Since most transactions on the interbank market are made up of large trades with values of $5 million or more, most retail investors and small businesses do not have direct access to this part of the foreign exchange market. As a result, many in need of foreign exchange deal with small regional banks or branches of banks that quote less advantageous rates than those that prevail on the interbank market. Retail investors also participate in the foreign exchange markets through their stockbrokers, who place orders on derivative markets that trade in futures and options contracts.

Large multinational corporations (such as Anglo American) and very large money-management firms (such as Investec) can directly access the foreign exchange interbank market. In addition, some multinational companies also have their own foreign exchange trading desks to manage these transactions. An important recent development that has affected the market is the rapid growth in electronic trading, both in the interbank market (through an electronic brokering) and on the retail side of the market.

To contextualise the various components of the foreign exchange market, we can make use of Figure 1, which is taken from Bekaert and Hodrick (2012). This diagram shows that all the foreign exchange transactions are centered around the interbank market.

Figure 1: Structure of the Foreign Exchange Market

The foreign exchange market operates 24 hours per day because the major financial centres where currencies are traded have different geographic locations. The first market activity during the course of the daily trading schedule takes place in Sydney (Australia) and Wellington (New Zealand), and it is quickly followed by trading in Tokyo and Osaka (Japan), Hong Kong, and Singapore. An abrupt decline in activity then occurs after 4 hours of trading, when most of the participants in these markets take their midday meal. Market activity picks up again during the afternoon of the Eastern Asian trading session. As activity from Hong Kong and Singapore declines during the later part of the afternoon, activity from Frankfurt and London intensifies. Other foreign exchange centres in Europe include Zurich, Copenhagen and Paris. The aggregate volume of trading activity increases further when New York markets are trading and it eventually declines when the business day in New York draws to a close. Other trading centres in the United States include Chicago and Los Angeles. This cycle is then repeated when the Eastern Australasian markets open once again.

It is worth noting that the amount of activity on the foreign exchange markets is the largest and most liquid market in the world, measured by dollar volume of trade. This volume has increased rapidly since the 1970s. For example, during 1973 the estimated daily volume of currency trading was between $10 and $20 billion. By the late 1980s, daily volume had increased to around $500 billion. By September 1993, the estimated daily volume in all currencies had increased to over $1 trillion, and by 2004, it had increased to almost $2 trillion.1 The Bank for International Settlements (BIS) estimated that daily trading volume in April 2010 was $3.9 trillion and in 2016 it was estimated to be $5.1 trillion per day.2 This dollar volume of trade dwarfs the corresponding dollar volume of transactions on stock markets such as the New York Stock Exchange (NYSE), where average daily dollar volume was roughly $87 billion in 2015. In addition, it also dwarfs the annual Gross Domestic Product (GDP) of a country like South Africa, which at the end of 2016 was about \(\pm\) $232 billion.3

1.1 Communications and funds transfers

The enormous volume of trade in the foreign exchange market requires an extensive communication network between traders and a sophisticated settlement system to transfer payments in different currencies between the buyers and sellers of different currencies. Traders are able to obtain information that is provided by major commercial distributors such as Reuters and Bloomberg. The traders are then able to contact each other, to obtain actual prices and negotiate deals. In addition, they could approach a foreign exchange broker to broker a deal, or they can trade on an electronic brokerage system, where quotes on a screen are transactable. When a trade is agreed upon, banks communicate and transfer funds electronically, using systems such as the Society of Worldwide Interbank Financial Telecommunications (SWIFT), which confirm trades and facilitate payment.

As Cross-Currency transactions may involve the simultaneous exchange of currencies, there is a risk that only one leg of the transaction may be completed, due to the possibility that parties use different systems in different countries that operate out of different time zones. This is known as cross-currency settlement risk, or Herstatt risk. Recently, foreign exchange dealers, encouraged by the BIS, have developed a number of practices to limit settlement risk. These measures include: firstly, banks now have strict limits on the amount of transactions they are willing to settle with a single counterparty on a given day. Secondly, banks have started to engage in a variety of netting arrangements, in which they agree to wire the net traded amounts only at the end of a trading day. Thirdly, settlement risk is eliminated if the exchange of the two monies occur simultaneously in a process known as payment versus payment (PvP).

More recently, we have witnessed the foundation of the Continuous Linked Settlement (CLS) Bank, which is owned by the world’s largest financial groups. CLS is the largest multi-currency cash settlement system, eliminating settlement risk for over half of the worlds foreign exchange payment instructions and its members include central banks, large commercial banks and other large corporations. The CLS daily settlement cycle operates with settlement and funding occurring during a five-hour window when all real time gross settlement systems are able to make and receive payments. This enables simultaneous settlement of the payments on both sides of a foreign exchange transaction. Each member holds a single multicurrency account with CLS, which has a zero balance at the start and the end of trading day. The settlement of the payment instructions and the associated payments are final and irrevocable.

1.2 Foreign exchange dealers and brokers

Traditionally, the main participants in the foreign exchange market are the commercial banks, investment banks, and brokerage firms in the major financial cities around the world. Traders at these banks and firms function as foreign exchange dealers, simultaneously “making-a-market” in several currencies. These market makers stand ready to buy and sell the currencies in which they specialize. By standing ready to transact with retail customers or other dealers, they provide liquidity to the market, which makes it easier and less costly to match buyers and sellers. When there are large numbers of buyers and sellers, markets are usually very liquid, and transaction costs are low. The foreign exchange markets for the major currencies of the world, such as the markets for the U.S. dollar, the euro, the Japanese yen, and the British pound, are among the most liquid markets in the world.

Foreign exchange brokers do not attempt to buy low and sell high. Instead, brokers satisfy the role of a financial intermediary. They match buyers and sellers but do not put their own money at risk. They then receive a brokerage fee on their transactions. Foreign exchange brokers typically have many lines of communication open to various foreign exchange dealers, and they provide information to dealers on the best available prices. Foreign exchange dealers often use these brokers to unwind very large positions in a particular currency in order to preserve their anonymity.

While the brokers continue to play an important role in foreign exchange trading, a large part of the brokering business now happens through computerized trading systems. In the early 1990s, Reuters (now Thomson Reuters), a large financial information provider, and Electronic Brokering Service (EBS), started by a consortium of 12 banks but now part of the interdealer broker ICAP, launched the first anonymous electronic brokering systems for trading spot foreign exchange. Trading is carried out through a network of linked computer terminals among the participating foreign exchange dealers. Currency prices are displayed on computer screens, and deals are completed by keystroke or by automatic deal matching within the system. Before a trade gets executed, either the systems check for mutual credit availability between the initiator of the deal and the counterparty of the deal; or each counterparty must have its creditworthiness pre-screened.

Trading in each major currency pair has over time become very highly concentrated on only one of the two systems. The top two traded currency pairs, euro-dollar and dollar-yen, trade primarily on the EBS, whereas the third most traded pair, the pound-dollar, trades primarily on Reuters. As a result, the exchange rates on EBS and Reuters for these particular currency pairs have become the reference rates for dealers across the world. When EBS allowed institutional investors and hedge funds onto its platform in 2005, it confirmed a trend towards the blurring of the distinction between the interbank and retail side of the foreign exchange market, ushered in by the emergence of electronic trading.

The central banks of different governments around the world periodically participate in the foreign exchange market as they try to influence the foreign exchange value of their currencies. Other participants include multinational corporations, which need to exchange currencies to conduct their international trade; institutional investors buying and selling foreign securities; hedge funds speculating on currency movements; and smaller domestic banks that service firms or individuals wanting to exchange currencies.

The use of internet platforms lead to the development of a number of Electronic Foreign Exchange Trading (eFX) systems. This is one of the fastest growing segments of the foreign exchange market, representing more than 30% of all trading volume (and more than 50% of the trades in spot markets) during 2010.

1.3 The costs of transactions

As there are many different types of monetary instruments which may be used to exchange currencies, the respective rate of exchange may be dependent on the form of currency. For example the exchange rate for cash may be different to the rate that is charged on travellers cheques and these rates may also differ to the rate that pertains to bank transfer orders. These differences depend on various elements that include the costs of transferring funds and the carrying costs in a broad sense. For example, if a bank keeps foreign currency in the form of banknotes in its vaults (rather than in the form of demand deposits), it not only loses interest but also has to bear the custody costs.

In most instances, a foreign exchange trader is typically responsible for buying and selling a particular currency or a small group of currencies. They would usually hold an inventory or portfolio of positions in those currencies. One reason for this activity in the interbank market is that foreign exchange traders at one bank use foreign exchange traders at other banks to adjust their portfolios in response to transactions that arise from their customers in the corporate market (to maintain an inventory of a particular currency or portfolio position). The repeated passing of inventory imbalances among dealers has been dubbed “hot potato trading” and may be one reason for the large trade volumes that we see on the interbank market.4

Ultimately, traders in the interbank market try to buy and sell various foreign currencies with the goal of generating profits. To do so, they quote two-way prices. The bid rate is the rate at which they want to buy a base currency, and the ask rate is the rate at which they sell base currency. The difference between these two rates is known as the bid-ask or bid-offer spread. The bid price is always less than the ask price because the trader bids for the base currency when they buy it and asks a price for the base currency when they sell it.

The competitive nature of the foreign exchange market and the growth of electronic trading have greatly compressed bid-ask spreads over the last decade. In the interbank market, spreads for major currencies have become negligible. In addition, even in the customer market, bid-ask spreads are now also within 5 pips (for the major currencies of large transaction sizes) where a pip is jargon for the fourth decimal point in a currency quote.

By way of example, if the midpoint of the USDZAR quote is R13.50, and we have a spread of 20 pips then the ask rate is R13.501 and the bid rate is R13.499. Therefore, 1 pip reflects 1/100 of a South African cent in this case. However, to get an idea of transaction costs involved in trading currencies, its usually better to express the bid-ask spread in percentage points, where the spread is computed as:

\[\begin{eqnarray} \nonumber \text{Percentage spread} &=& \frac{(\text{ask} - \text{bid})}{\text{midpoint}} \\ \nonumber &=& \frac{(13.501 - 13.499)}{13.5} \\ \nonumber &=& 0.00015 \end{eqnarray}\]

Hence, the bid-ask spread represents 0.015% or 1.5 basis points. The cheapest currencies to trade are the major ones like the EUR versus USD (with spreads sometimes as low as 1 pip), the GBP versus USD, and the USD versus JPY. In addition, the “G10” currencies that include the AUD, CHF, CAD, NZD, SEK, and NOK are also highly liquid and typically trade at less than 10 pips. Emerging market currencies trade at higher spreads.

As these differences are usually very small, the examples that are contained below largely assume that there is only one exchange rate for each foreign currency. Therefore, we largely exclude the effects of the bid-ask spread on transactions.

2 The exchange rate

The exchange rate is a price of one currency in terms of another. Since two currencies are involved, there are two different ways of giving the quotation of foreign exchange. One is called the direct or price quotation system, and defines the exchange rate as the number of units of domestic currency per unit of foreign currency. When taking South Africa as the home country, we have, say, R 13.24 per US dollar, R 0.12 per Japanese yen, R 15.50 per euro, etc. This amounts to defining the exchange rate as the it price of foreign currency in terms of domestic currency.

The alternative method makes use of the indirect or volume quotation system, and it defines the exchange rate as the number of units of foreign currency per unit of domestic currency, which is obviously the reciprocal of the previous case. Once again, after taking South Africa as the home country, we would have 0.076 US dollar per South African rand, 8.37 Japanese yens per South African rand, 0.065 euros per South African rand, etc. With this definition the exchange rate is the price of domestic currency in terms of foreign currency.

In what follows we will adopt the direct or price quotation system, and a good piece of advice to the reader is to always ascertain which definition is being (explicitly or implicitly) adopted, to avoid confusion.

Rather than write out the full name of these currencies, contractual parties use abbreviations that were established by the International Organization for Standardization (called ISO from the Greek word equal). For example, the notation for the U.S. dollar is USD, the euro is EUR, and the South African rand is ZAR.

If it takes 13.24 South African rand to purchase 1 U.S. dollar, we can write ZAR13.24 = USD1, or ZAR13.24/USD. Note that we treat the slash symbol (/) as a divisor in a ratio to indicate the amount of the first currency that is necessary to purchase one unit of the second currency. In practice, you will encounter foreign exchange quotations, such as USDZAR, in which the first currency in the quote is the base currency and the second currency is the numerator currency or “quote currency”. In other words, if you type USDZAR into Google, it will return the price of the U.S. dollar in terms of South African rands, or how many South African rands you can buy with one U.S. Dollar (which is makes use of a direct quotation). Indeed, most practitioners in South Africa, including dealers and brokers, would make use of the term USDZAR, when quoting the exchange rate for these currencies.

When quoting against the U.S. dollar, we make use of two additional terms, where the European quote refers to the amount of foreign currency needed to buy dollars. This is due to the fact that most former European currencies, such as the Deutsche mark and the French franc, were quoted this way relative to the dollar. Similarly, the phrase American quote refers to the dollar price of a foreign currency, which is the number of dollars it takes to purchase one unit of the foreign currency.

The focus on the U.S. dollar exchange rates in comparison to other currencies of the world is warranted because the U.S. dollar is a vehicle currency, which implies that it is actively used in many international financial transactions around the world. Exchange rates between two currencies that do not involve the dollar are often called cross-rates. Although there appears to be a trend toward more cross-rate transactions, it is estimated that 85% of all transactions have the dollar as one side. Some analysts think the euro, which replaced 11 different currencies in Europe in 1999, may someday replace the dollar as a vehicle currency. In fact, a BIS survey of foreign exchange activity from 2010 suggests that about 40% of all trades involved the euro during that year.

Triangular arbitrage is a process that keeps cross-rates (such as euros per British pound) in line with exchange rates quoted relative to the U.S. dollar. A trader can conduct a triangular arbitrage in many ways. For example, a trader might start with euros, buy pounds with the euros, then simultaneously sell those pounds for dollars and sell those dollars for euros. In other words, instead of exchanging just two currencies, the trader exchanges three (hence the term “triangular” arbitrage). If the number of euros the trader has at the end of these three transactions is greater than the number of euros at the beginning, then the trader has earned a profit.

If such transactions can be done profitably, the trader can generate pure arbitrage profits to earn risk-free profits. Obviously, in perfectly competitive financial markets, one would not be able to earn arbitrage profits for very long. If the euro price of the pound were not equal to the euro price of the U.S. dollar multiplied by the U.S. dollar price of the pound, arbitrage activity would immediately restore equality between the quoted cross-rate and the cross-rate implied by the two dollar quotes.

2.1 Rates of change

As exchange rates are relative prices, it would imply that there are always two ways to describe the same situation. For example, if we are told that currency \(x\) has depreciated, we may infer from the point of view of those in country \(x\) that the currency has experienced some form of exchange rate depreciation. This means that value of currency \(x\) has declined in terms of foreign currency, such that a greater amount of currency \(x\) is required to buy one unit of foreign currency. Similarly, this would also imply that a smaller amount of foreign currency is required to buy one unit of currency \(x\). Hence, the foreign currency would have appreciated, relative to the domestic currency in this case.

The terms appreciation and depreciation are typically used to describe changes in flexible exchange rates, where they are able to fluctuate freely in response to changes in demand and supply. This terminology would also usually apply to those exchange rates that are subject to a managed framework, where the central bank of a country would make a number of transactions on the foreign exchange market to alter the value of the domestic currency. This framework is currently employed by the South African Reserve Bank, although it would only usually enter the foreign exchange market in extreme cases.

Alternatively, the government authorities of a particular country may fix or peg the exchange rate of their currency relative to a foreign currency. Discrete changes in the values of exchange rates under such a fixed exchange rate system are termed devaluations or revaluations, for corresponding downward and upward movements.

The rate of appreciation or depreciation of one currency relative to another can be calculated as the percentage rate of change of the exchange rate. For example, if the USDZAR exchange rate moves from R14 to R15 per dollar, then we can show that the % change is:

\[\begin{eqnarray} \nonumber \% \; \text{Change in the rand} &=& \frac{(\text{New exchange rate} - \text{Old exchange rate})}{\text{Old exchange rate}} \\ \nonumber &=& \frac{(15 - 14)}{14} \\ \nonumber &=& 7.14 \% \end{eqnarray}\]

In this case, the South African rand is said to have depreciated by 7.14%.

2.2 Continuously Compounded Rates of Appreciation

In certain instances, we might also want to describe the rate of appreciation using a compounded monthly basis, while still expressing the percentage change at an annual rate. For example, if we want to calculate the annualised value of appreciation of the dollar then we would make use of the following expression:

\[\begin{eqnarray} \nonumber (\text{USDZAR}\; 14.00) \Big[1 + ( \eta /12) \Big]^{12} = (\text{USDZAR} \; 15) \end{eqnarray}\]

where \(\eta\) is the rate of appreciation. To solve for \(\eta\), we first divide both sides by USDZAR 14.00 and then take the (1/12) power of each side:

\[\begin{eqnarray} \nonumber \Big[1 + ( \eta /12) \Big] &=& \Big[(\text{USDZAR} \; 15) \; / \; (\text{USDZAR} \; 14) \Big]^{1/12} \\ \nonumber \eta &=& \Big(\big[(\text{USDZAR} \; 15) \; / \; (\text{USDZAR} \; 14) \big]^{1/12} -1 \Big) \times 12 \\ \nonumber &=& 0.0692 \end{eqnarray}\]

This provides an answer is \(\eta =\) 0.0692, or an annualized compound monthly rate of appreciation of the dollar of 6.92%. To calculate the annualized compound monthly rate of depreciation of the rand, relative to the dollar, we could assume in this example that the equivalent values for the ZARUSD are \(1/14\) and \(1/15\), respectively. This would imply that the depreciation of the rand could then be calculated as

\[\begin{eqnarray} \nonumber (\text{ZARUSD} \; 0.07142857) \Big[1 - (\delta / 12)\Big]^{12} = (\text{ZARUSD} \; 0.0\dot{6}) \end{eqnarray}\]

where after using similar steps, \(\delta =\) 0.0688 or 6.88%.

If we drive the compounding interval to one that is infinitely small, we will eventually be able to derive the continuous rate of appreciation of the dollar relative to the rand. Continuous compounding uses the symbol \(e\), which represents the base of natural logarithms, where \(e \sim\) 2.71828.

In this case the annualized continuously compounded rate of appreciation of the dollar is the value of \(\eta\) that satisfies

\[\begin{eqnarray} \nonumber (\text{USDZAR} \; 14.00)e^{\eta} = \text{USDZAR} \;15.00 \end{eqnarray}\]

To solve for the value of \(\eta\), we take the natural logarithm of both sides of the equation and find5

\[\begin{eqnarray} \nonumber \eta = \log (\text{USDZAR} \; 15.00) - \log (\text{USDZAR} \; 14.00) = 0.069 \end{eqnarray}\]

or 6.9%. Similarly, to calculate the annualized continuously compounded rate of depreciation of the rand, we could again assume that the equivalent values for the ZARUSD are \(1/14\) and \(1/15\), respectively. To perform this calculate we need to find the value of \(\delta\) that satisfies

\[\begin{eqnarray} \nonumber (\text{ZARUSD} \; 0.07142857)e^{-\delta} = \text{ZARUSD} \; 0.0\dot{6} \end{eqnarray}\]

To solve for the value of \(\delta\), we take the natural logarithm of both sides of the equation and find

\[\begin{eqnarray} \nonumber \delta = - \Big[ \log (\text{ZARUSD} 0.0\dot{6}) - \log (\text{ZARUSD} 0.07142857) \Big] = 0.069 \end{eqnarray}\]

or 6.9%. With continuous compounding, the rates of appreciation of the dollar and depreciation of the rand are the same.

In finance, we often encounter the natural logarithm because of continuous compounding. Banks usually quote interest rates at annual rates such as 10%, and they specify a compounding period, which might be annual, semiannual, monthly, daily, or even continuously. We know that the more often the bank credits interest to our account, the more money we will have at the end of a year because we will earn interest on previously credited interest.

For example, if the quoted interest rate is 10%, at the end of 1 year, we will have the following amounts, depending on the compounding interval:

Compounding Interval \(\hspace{3cm}\) Amount in 1 Year
Annual [1 + 0.1] = 1.1
Semiannual [1 + (0.1/2)]2 = 1.1025
Quarterly [1 + (0.1/4)]4 = 1.1038
Monthly [1 + (0.1/12)]12 = 1.1047
Daily [1 + (0.1/365)]365 = 1.10516

Table 1: Compounding Rates

The return from continuously compounding at an interest rate, \(i\), is obtained by taking the limit of the following expression as the number of compounding intervals, \(n\), goes to infinity:

\[ \lim_{n \rightarrow \infty} \big(1 + (i / n) \big)^n = e^i \]

where \(e\) is the base for natural logarithms. In our example, a 10% annual interest rate that is continuously compounded will provide an amount of money after 1 year of \(e\)0.1 = 1.10517.6 Note that practitioners sometimes write \(exp( i )\) rather than \(e^i\), where both imply that they are looking to raise the number \(e\) to the \(i\)th power.

Raising the number \(e\) to a power tells you the rate at which your principal grows when it is continuously compounded at a particular interest rate. Hence, the exponential function can be used to describe other growth rates, such as rates of appreciation or depreciation of currencies and rates of inflation. For example, if the value of the South African rand relative to the US dollar were to grow at a continuous rate of \(5\%\) during 2017, then for a starting value of R13.50, the exchange rate at the end of the year would be

\[\begin{eqnarray} \nonumber (\text{USDZAR}_{end}) = (R 13.50) \; e^{0.05} = R 14.19 \end{eqnarray}\]

2.3 The real exchange rate

In general, real magnitudes are obtained from the corresponding nominal magnitudes after we eliminate the change that is due to price changes, which can be done in a variety of ways. In the case of exchange rates the question is slightly more complicated, due to the fact that the exchange rate is intrinsically a nominal concept. This has lead to development of a number of different definitions for the real exchange rate.7

The oldest notion of real exchange rate, and the one that we will use in what follows is defined as the nominal exchange rate multiplied by the ratio of the price levels. For example, if we would want to calculate the real exchange rate for the South Africa rand relative to the euro, we would use the expression:

\[\begin{eqnarray} \text{RS}_{t,\text{R} / {\unicode{0x20AC}}} =\frac{S_{t,{\text{R}}/ {\unicode{0x20AC}}} \times P_{t,{\unicode{0x20AC}}}}{P_{t,{\text{R}}}} \tag{2.1} \end{eqnarray}\]

where \(S_{t,{\text{R}}/ {\unicode{0x20AC}}}\) is the spot exchange rate, while \(P_{t,\text{R}}\) and \(P_{t,\text{euro}}\) are the price levels in South Africa and in Europe. This definition is obviously linked to the concept of Purchasing Power Parity (PPP) and it will be discussed at a later point in the course.

According to another definition, the real exchange rate is the (domestic) relative price of tradable and nontradable goods,

\[\begin{eqnarray} \nonumber \text{RS}_{t, \text{T/NT}} = \frac{P_{t,\text{T}}}{P_{t,\text{NT}}} \end{eqnarray}\]

The rationale of this definition is that, in a two-sector (tradables-nontradables) model, the balance of trade depends on \(P_{\text{T}}/P_{\text{NT}}\) because this relative price measures the opportunity cost of domestically producing tradable goods, and the ex ante balance of trade depends on the ex ante excess supply of tradables. According to the Harrod-Balassa-Samuelson hypothesis, the long run PPP holds only for traded goods, and the real exchange rate in the long run is a function of the relative productivity of traded to non-traded goods in the home and foreign countries.

A widely held opinion is that the real exchange rate should give a measure of the external competitiveness of a country’s goods (if non-traded goods are also present, only tradables should be considered). However, it is by no means obvious which index should be taken. In the simple exportables-importables model of trade the real exchange rate reduces to the notion of terms of trade defined in the theory of international trade, namely

\[\begin{eqnarray} \text{RS}_{t, \text{TOT}} = \pi = \frac{P_{t,\text{EX}}}{S_t \times P_{t,\text{M}}} \tag{2.2} \end{eqnarray}\]

where \(P_{t, \text{EX}}\) represents export prices (in terms of domestic currency), \(P_{t, \text{M}}\) import prices (in terms of foreign currency), and \(S_t\) the nominal exchange rate of the country under consideration. From the point of view of the consumer, \(\pi\) represents the relative price of foreign and domestic goods on which (in accordance with standard consumer theory) demand will depend. From the point of view of the country as a whole, \(\pi\) represents the amount of imports that can be obtained in exchange for one unit of exported good (or the amount of exports required to obtain one unit of imports). Therefore, an increase in \(\pi\) is also defined as an improvement in the terms of trade, as it means that a greater amount of imports can be obtained per unit of export (or, equivalently, that a smaller amount of exports are required per unit of import).

The terms of trade \(\pi\) can serve both the domestic and the foreign consumer (country) for the relevant price-comparison, because the prices of domestic and foreign goods are expressed in domestic currency (for which we have a reciprocal). In addition, it should also be clear to see how (2.1) is related to (2.2) since exports are an essential part of domestic output and prices, \(P_{t,\text{EX}} = P_{t,{\text{R}}}\) and imports are related to foreign output and prices \(P_{t,\text{M}} = P_{t,{\unicode{0x20AC}}}\)

Another definition of the real exchange rate takes the ratio of unit labour costs at home, \(W_{h}\), to unit labour costs abroad, \(W_{f}\), expressed in a common monetary unit through the nominal exchange rate, \(S_t\). Hence, this expression would take the form

\[\begin{eqnarray} \nonumber RS_t = S_t \left( \frac{W_{f}}{W_{h}} \right) \end{eqnarray}\]

Note that the real exchange rate in this case is defined such that an increase (decrease) would imply an improvement (deterioration) in the external competitiveness of domestic goods. In fact, ceteris paribus, a decrease in domestic unit labour costs relative to foreign unit labour costs is reflected (in both perfectly and imperfectly competitive markets) in a decrease of the relative price of domestic goods with respect to foreign goods. The same result is obtained when, at given \(W_{f}/W_{h}\), the exchange rate depreciates (i.e. the nominal exchange rate \(S_t\) increases).

2.4 The effective exchange rate

The concept of effective exchange rate must not be confused with the real exchange rate, since the effective exchange rate can be nominal or real.

While the (nominal or real) exchange rate involves two currencies only, it may be desirable to have an idea of the overall external value of a currency, namely with respect to the rest of the world (or a subset of it, for example with that countries main trading partners) and not only with respect to another country’s currency. The presence of flexible exchange rates makes it difficult to ascertain the behaviour of the external value of a currency. In fact, in a flexible regime a currency may simultaneously depreciate with respect to one (or more) foreign currency and appreciate with respect to another (or several others).

In such a situation it is necessary to have recourse to an index number, in which the bilateral exchange rates of the currency under consideration with respect to all other currencies enter with suitable weights. This index is called an effective exchange rate. Let us begin with the nominal effective exchange rate, which is given by the formula

\[\begin{eqnarray} \nonumber R_{E,i}= \sum_{j=1, j\neq i}^{n} w_{j}R_{j,i}, \hspace{1cm} \sum_{j=1, j\neq i}^{n}w_{j}=1 \end{eqnarray}\]

The respective variables in this expression refer to:

  • \(R_{E,i} =\) (nominal) effective exchange rate of currency \(i\)
  • \(R_{ji} =\) nominal exchange rate of currency \(i\) with respect to currency \(j\)
  • \(w_{j}=\) weight given to currency \(j\) in the construction of the index; by definition, the sum of the weights equals one.

Usually the effective exchange rate is given as an index number with a base of 100 and presented in such a way that an increase (decrease) in it means an appreciation (depreciation) of the currency under consideration with respect to the other currencies as a whole.

Unfortunately it is not possible to determine the weights unambiguously: this is an ambiguity inherent in the very concept of index number. Many effective exchange rates thus exist in theory; however, the weights are related to the share of the foreign trade of country \(i\) with country \(j\) in the total foreign trade of country \(i\). Effective exchange rates are computed and published by the International Monetary Fund, central banks and private institutions.

If we carry out the same operation with the use of real rather than nominal bilateral exchange rates we shall of course obtain the real effective exchange rate. This will give a measure of the overall competitiveness of domestic goods on world markets rather than with respect to another country’s goods.

3 Foreign exchange contracts

Many different types of trades can be conducted in the foreign exchange market. The spot market has traditionally been a very important part of the market for the immediate exchanges of currencies. Another part of the interbank foreign exchange market involves trade in swaps and forward contracts. These transactions involve the exchange of currencies in the future. Together the spot and the forward market constitute the foreign exchange market. Another important type of contract involves derivative securities that would include foreign currency futures and options. These aspects of the market are considered below.

3.1 The spot market

The spot market for the foreign exchange of currencies relates to the exchange of two currencies on the spot (i.e. for immediate delivery). In most instances, this transaction would be reflected by a change in the demand deposits of banks that denominated in the two currencies. It is worth noting that although there may be the possibility of earning arbitrage profits on the spot market, due to time lags or other possible frictions, the reality is that these discrepancies are incredibly small, which would imply that it is unlikely that any potential profit will exceed the transaction costs.

A number of important rules are associated with trades on the interbank spot market. For example, when the trade involves the U.S. dollar, these rules dictate that spot contracts are settled after 2 business days. Hence, the payment of one currency and receipt of the other currency occurs after 2 business days. One business day is necessary because of the back-office paperwork involved in any financial transaction. The second day is then needed because of the time zone differences around the world. All transactions need to be settled within this time frame.8

3.2 The forward market

The forward foreign exchange market (or the forward market, for short) is the market for exchanges of currencies in the future. The main function of the forward exchange market is to allow economic agents engaged in international transactions (whether these are commercial or financial) to cover themselves against the exchange rate risk that is due to possible future variations in the spot exchange rate. If it were the case that the spot exchange rate were permanently and rigidly fixed, then the agent who is about to make or receive a foreign currency payment in the future (or, more generally, who has liabilities and/or assets in foreign currency) does not incur any exchange risk. However, as most exchange rates will change over time, as is usually the case, an exchange rate risk will usually arise.

From the point of view of the agent who has to make a future payment in foreign currency (for example, an importer who will have to pay in three months’ time for the goods imported now), the risk is that the exchange rate will depreciate at the time of the payment. In this case he will have to pay out a greater amount of domestic currency to purchase the required amount of foreign currency. From the point of view of the agent who is to receive a future payment in foreign currency (for example, an exporter who will be paid in three months time for the goods exported now) the risk is that the exchange rate will have appreciated at the time of the payment, in which case he will get a smaller amount of domestic currency from the sale of the given amount of foreign currency.

Naturally the agent who has to make a future payment in foreign currency will benefit from an appreciation of the domestic currency and, similarly, a depreciation will benefit the agent who is to receive a future payment in foreign currency. However, if we exclude the category of speculators and only consider economic agents who are risk averse, then such an agent will assign a greater weight on the eventuality of a loss than on a gain derived from future variations in the exchange rate. This would be achieved with the aid of some form of hedging transaction. For example, to hedge against currency risks, the agent would enter into an additional contract that provides profits when the underlying transaction on the spot market results in a realised pecuniary loss. To evaluate the costs and benefits of hedging for a future transaction involving foreign currencies, the hedging party must have some way to quantify the degree of uncertainty it faces about future spot exchange rates. It accomplishes this by figuring out the likelihood of observing various ranges for future exchange rates.

The need for some form of hedging transaction arises whenever the liabilities and assets of an agent that are denominated in foreign currency are not matched. Of course, this exact balance must hold for each of the foreign currencies that the agent holds. These positions are usually described with financial jargon, where to have no open position in foreign exchange would imply that you have neither a long position (more assets than liabilities in foreign currency) nor a short position (more liabilities than assets in foreign currency).

3.3 Derivative contracts

Foreign currency futures and options can be used for hedging or speculative purposes. Since the profits and losses earned on futures and option contracts, as well as those earned on forward contracts, depend on how the spot exchange rate evolves over time, all these instruments are considered derivative securities. Derivative securities are securities whose values depend on the values of other, more basic underlying variables-in this case, the spot exchange rate.

As with other instruments in the foreign exchange market, much of the trade in futures contracts and options are conducted by banks. Commercial and investment banks deal aggressively in foreign currency options in order to meet the demands of their corporate and institutional customers, who use them to hedge their foreign exchange risks. In addition to banks, hedge funds and other investors trade foreign currency futures and options purely for speculative purposes (in order to earn a profit).

4 Recent Developments on the Foreign Exchange Markets

The main factor behind the large increase in volumes is undoubtedly the globalization process, which led to increased cross-border trades in goods, services, and securities, all requiring transactions in the foreign exchange market. In addition, the speculative activities and high-volume, high-frequency trading by hedge funds have also played an increasingly important role. More recently, non-dealer financial institutions, including smaller banks, institutional investors and hedge funds, have grown into the largest and most active counterparty segment. Hence, the once clear-cut divide between inter-dealer and customer trading has dissipated as technological change has increased the connectivity of participants, bringing down search costs. This implies that the foreign exchange market is now largely characterised by heterogeneous agents.

Rime and Schrimpf (2013) make use of the 2013 Triennial Central Bank Survey of Foreign Exchange and Derivatives Market Activity to show that the increase in trading activity has risen fairly evenly across instruments, since 2010. That said, spot was the largest contributor to turnover growth, accounting for 41% of the turnover rise, such that the volume is now similar to that of foreign exchange swaps. Turnover in foreign exchange OTC derivatives, such as forwards (up 43%) and foreign exchange options (up 63%), also grew strongly, albeit from a lower base.9

Figure 2: Global FX Market Turnover (by instrument)

The survey also suggested that trading in currency markets is increasingly dominated by financial institutions outside the dealer community (“other financial institutions” in the survey terminology). Transactions with non-dealer financial counterparties grew by 48% to $2.8 trillion per day in 2013, up from $1.9 trillion in 2010, and accounted for roughly two thirds of the rise in the total. These non-dealer financial institutions are very heterogeneous in their trading motives, patterns and horizons. They include lower-tier banks, institutional investors (e.g. pension funds and mutual funds), hedge funds, high-frequency trading (HFT) firms and official sector financial institutions (e.g. central banks or sovereign wealth funds).

The flip-side of the relative growth in the non-dealer financial institution transactions is the relative decline in the inter-dealer share, which is down to 39% (much lower than the 63% in the late 1990s). The primary reason is that major dealing banks net more trades internally. Due to higher industry concentration, top-tier dealers are able to match more customer trades directly on their own books. This reduces the need to offload inventory imbalances and hedge risk via the traditional inter-dealer market.

A significant fraction of transactions with the relatively heterogeneous non-dealer financial customers is with lower-tier banks. While these “non-reporting banks” tend to trade smaller amounts and/or only sporadically, in aggregate they account for roughly one quarter of global foreign exchange volumes. These smaller banks do not engage in market-making, but mostly serve as clients of the large foreign exchange dealing banks. As they find it hard to rival dealers in offering competitive quotes in major currencies, they concentrate on niche business and mostly exploit their competitive edge. Like dealers, they extensively trade short-tenor foreign exchange swaps (less than one week), which are commonly used for short-term liquidity management.

Figure 3: Trading of non-dealer financials

The most significant non-bank foreign exchange market participants are professional asset management firms, captured under the two labels “institutional investors” (e.g. mutual funds, pension funds and insurance companies) and “hedge funds”. The two groups each accounted for about 11% of turnover, as is shown in Figure 3.

\(\hspace{.5cm}\) Share \(\hspace{.5cm}\) \(\hspace{.5cm}\) Top 10 \(\hspace{.5cm}\)
Deutsche Bank (Autobahn) \(\hspace{1cm}\) 36 7
UBS (FX Trader) 22 7
Barclays Capital (BARX) 12 7
Citi (Velocity) 6 7
JPMorgan (MorganDirect) 3 3
Goldman Sachs (REDI) 3 5
RBS (SmartPrime) 3 6
HSBC (HSBCnet FXHub) 2 7
Credit Suisse (PrimeTrade FX) 2 3
Morgan Stanley (Passport) 2 2

Table 2: Average market share and years with top-10 ranking for single-bank platforms

Though many small banks have withdrawn from market making in the most liquid currencies, they ensure their customers have access to liquidity by providing the single-bank trading platforms of major banks under their own name. This practice, called “white labelling”, has numerous advantages for the major banks. First, it lets them view the small banks’ trading flows, and to extract any relevant information, without the expense of evaluating each counterparty’s creditworthiness. It also provides major banks with a new revenue stream, supporting the investments required to develop their single-bank trading systems. King, Osler, and Rime (2011) show the extent of white labelling, which is displayed in Table 2, where the combined market share of the three largest single-bank trading platforms, at roughly 70%, is double the overall market share of the three largest banks, at roughly 35%.10

4.1 Trading of Emerging Market Currencies

The trend towards more active foreign exchange trading by non-dealer financial institutions and a concentration in financial centres is particularly visible for emerging market (EM) currencies, where the trading of EM currencies is increasingly conducted from offshore centres. The ease of trading minor currencies has improved significantly as a result, and transaction costs for trading in EM currencies, measured by bid-ask spreads, have steadily declined and converged to almost the levels for developed currencies. This is shown in Figure 4, where the relative bid-ask spreads of advanced and emerging market economies are displayed.11

Figure 4: Relative Bid-Ask Spreads

As liquidity in EM currencies has improved, these markets have attracted the attention of international investors. Naturally, this has also boosted the share of key EM currencies in total global turnover, from 12% in 2007 to 17% in 2013. The strong growth is particularly visible in the case of the Mexican peso, whose market share now exceeds that of several well established advanced economy currencies. Another case is the renminbi, where most of the 250% growth is due to a surge in offshore trading. In addition, China set itself to promote more international use of its currency and introduced offshore renminbi (CNH) in 2010.

4.2 Retail trading in the FX market

In the late 1990s, foreign exchange trading was mainly the domain of large corporations and financial institutions. Banks charged small “retail” investors prohibitively high transaction costs, as their trades were considered too tiny to be economically interesting. This changed when retail-oriented platforms (e.g. FXCM and OANDA) started offering online margin brokerage accounts to private investors around 2000, streaming prices from major banks and EBS. Their business model was to bundle many small trades together and lay them off in the inter-dealer market. With trade sizes now much larger, dealers were willing to provide liquidity to such “retail aggregators” at attractive prices.

Retail foreign exchange trading has since grown quickly. New breakdowns collected in the 2013 Triennial show that retail trading accounted for 3.5% and 3.8% of total and spot turnover, respectively. The largest retail volumes in absolute terms are in the United States and Japan. That said, Japan, which has a very active retail segment, is clearly one of the biggest players in this spot market. In April 2013, retail trading in Japan accounted for 10% and 19% of total and spot, respectively.

Retail investors differ from institutional investors in their foreign exchange trading patterns. They tend to trade directly in relatively illiquid currency pairs rather than via a vehicle currency (as shown in Figure 5). In addition, it is also worth noting that the boundaries of what constitutes a retail trade are also becoming more blurred. Regulatory changes (e.g. leverage limits for margin brokerage accounts for private investors) in countries such as the United States have slowed growth in the retail segment and led some platforms to target themselves towards professional investors (e.g. small hedge funds). Furthermore, the recent poor returns on popular strategies, such as momentum and carry trades, suggest that growth in the retail segment may have slowed.

Figure 5: Retail trading in the FX market

5 Crytocurrencies

Less than 10 years after their inception, cryptocurrencies have emerged from obscurity to attract intense interest on the part of businesses and consumers, as well as central banks and other authorities. They garner attention because they promise to replace trust in long-standing institutions, such as commercial and central banks, with trust in a new, fully decentralised system founded on the blockchain and related distributed ledger technology (DLT).

We consider the potential usefulness of cryptocurrencies and discusses the economic limitations inherent in the decentralised creation of trust which they entail. For the trust to be maintained, honest network participants need to control the vast majority of computing power, each and every user needs to verify the history of transactions and the supply of the cryptocurrency needs to be predetermined by its protocol. Trust can evaporate at any time because of the fragility of the decentralised consensus through which transactions are recorded. Not only does this call into question the finality of individual payments, it also means that a cryptocurrency can simply stop functioning, resulting in a complete loss of value. Moreover, even if trust can be maintained, cryptocurrency technology comes with poor efficiency and vast energy use. Cryptocurrencies cannot scale with transaction demand, are prone to congestion and greatly fluctuate in value. Institutions such as the Bank of International Settlements (BIS) suggest that the decentralised technology of cryptocurrencies, however sophisticated, is a poor substitute for the solid institutional backing of money.

That said, the underlying technology could have promise in other applications, such as the simplification of administrative processes in the settlement of financial transactions. However, such uses of this technology would still need to be tested.

5.1 The elusive promise of decentralised trust

While most modern-day transactions occur through means ultimately supported by central banks, over time a wide range of public and private payment means has emerged. These can be best summarised by a taxonomy characterised as the “money flower”, which is shown in Figure 6.

Figure 6: The money flower: a taxonomy of money

The money flower distinguishes four key properties of moneys: the issuer, the form, the degree of accessibility and the payment transfer mechanism. The issuer can be a central bank, a bank or nobody, as was the case when money took the form of a commodity. Its form can be physical, e.g. a metal coin or paper banknote, or digital. It can be widely accessible, like commercial bank deposits, or narrowly so, like central bank reserves. A last property regards the transfer mechanism, which can be either peer-to-peer, or through a central intermediary, as for deposits. Money is typically based on one of two basic technologies: so called “tokens” or accounts. Token-based money, for example banknotes or physical coins, can be exchanged in peer-to-peer settings, but such exchange relies critically on the payee’s ability to verify the validity of the payment object - with cash, the worry is counterfeiting. By contrast, systems based on account money depend fundamentally on the ability to verify the identity of the account holder.

Do cryptocurrencies deliver what they promise? Or will they end up as short-lived curiosities?

In order to answer these questions, it is necessary to define them more precisely, to understand their supporting technology and to examine the associated economic limitations.

5.2 A new petal in the money flower?

Cryptocurrencies aspire to be a new form of currency and promise to maintain trust in the stability of their value through the use of technology. They consist of three elements. First, a set of rules (the “protocol”), computer code specifying how participants can transact. Second, a ledger storing the history of transactions. And third, a decentralised network of participants that update, store and read the ledger of transactions following the rules of the protocol. With these elements, advocates claim, a cryptocurrency is not subject to the potentially misguided incentives of banks and sovereigns.

In terms of the money flower taxonomy, cryptocurrencies combine three key features. First, they are digital, aspiring to be a convenient means of payment and relying on cryptography to prevent counterfeiting and fraudulent transactions. Second, although created privately, they are no one’s liability, i.e. they cannot be redeemed, and their value derives only from the expectation that they will continue to be accepted by others. This makes them akin to a commodity money (although without any intrinsic value in use). And, last, they allow for digital peer-to-peer exchange.

Compared with other private digital moneys such as bank deposits, the distinguishing feature of cryptocurrencies is digital peer-to-peer exchange. Digital bank accounts have been around for decades. And privately issued “virtual currencies” - e.g. as used in massive multiplayer online games like World of Warcraft - predate cryptocurrencies by a decade. In contrast to these, cryptocurrency transfers can in principle take place in a decentralised setting without the need for a central counterparty to execute the exchange.

5.3 Distributed ledger technology in cryptocurrencies

The technological challenge in digital peer-to-peer exchange is the so-called “double-spending problem”. Any digital form of money is easily replicable and can thus be fraudulently spent more than once. Digital information can be reproduced more easily than physical banknotes. For digital money, solving the double-spending problem requires, at a minimum, that someone keep a record of all transactions. Prior to cryptocurrencies, the only solution was to have a centralised agent do this and verify all transactions.

Cryptocurrencies overcome the double-spending problem via decentralised record-keeping through what is known as a distributed ledger. The ledger can be regarded as a file (think of a massive spreadsheet) that starts with an initial distribution of cryptocurrency and records the history of all subsequent transactions. An up-to-date copy of the entire ledger is stored by each user (this is what makes it “distributed”). With a distributed ledger, peer-to-peer exchange of digital money is feasible: each user can directly verify in their copy of the ledger whether a transfer took place and that there was no attempt to double-spend.

While all cryptocurrencies rely on a distributed ledger, they differ in terms of how the ledger is updated. One can distinguish two broad classes, with substantial differences in their operational setup, as displayed in Figure 7.

Figure 7: Centralised ledger and permissioned/permissionless decentralised ledgers

One class is based on “permissioned” DLT. Such cryptocurrencies are similar to conventional payment mechanisms in that, to prevent abuse, the ledger can only be updated by trusted participants in the cryptocurrency - often termed “trusted nodes”. These nodes are chosen by, and subject to oversight by, a central authority, which could be the firm that developed the cryptocurrency. Thus, while cryptocurrencies based on permissioned systems differ from conventional money in terms of how transaction records are stored (decentralised versus centralised), they share with it the reliance on specific institutions as the ultimate source of trust.

In a much more radical departure from the prevailing institution-based setup, a second class of cryptocurrencies promises to generate trust in a fully decentralised setting using “permissionless” DLT. The ledger recording transactions can only be changed by a consensus of the participants in the currency: while anybody can participate, nobody has a special key to change the ledger.

The concept of permissionless cryptocurrencies was laid out for the case of Bitcoin in a white paper by an anonymous programmer (or group of programmers) under the pseudonym Satoshi Nakamoto, who proposed a currency based on a specific type of distributed ledger, the “blockchain”. The blockchain is a distributed ledger that is updated in groups of transactions called blocks. Blocks are then chained sequentially via the use of cryptography to form the blockchain. This concept has been adapted to countless other cryptocurrencies.

Blockchain-based permissionless cryptocurrencies have two groups of participants: “miners” who act as bookkeepers and “users” who want to transact in the cryptocurrency. At face value, the idea underlying these cryptocurrencies is simple: instead of a bank centrally recording transactions (Figure 8, left-hand panel), the ledger is updated by a miner and the update is subsequently stored by all users and miners (right-hand panel).

Figure 8: Valid transactions in a centralised ledger/bank account and in a permissionless cryptocurrency

Underlying this setup, the key feature of these cryptocurrencies is the implementation of a set of rules (the protocol) that aim to align the incentives of all participants so as to create a reliable payment technology without a central trusted agent. The protocol determines the supply of the asset in order to counter debasement - for example, in the case of Bitcoin, it states that no more than 21 million bitcoins can exist. In addition, the protocol is designed to ensure that all participants follow the rules out of self-interest, i.e. that they yield a self-sustaining equilibrium. Three key aspects are the following.

First, the rules entail a cost to updating the ledger. In most cases, this cost comes about because updating requires a “proof-of-work”. This is mathematical evidence that a certain amount of computational work has been done, in turn calling for costly equipment and electricity use. Since the proof-of-work process can be likened to digging up rare numbers via laborious computations, it is often referred to as mining. In return for their efforts, miners receive fees from the users - and, if specified by the protocol, newly minted cryptocurrency.

Second, all miners and users of a cryptocurrency verify all ledger updates, which induces miners to include only valid transactions. Valid transactions need to be initiated by the owners of funds and must not be attempts to double-spend. If a ledger update includes an invalid transaction, it is rejected by the network and the miner’s rewards are voided. The verification of all new ledger updates by the network of miners and users is thus essential to incentivise miners to add only valid transactions.

Third, the protocol specifies rules to achieve a consensus on the order of updates to the ledger. This is generally done by creating incentives for individual miners to follow the computing majority of all other miners when they implement updates. Such coordination is needed, for example, to resolve cases where communication lags lead to different miners adding conflicting updates - i.e. updates that include different sets of transactions.

With these key ingredients, it is costly - though not impossible - for any individual to forge a cryptocurrency. To successfully double-spend, a counterfeiter would have to spend their cryptocurrency with a merchant and secretly produce a forged blockchain in which this transaction was not recorded. Upon receipt of the merchandise, the counterfeiter would then release the forged blockchain, i.e. reverse the payment. But this forged blockchain would only emerge as the commonly accepted chain if it were longer than the blockchain the rest of the network of miners had produced in the meantime. A successful double-spend attack thus requires a substantial share of the mining community’s computing power. Conversely, in the words of the original Bitcoin white paper, a cryptocurrency can overcome the double-spending problem in a decentralised way only if “honest nodes control a majority of [computing] power”.

5.4 Assessing the economic limitations of permissionless cryptocurrencies

Cryptocurrencies such as Bitcoin promise to deliver not only a convenient payment means based on digital technology, but also a novel model of trust. Yet delivering on this promise hinges on a set of assumptions: that honest miners control the vast majority of computing power, that users verify the history of all transactions and that the supply of the currency is predetermined by a protocol. Understanding these assumptions is important, for they give rise to two basic questions regarding the usefulness of cryptocurrencies. First, does this cumbersome way of trying to achieve trust come at the expense of efficiency? Second, can trust truly and always be achieved?

As the first question implies, a key potential limitation in terms of efficiency is the enormous cost of generating decentralised trust. One would expect miners to compete to add new blocks to the ledger through the proof-of-work until their anticipated profits fall to zero. Individual facilities operated by miners can host computing power equivalent to that of millions of personal computers. At the time of writing, the total electricity use of bitcoin mining equalled that of mid-sized economies such as Switzerland, and other cryptocurrencies also use ample electricity (Figure 9, left-hand panel). Put in the simplest terms, the quest for decentralised trust has quickly become an environmental disaster.

Figure 9: Energy consumption and scaling issues

However, the underlying economic problems go well beyond the energy issue. They relate to the signature property of money: to promote “network externalities” among users and thereby serve as a coordination device for economic activity. The shortcomings of cryptocurrencies in this respect lie in three areas: scalability, stability of value and trust in the finality of payments.

First, cryptocurrencies simply do not scale like sovereign moneys. At the most basic level, to live up to their promise of decentralised trust cryptocurrencies require each and every user to download and verify the history of all transactions ever made, including amount paid, payer, payee and other details. With every transaction adding a few hundred bytes, the ledger grows substantially over time. For example, at the time of writing, the Bitcoin blockchain was growing at around 50 GB per year and stood at roughly 170 GB. Thus, to keep the ledger’s size and the time needed to verify all transactions (which increases with block size) manageable, cryptocurrencies have hard limits on the throughput of transactions (Figure 9, centre panel).

A thought experiment illustrates the inadequacy of cryptocurrencies as an everyday means of payment (Figure 9, right-hand panel). To process the number of digital retail transactions currently handled by selected national retail payment systems, even under optimistic assumptions, the size of the ledger would swell well beyond the storage capacity of a typical smartphone in a matter of days, beyond that of a typical personal computer in a matter of weeks and beyond that of servers in a matter of months. But the issue goes well beyond storage capacity, and extends to processing capacity: only supercomputers could keep up with verification of the incoming transactions. The associated communication volumes could bring the internet to a halt, as millions of users exchanged files on the order of magnitude of a terabyte.

Another aspect of the scalability issue is that updating the ledger is subject to congestion. For example, in blockchain-based cryptocurrencies, in order to limit the number of transactions added to the ledger at any given point in time, new blocks can only be added at pre-specified intervals. Once the number of incoming transactions is such that newly added blocks are already at the maximum size permitted by the protocal, the system congests and many transactions go into a queue. With capacity capped, fees soar whenever transaction demand reaches the capacity limit (Figure 10). And transctions have at times remained in a queue for several hours, interruptin the payment process. This limits cryptocurrencies’ usefulness for day-to-day transactions such as paying for a coffee or a conference fee, not to mention for wholesale payments. Thus, the more people use a crytocurrency, the more cumbersome payments become. This negates an essential property of present-day money: the more people use it, the stronger the incentive to use it.

Figure 10: Transaction fees over time and in relation to transaction throughput

The second key issue with cryptocurrencies is their unstable value. This arises from the absence of a central issuer with a mandate to guarantee the currency’s stability. Well run central banks succeed in stabilising the domestic value of their soveriegn currency by adjusting the supply of the means of payment in line with transaction demand. They do so at high frequency, in particular during times of market stress but also during normal times.

This contrasts with a crytocurrency, where generating some confidence in its value requires that supply be predetermined by a protocol. This prevents it from being supplied elastically. Therefore, any fluctuation in demand translates into changes in valuation. This means that crytocurrencies’ valuations are extremely volatile (Figure 11, left-hand panel). And the inherent instability is unlikely to be fully overcome by better protocols or financial engineering, as exemplified by the experience of the Dai cryptocurrency. While engineered to be fixed to the US dollar at a rate of one to one, it reached a low of $0.72 just a few weeks after its launch in late 2017. Other cryptocurrencies designed to have a stable value have also fluctuated substantially (centre panel).

Figure 11: Volatility of select cryptocurrencies and number of cryptocurrencies

This outcome is not coincidental. Keeping the supply of the means of payment in line with transaction demand requires a central authority, typically the central bank, which can expand or contract its balance sheet. The authority needs to be willing at times to trade against the market, even if this means taking risk onto its balance sheet and absorbing a loss. In a decentralised network of cryptocurrency users, there is no central agent with the obligation or the incentives to stabilise the value of the currency: whenever demand for the cryptocurrency decreases, so does its price.

Further contributing to unstable valuations is the speed at which new cryptocurrencies - all tending to be very closely substitutable with one another - come into existence. At the time of writing, several thousand existed, though proliferation makes reliable estimates of the number of outstanding cryptocurrencies impossible (Figure 11, right-hand panel). Recalling the private banking experiences of the past, the outcome of such liberal issuance of new moneys is rarely stability.

The third issue concerns the fragile foundation of the trust in cryptocurrencies. This relates to uncertainty about the finality of individual payments, as well as trust in the value of individual cryptocurrencies.

In mainstream payment systems, once an individual payment makes its way through the national payment system and ultimately through the central bank books, it cannot be revoked. In contrast, permissionless cryptocurrencies cannot guarantee the finality of individual payments. One reason is that although users can verify that a specific transaction is included in a ledger, unbeknownst to them there can be rival versions of the ledger. This can result in transaction rollbacks, for example when two miners update the ledger almost simultaneously. Since only one of the two updates can ultimately survive, the finality of payments made in each ledger version is probabilistic.

The lack of payment finality is exacerbated by the fact that cryptocurrencies can be manipulated by miners controlling substantial computing power, a real possibility given the concentration of mining for many cryptocurrencies (Figure 12, left-hand panel). One cannot tell if a strategic attack is under way because an attacker would reveal the (forged) ledger only once they were sure of success. This implies that finality will always remain uncertain. For cryptocurrencies, each update of the ledger comes with an additional proof-of-work that an attacker would have to reproduce. Yet while the probability that a payment is final increases with the number of subsequent ledger updates, it never reaches 100%.

Figure 12: Mining concentration and bitcoin value during a temporary fork

Not only is the trust in individual payments uncertain, but the underpinning of trust in each crytocurrency is also fragile. This is due to “forking”, which is a process whereby a subset of crytocurrency holders coordinate on using a new version of the ledger and protocol, while others stick to the original one. In this way, a cryptocurrency can split into two subnetworks of users. While there are many recent exmaples, an episode on 11 March 2013 is noteworthy because - counter to the idea of achieving trust by decentralised means - it was undone by centralised coordination of the miners. On that day, an erroneous software update led to incompatibilities between one part of the Bitcoin network mining on the legacy protocol and another part mining using an updated one. For several hours, two separate blockchains grew; once news of this fork spread, the price of bitcoin tumbled by almost a third (Figure 12, right-hand panel). The fork was ultimately rolled back by a coordinated effort whereby miners temporarily departed from protocol and ignored the longest chain. But many transactions were voided hours after users had believed them to be final. This episode shows just how easily crytocurrencies can split, leading to significant valuation losses.

An even more worrying aspect underlying such episodes is that forking may only be symtomatic of a fundamental shortcomming: the fragility of the decentralised consensus involved in updating the ledger, and with it, of the underlying trust in the crytocurrency. Theoretical analysis suggests that coordination on how the ledger is updated could break down at any time, resulting in a complete loss of value.

Overall, decentralised cryptocurrencies suffer from a range of shortcomings. The main inefficiencies arise from the extreme degree of decentralisation: creating the required trust in such a setting wastes huge amounts of computing power, decentralised storage of a transaction ledger is inefficient and the decentralised consensus is vulnerable. Some of these issues might be addressed by novel protocols and other advances. But others seem inherently linked to the fragility and limited scalability of such decentralised systems. Ultimately, this points to the lack of an adequate institutional arrangement at the national level as the fundamental shortcoming.

5.5 Beyond the bubble: making use of distributed ledger technology

While cryptocurrencies do not work as money, the underlying technology may have promise in other fields. A notable example is in low-volume cross-border payment services. More generally, compared with mainstream centralised technological solutions, DLT can be efficient in niche settings where the benefits of decentralised access exceed the higher operating cost of maintaining multiple copies of the ledger.

To be sure, such payment solutions are fundamentally different from cryptocurrencies. A recent non-profit example is the case of the World Food Programme’s blockchain-based Building Blocks system, which handles payments for food aid serving Syrian refugees in Jordan. The unit of account and ultimate means of payment in Building Blocks is sovereign currency, so it is a “cryptopayment” system but not a cryptocurrency. It is also centrally controlled by the World Food Programme, and for good reason: an initial experiment based on the permissionless Ethereum protocol resulted in slow and costly transactions. The system was subsequently redesigned to run on a permissioned version of the Ethereum protocol. With this change, a reduction of transaction costs of about 98% relative to bank-based alternatives was achieved.

Permissioned cryptopayment systems may also have promise with respect to small-value cross-border transfers, which are important for countries with a large share of their workforce living abroad. Global remittance flows total more than $540 billion annually (Figure 13, left-hand and centre panels). Currently, forms of international payments involve multiple intermediaries, leading to high costs (right- hand panel). That said, while cryptopayment systems are one option to address these needs, other technologies are also being considered, and it is not clear which will emerge as the most efficient one.

Figure 13: Indicators of the volume and cost of remittances

More important use cases are likely to combine cryptopayments with sophisticated self-executing codes and data permission systems. Some decentralised cryptocurrency protocols such as Ethereum already allow for smart contracts that self-execute the payment flows for derivatives. At present, the efficacy of these products is limited by the low liquidity and intrinsic inefficiencies of permissionless cryptocurrencies. But the underlying technology can be adopted by registered exchanges in permissioned protocols that use sovereign money as backing, simplifying settlement execution. The added value of the technology will probably derive from the simplification of administrative processes related to complex financial transactions, such as trade finance. Crucially, however, none of the applications require the use or creation of a cryptocurrency.

6 Conclusion

Trading activity in the foreign exchange market reached an all-time high of $5.3 trillion in April 2013, 35% higher than in 2010. The results of the 2013 Triennial Survey confirm a trend in the market already seen in prior surveys: first, a growing role of non-dealer financial institutions (smaller banks, institutional investors and hedge funds); second, a further internationalisation of currency trading and at the same time a rising concentration in financial centres; and lastly, a fast-evolving market structure driven by technological innovations that accommodates the diverse trading needs of market participants.

New and more granular breakdowns introduced in the 2013 Triennial allow a more detailed analysis of these developments. With more detailed information on non-dealer financial institutions, the linkages between their trading motives and foreign exchange turnover growth can be better understood. The once clear-cut two-tier structure of the market, with separate inter-dealer and customer segments, no longer exists. At the same time, the number of ways the different market participants can interconnect has increased significantly, suggesting that search costs and trading costs are now considerably reduced. This has paved the way for financial customers to become liquidity providers alongside dealers. Hence, financial customers contribute to increased volumes not only through their investment decisions, but also by taking part in a new hot potato trading process, where dealers no longer perform an exclusive role.

7 References

Bekaert, Geert, and Robert J. Hodrick. 2012. International Financial Management. New York: Prentice Hall.

Chinn, M. D. 2006. “A Primer on Real Effective Exchange Rates: Determinants, Overvaluation, Trade Flows and Devaluation.” Open Economies Review 17: 15–143.

Hinkle, E. L., and P. Montiel. 1999. Exchange Rate Misalignment: Concepts and Measurement for Developing Countries. Oxford: Oxford University Press for the World Bank.

International Settlements, Bank of. 2016. “Triennial Central Bank Survey of Foreign Exchange and Otc Derivatives Markets in 2016.” Bank of International Settlements.

King, Michael R., Carol Osler, and Dagfinn Rime. 2011. “Foreign Exchange Market Structure, Players and Evolution.” Working Paper 2011/10. Norges Bank.

Lipschitz, L., and D. McDonald. 1992. “Real Exchange Rates and Competitiveness: A Clarification of Concepts, and Some Measurements for Europe.” Empirica 19: 37–69.

Lyons, R. 1997. “A Simultaneous Trade Model of the Foreign Exchange Hot Potato.” Journal of International Economics 42: 275–98.

Marsh, I. W., and S. P. Tokarick. 1996. “An Assessment of Three Measures of Competitiveness.” Weltwirtschaftliches Archiv 132: 700–732.

Rime, Dagfinn, and Andreas Schrimpf. 2013. “The Anatomy of the Global Fx Market Through the Lens of the 2013 Triennial Survey.” BIS Quarterly Review, 27–43.

  1. One trillion is equal to 1 000 000 000 000.

  2. This is down from $5.4 trillion in April 2013. This represents the first decline in FX spot trading activity, since 2001. However, activity in FX derivatives has continued to increase. Trading in OTC interest rate derivatives averaged $2.7 trillion per day in April 2016, up from $2.3 trillion in April 2013 (International Settlements (2016)).

  3. When using an exchange rate of R13.24 to the US dollar.

  4. This term was coined by Lyons (1997).

  5. The natural logarithm is denoted “\(\log\)” in this calculation.

  6. The natural logarithm of 1.10517 is 0.1 because raising 2.71828 to the 0.1 power is 1.10517.

  7. For surveys see, e.g., Lipschitz and McDonald (1992); Marsh and Tokarick (1996); Hinkle and Montiel (1999); Chinn (2006).

  8. Several exceptions to the 2-business-day rule are noteworthy. First, for exchanges between the U.S. dollar and the Canadian dollar or the Mexican peso, in these cases the rule is 1 business day. Second, if the transaction involves the dollar and the first of the 2 days is a holiday in the United States (but not in the other settlement centre), the first day is counted as a business day for settlement purposes. Third, Fridays are not part of the business week in most Middle Eastern countries, although Saturdays and Sundays are. Hence, Middle Eastern currencies transacted on Wednesday settle on Saturday, not on Friday.

  9. The surge in options reflects the period of intense yen trading in April 2013, where almost half of the options traded that month were linked to JPY/USD. At the time, hedge funds expressed their directional views via the options market, where they accounted for 24% of turnover.

  10. The table show the average market share (in percentage points) since 2004 up to 2010. The column “Top 10” states how many years, out of the seven possible years, that a bank’s single-bank platform has ranked Top 10. Source: Euromoney FX Survey.

  11. Advanced economy currencies: Australian dollar, Canadian dollar, Danish krone, euro (from 2004), Japanese yen, New Zealand dollar, Norwegian krone, Swedish krone, Swiss franc, pound sterling and US dollar. Emerging market currencies: Brazilian real, Chinese renminbi (from 2007), New Taiwan dollar, Hong Kong dollar, Hungarian forint, Indian rupee, Korean won, Mexican peso, Polish zloty, Russian rouble, Singapore dollar, South African rand and Turkish lira.