Exchange Rates and Volatility: Why Currencies Don’t Really Follow the Real Economy

The global financial crisis, trade wars, the pandemic, energy shocks, and geopolitical tensions have turned uncertainty into one of the dominant forces in the global economy. When a country’s growth prospects become more fragile, it would be natural to expect an equally noticeable reaction in its currency. Yet, in international markets, this link continues to elude us.

For decades, economists have debated the so-called “foreign exchange disconnect”: real exchange rates fluctuate widely, but their movements seem to have little to do with differences in consumption and other economic fundamentals across countries. Currencies, in other words, often appear disconnected from the real economy they are supposed to reflect.

A new study published in Management Science shows that the problem runs even deeper. It concerns not only the direction in which exchange rates and consumption move but also their volatility—that is, the way uncertainty increases or decreases over time.

According to the study, the volatility of exchange rates tracks that of economic fundamentals more closely than their levels do, but the connection remains surprisingly weak.

A fresh look at the “foreign exchange disconnect”

The research was conducted by M. Massimiliano Croce of the Department of Finance at Bocconi University, who is affiliated with the IGIER and Baffi research centers. He was joined by Riccardo Colacito of the Kenan-Flagler Business School at the University of North Carolina at Chapel Hill and the NBER; Yang Liu of the HKU Business School; and Ivan Shaliastovich of the University of Wisconsin–Madison.

The paper, titled Volatility (Dis)Connect in International Markets, begins with a question that is simple only at first glance: when uncertainty about one country’s economy increases relative to another’s, does uncertainty about their exchange rate also increase consistently?

To answer this, the authors do not merely observe how consumption and exchange rates change, but also measure how the volatility of both varies over time. This is an important step: two variables may not move in the same direction, but they may still react together during periods of greater instability.

The authors describe the phenomenon with a particularly effective phrase: the ‘disconnect of disconnects’. In other words, there is not just a single disconnect. There are at least two: one between exchange rate movements and consumption, and another between their respective volatilities. And a country that exhibits a strong correlation in the first case does not necessarily show an equally strong correlation in the second. This distinction changes the way we interpret currency markets. It is not enough to ask whether a currency rises or falls in tandem with the economy; one must understand whether its degree of volatility truly reflects the underlying economic uncertainty.

Nearly half a century of international data

The main analysis uses quarterly data from 1971 to 2019 covering 17 advanced economies, including Italy, the United States, Germany, France, the United Kingdom, Japan, Canada, Australia, and Switzerland. Since 1999, the euro area countries have been treated as a single currency unit. The authors also examine a group of emerging markets, including, among others, China, India, Brazil, Mexico, Russia, South Africa, and Turkey. The main sample ends before the pandemic to prevent an exceptional shock from skewing the results; however, extensions of the analysis through 2022 confirm or reinforce the conclusions.

The initial figures show just how different the scales are on which the real economy and currencies operate. In the countries analyzed, the average annual growth of consumption and gross domestic product stands at around 2.2%. Their average volatility is just under 1.9%. The volatility of real exchange rates, on the other hand, reaches 11.2% per year.

Currencies, therefore, fluctuate about six times more than consumption and output. This is one of the reasons why explaining exchange rates solely through macroeconomic fundamentals has proven so difficult.

Volatility responds to the economy, but only very weakly

In the sample, the average correlation between changes in real exchange rates and differences in consumption growth is just 0.04. A value this close to zero indicates that the two phenomena, in terms of their levels, move almost independently of one another.

When comparing volatilities, the correlation rises to 0.20. It is five times higher, but remains far from one—the value that would indicate a perfect connection. Exchange rates therefore seem to reflect the rise in macroeconomic uncertainty, without, however, faithfully replicating it. A significant portion of their volatility stems from other sources.

International differences are striking: the link between exchange rate volatility and consumption volatility is relatively strong in some countries, almost nonexistent in others, and, in some cases, negative. The volatility disconnect therefore does not manifest itself with the same intensity across all markets.

Furthermore, the observed correlation in levels explains only between 10% and 20% of the international differences in the relationship between volatilities. Knowing how closely a currency tracks consumption today does little to predict whether it will react in the same way when uncertainty rises.

Incomplete markets do not explain everything

A traditional explanation for the disconnect between exchange rates and the real economy concerns the incompleteness of international financial markets. Households and investors cannot purchase instruments capable of protecting them from every possible risk originating abroad. International risk-sharing therefore remains imperfect.

This idea may help explain the weak correlation between consumption levels and exchange rates. But, according to the paper, it is not enough to account for what happens to volatility.

In the simplest models, if the relationship between consumption and exchange rates does not change much over time, the volatilities of the two variables should be almost perfectly correlated. The data, however, show an average correlation of just 0.20. The authors’ theoretical conclusion, therefore, is that

“a richer risk-sharing model […] is required to better align the second moments of macroeconomic variables and exchange rates.”

The point is not merely to add a new mathematical complication. It is to recognize that future uncertainty influences economic decisions differently than an immediate change in income or consumption. Models must therefore capture not only current risks but also how households and investors react to news over the long term.

Countries also trade uncertainty

The model developed in the paper distinguishes between two types of shocks. The first concerns news about future growth. If a country’s long-term prospects improve, the way in which resources and risks are distributed across economies also changes. An expected benefit may reduce certain future risks but increase the volatility of consumption in the short term. The authors call this mechanism the “reallocation effect.”

The second type of shock directly affects output volatility. In this case, increased uncertainty tends to spread beyond national borders and simultaneously heighten the instability of multiple economic and financial variables. Investor behavior is summarized as follows:

“When news shocks hit the economy, agents have an incentive to trade to reduce the uncertainty of their future utility.”

International trade, therefore, serves not only to transfer goods or finance investments. It also serves to redistribute exposure to future adverse scenarios. When major news breaks, what flows from one country to another is not just capital—it is also uncertainty.

The two forces identified by the study act in opposite directions. News about long-term growth can reduce volatility across different countries; uncertainty shocks, on the other hand, tend to cause volatility to rise across the board. The result of their interaction is a positive but moderate correlation, just like the one observed in the data.

A new way to interpret currency markets

The message, therefore, is that not all macroeconomic uncertainty is the same. A revision of growth forecasts does not produce the same effects as a sudden increase in volatility. Both shocks can influence consumption and currencies, but through different mechanisms.

When we look at exchange rates, it is therefore not enough to simply measure how risky an economy is. We need to understand which risk has changed, how long it might last, and how market participants can share it with the rest of the world. Currencies, for their part, do not completely ignore the real economy. They pick up on some signals from it, but filter them through long-term expectations, investor preferences, and international risk-sharing mechanisms.