Federal Reserve Economic Data

The FRED® Blog

Dollar strength and the trade balance

Has the exchange rate shifted the U.S. trade balance?

The exchange rate is the price of one country’s currency in terms of another country’s currency. For example, an exchange rate of 100 Japanese yen to the U.S. dollar means that you can exchange a single U.S. dollar for 100 Japanese yen. The exchange rate is important for international trade because changes in exchange rates often alter the prices of imported and exported goods between countries. For example, if the U.S. dollar appreciates with respect to the Japanese yen, Japanese consumers have to give up more Japanese yen to buy the same dollar value of U.S. goods exported to Japan. In other words, appreciation of the dollar implies that U.S. goods become more expensive to foreigners. On the other hand, appreciation of the dollar tends to make goods imported from other countries cheaper for U.S. consumers. Because of these changes in relative prices, appreciation of the dollar tends to increase imports and decrease exports, thereby deteriorating the trade balance. The trade balance is the total value of imported goods minus the total value of exported goods. Depreciation of the dollar has the opposite effect, likely improving the trade balance.

The graph above shows this relationship between the trade balance and the exchange rate. The green line plots the trade-weighted U.S. dollar index, which is “a weighted average of the foreign exchange value of the U.S. dollar against the currencies of a broad group of major U.S. trading partners.” A higher value of the index indicates a stronger dollar. The blue line is the trade balance-to-trade volume ratio. The trade volume is the sum of the total value of imports and exports. We look at the ratio instead of the trade balance directly because globalization has led to higher volumes of international trade over time. The ratio gives the difference between exports and imports as a share of total trade, thereby controlling for higher volumes.

Over the past three decades, the trade-weighted dollar index has varied significantly. For example, from the second quarter of 1995 to the first quarter of 2002, the index increased from 90 to 127, an appreciation of the dollar of over 40 percent. The corresponding trade balance-to-trade ratio drops from around –6 percent to –16 percent. In general, we see a negative relationship between the exchange rate and the trade balance.

However, the influence of the exchange rate on the trade balance varies over time. The recent appreciation of the dollar of 20 percent from 2014 to 2016 worsened the trade balance ratio only slightly. The trade balance’s tepid response is likely because of other changes to trade conditions, such as tariffs and regulations. The persistence of the U.S. trade deficit is also noteworthy. Throughout the 22-year span covered in our sample period, the U.S. continuously ran a trade deficit despite the large variation present in the exchange rate. In other words, adjustments to the exchange rate have not removed the U.S. trade deficit even in the long run.

How this graph was created: Search for and add “Trade Weighted U.S. Dollar Index: Broad (TWEXB)” to the graph on the left axis. From the “Edit Graph” tab, add “Exports of Goods and Services (EXPGS) and Imports of Goods and Services (IMPGS) as Line 2. To do this, enter the formula (a-b)/(a+b) in the Line 2 tab. Finally, change the starting date to “1995-01-01.”

Suggested by Yili Chien.

View on FRED, series used in this post: EXPGS, IMPGS, TWEXB

A precise measure of uncertainty?

The Economic Policy Uncertainty Index tries to quantify unpredictability

FREDcast, FRED’s forecasting game, asks players to forecast four major macroeconomic variables by the 20th of every month. Some players may be frustrated by the erratic behavior of some of these indicators as they attempt to make their guesses. Fair enough. Plenty of other data analysts are in the same boat. Well, a team of researchers has been trying to quantify this sense of uncertainty using the Economic Policy Uncertainty Index, graphed above for the world’s five largest economies.

And how, exactly, does one transform a feeling into a number? According to the researchers, articles from major newspapers are analyzed for mentions of uncertainty related to aspects of economic policy, including the decisionmakers themselves, the actions undertaken, and the effects of those policies. The number of articles expressing uncertainty is standardized to the total number of articles written by each source; the accuracy of the resulting index is then tested by making comparisons with relevant indexes constructed through other methods.

The index associates historical events with the economic data. For example, the first significant spike outside of a recession in the above graph shows up in the first quarter of 2003. The spike is highest for Europe and occurred when 10 European countries were in talks to join the European Union under the condition that they’d eventually join the eurozone. The media covered these discussions in a way that uncertainty regarding policymakers, policies, and their effects came across in the index.

In 2008, as the financial crisis was unraveling, there was still much debate on what policies should be adopted and how effective those policies might be. This translated into an elevated index. Soon after, in 2011, the index spiked again in all countries. It’s likely that the uncertainty surrounding the economic and political events at the time, such as the European sovereign debt crisis and the U.S. debt-ceiling discussions, was captured by the index in this case as well.

After 2011, many nations appear to have maintained high levels of economic policy uncertainty that well surpass levels before the Great Recession. The impact of Britain’s vote to leave the European Union, several significant elections around the world, and similar newsworthy events are plainly visible in the all-time highs of the index in the past two years. If you’re not doing too well on FREDcast, you can use the pretty good excuse that there’s certainly a lot of policy uncertainty out there.

How this graph was created: Search for “economic policy uncertainty” and check the boxes next to the monthly series for the United States, Europe, China, India, and Japan. Select “Add to Graph.” Adjust the time range to begin in 1990.

Suggested by Maria Hyrc and Christian Zimmermann.

View on FRED, series used in this post: CHIEPUINDXM, EUEPUINDXM, INDEPUINDXM, JPNEPUINDXM, USEPUINDXM

How much do we spend on new houses?

The highs and lows in the numbers and values of new construction

Do we spend more on new houses than we used to? It can feel like it, especially because houses have become larger and available land has become more scarce. For a quantitative answer to this question, we can use FRED’s data on the number of new houses being sold across the U.S. and the median value of those houses. Multiplying these two indicators yields the total value of all houses sold in a given period. (Well, at least approximately: The mean would give us a better measure, but if the price distribution of new houses doesn’t change too much, this method will do.) Now, prices and incomes have generally increased, so we want to divide the total value of all houses sold by nominal GDP. The result is the series that we display in the graph above, with data normalized to 100 for the start of the sample.

What do we learn? 1) We spend relatively less on houses now, but we’re getting back to the trend. 2) There are strong seasonal factors in the sales volume of new houses. 3) Recessions are really bad for new house sales. 4) The U.S. spent historically high amounts for new houses just before the previous recession and then they dropped to historical lows. 5) Although this recent drop was extraordinarily severe, from a value of 183 to a value of 28 in the matter of a few years, the movements are also very large in other years and some values have doubled within a business cycle. After all, construction is known to be a very volatile sector, and this is especially true for new construction.

How this graph was created: search for “new houses sold, select the series and open the graph. Click on “edit graph” and add a series to the line by searching for median value, then again by searching for “nominal GDP.” Apply formula a*b/c. Finally, change the units ate the very bottom of the form to “Index” setting 100 on 1963-01-01.

Suggested by Christian Zimmermann.

View on FRED, series used in this post: GDP, HSN1FNSA, MSPNHSUS


Back to Top