Should we really worry about currency wars?
In September 2010, Brazil’s former finance minister, Guido Mantega, made headlines when he accused the Federal Reserve, the European Central Bank and the Bank of England of engaging in a currency war. The complaint was that easy monetary policy was driving down the value of the dollar, the euro and the pound, at the expense of his country and those like it. More recently, similar charges have been levied against Japan: namely, that the Bank of Japan’s extraordinary balance sheet expansion is aimed at driving down the value of the yen, damaging the country’s trading partners and competitors.
At first blush, these “beggar thy neighbor” charges might seem plausible. Many emerging market countries survive on exports, so an appreciation of their currency makes their products more expensive to foreign buyers. This naturally depresses sales, putting a drag on growth. Cheaper imports add to that drag. Finally, with policy interest rates in the advanced economies at the zero bound, the exchange rate channel has become more important than usual as a means of transmitting monetary policy stimulus.
But the charge was and is difficult to justify. First, the aggregate economic impact of currency changes is limited. Second, there is no evidence that advanced economies are manipulating their exchange rates to achieve a competitive advantage.
Let’s start with the impact of exchange rates. Their influence on a country depends on how open its economy is to the outside world. The table below lists a broad array of countries ranked by a common measure of trade openness: exports plus imports of goods and services as a fraction of GDP. For small economies that engage in enormous amounts of trade, where they import parts and export assembled products, this number is frequently above 100%. Hong Kong SAR and Singapore are the top examples. The case of Luxembourg is explained by the fact that it provides financial services to much of Europe.
| Hong Kong SAR, China | 450 | Mexico | 66 | |
| Singapore | 368 | United Kingdom | 66 | |
| Luxembourg | 325 | Canada | 62 | |
| Ireland | 191 | South Africa | 62 | |
| Netherlands | 168 | Italy | 59 | |
| Malaysia | 162 | France | 57 | |
| Vietnam | 157 | India | 55 | |
| Austria | 111 | China | 52 | |
| Korea, Rep. | 110 | Indonesia | 50 | |
| Germany | 98 | Australia | 43 | |
| Switzerland | 94 | Japan | 31 | |
| Sweden | 91 | United States | 30 | |
| Kenya | 78 | Argentina | 30 | |
| Chile | 68 | Brazil | 27 | |
| Norway | 68 | Sudan | 22 |
Looking down the table, we see that European countries tend to be fairly open, but large economies like the United States, Japan, or the Euro Area as a whole, do not. (The ECB’s estimate of extra-euro area imports plus exports relative to GDP is a modest 39%.) As it happens, Brazil stands out. Of the 150 or so countries for which we have reliable data, Sudan is the only country in the world that is more closed than Brazil. Even the notoriously self-sufficient United States is more open!
What are we to make of this? Large economies are relatively closed and benefit less than others from a depreciation of their exchange rates.
Given the closed character of the Brazilian economy, why might the Brazilian government have been so concerned about its exchange rate?
We can think of three potential motives: (1) concern about the impact on inflation; (2) concern about the impact on aggregate output; and (3) concern about the impact on favored economic sectors. The first two are straightforward economic motivations. The third one reflects concern about specific winners and losers in the economy, so we think of it as political.
To see which of these explanations holds up, let’s focus on the Brazilian case. We start with their nominal effective exchange rate (NEER, see following chart). Unlike the bilateral exchange rates quoted each moment in currency markets, the NEER takes account of Brazil’s trade patterns, weighting more heavily those countries with which Brazil trades more. The plunge in 1998-1999 reflects Brazil’s currency crisis. When Minister Mantega made his accusation, Brazil’s currency had appreciated to the highest level since the crisis.
Brazil: Nominal Effective Exchange Rate (2010=100)
Source: BIS and FRED
It is important to understand that a nominal currency appreciation lowers inflation. The reason is that when the value of a country’s currency rises relative to that of its trading partners, it becomes cheaper to purchase imported goods. This both drives down the price that domestic consumers pay for imports and induces local producers of import-competing goods to lower their prices. This impact can be substantial – in very open advanced economies like Germany, roughly half of the exchange rate movement is passed through to import prices and about one-quarter to export prices. For Brazil, the numbers are even bigger – more than 80% of an exchange rate movement goes into import prices and nearly 40% into export prices. The Japanese case is somewhere in between. (See the analysis here.)
Yet, inflation can’t account for Minister Mantega’s concern. Having lived through a hyperinflation in the late 1980s and 1990s, when prices were doubling every 125 days, Brazilians are very sensitive to anything that could drive inflation up. (At the end of 1995, the Brazilian price level was 700 million times what it was at the end of 1985). However, in 2010, Brazilian inflation was a relatively modest 5.0%, close to the central bank’s 4½% inflation target and below the 6.7% average of the previous decade.
What about the impact of the currency on aggregate economic output? To analyze that effect, consider Brazil’s “real effective exchange rate” (REER, see chart below). Unlike the NEER, the REER also takes account of relative prices across countries. Even if the NEER is stable, Brazil’s REER will rise when it experiences higher inflation than the weighted average of its trading partners, and it will fall when Brazil’s inflation rate is relatively lower. Consequently, the REER is often used as an indicator of a country’s evolving trade competitiveness. (Both the NEER and the REER are computed by the BIS, and are described here.)
Brazil: Real Effective Exchange Rate (2010=100)
Source: BIS and FRED.
When Minister Mantega made his accusation, Brazil’s REER was higher than it has been in the 20-year history of the index. In fact, by late 2010, the real value of Brazil’s currency (called the real) was 10% above where it stood prior to the financial crisis.
What impact does such a real appreciation have on production and growth? Here again, the picture turns out to be benign. In 2010, Brazil grew at more than 5%. Since then, economic growth has slowed significantly, so that today it is even slightly negative. But over the intervening four-year period, the Brazilian REER depreciated. Perhaps it should fall further, but if the Brazilian central bank wishes to secure price stability, the closed nature of the economy suggests that it should pay relatively little attention to the exchange rate in adjusting monetary policy.
Because it is difficult to find a macroeconomic explanation for Minister Mantega’s reaction, we are drawn to a political one. That is, Brazil’s policy concern over the value of the real reflects the distributive (rather than the aggregate) impact of a change in its value. When thinking about trade, it is often helpful to ask who gains and who suffers. In many economies, whether advanced or emerging, the existence of barriers to trade reflects the political influence of those who are able to obtain government protection from foreign competition. Such barriers typically generate sizable benefits for a small number of producers and their employees, while imposing costs on millions of consumers that are modest individually but substantial in the aggregate.
Brazil is a big commodity exporter. The combination of minerals, petroleum and agricultural products accounts for nearly 40% of exports, or more than 5% of Brazilian GDP. The prices of these commodities are set in U.S. dollars. In September 2010, it took only 1.7 Brazilian reals to purchase one dollar, whereas three years earlier, it took more than 2 reals. That means that an equivalent dollar value of commodities would net exporters 15% less in domestic currency.
As we already mentioned, import and import-competing goods prices were falling, benefiting Brazilian consumers at large. So, what we have is a typical case where the losses are concentrated in an easily identifiable and politically influential group, while the gains are dispersed broadly across the population as a whole. In a circumstance like that, it is pretty clear who gets heard.
Does this analysis mean that cries of “currency war” are misleading and should be ignored by policymakers? There can be exceptions, but the answer is usually yes.
As for policymakers in Japan, they are targeting domestic price stability, just as U.S. and euro area central bankers do. The depreciation of the yen versus other currencies is a natural side-effect of these efforts. The yen’s weakness mostly reflects relative growth prospects, especially versus the United States.
The bottom line: when you hear someone complain about a “currency war,” beware the motives of those seeking protection from competition.