Backus–kehoe–kydland puzzle

What Is the Backus–Kehoe–Kydland Puzzle?

The Backus–Kehoe–Kydland puzzle refers to a significant discrepancy observed in macroeconomic theory, specifically within the context of international business cycles. It highlights that standard dynamic general equilibrium models of open economies predict a higher cross-country correlation of consumption than of output. However, empirical data consistently shows the opposite: cross-country output correlations are typically higher than cross-country consumption correlations. This paradox, named after economists David Backus, Patrick Kehoe, and Finn Kydland, challenges the ability of these models to fully capture the dynamics of global economic fluctuations.

History and Origin

The Backus–Kehoe–Kydland puzzle emerged from research conducted by David Backus, Patrick Kehoe, and Finn Kydland in the early 1990s. Their foundational work sought to determine whether international versions of real business cycle theory could simultaneously explain both domestic and international aspects of economic cycles. In their 1993 National Bureau of Economic Research (NBER) working paper, "International Business Cycles: Theory and Evidence," and subsequent publications, they meticulously documented several discrepancies between their theoretical predictions and observed empirical data.

A central ⁴finding was the puzzling observation that while their models, often driven by technological shocks, predicted that consumption across countries should move more closely together than output due to efficient risk sharing in global markets, real-world data contradicted this. The persistence of this puzzle has spurred extensive research in international macroeconomics, leading to modifications and alternative economic models to better reconcile theory with empirical evidence.

Key Takeaways

• The Backus–Kehoe–Kydland puzzle describes a persistent discrepancy where economic models predict higher cross-country consumption correlations than output correlations, contrary to empirical data.
• It originated from the work of David Backus, Patrick Kehoe, and Finn Kydland in the early 1990s as they tested international real business cycle models.
• The puzzle suggests that existing models may not fully account for international capital flows, financial market frictions, or other factors influencing global consumption patterns.
• Addressing the Backus–Kehoe–Kydland puzzle is crucial for improving the accuracy of open-economy macroeconomic models and understanding global economic synchronization.
• Solutions often involve incorporating incomplete markets, non-tradable goods, or other real-world complexities into theoretical frameworks.

Interpreting the Backus–Kehoe–Kydland Puzzle

Interpreting the Backus–Kehoe–Kydland puzzle involves understanding its implications for how economists model global economies. The puzzle highlights a fundamental tension between the theoretical ideal of international markets facilitating perfect consumption smoothing and the observed reality. In a world with frictionless financial markets, individuals and countries should be able to borrow and lend freely to smooth their consumption paths in response to idiosyncratic shocks. This would imply that consumption fluctuations would be highly correlated across countries, even if national outputs diverge due to localized shocks.

However, the empirical evidence consistently shows that national output tends to be more synchronized across countries than consumption. This suggests that the mechanisms for international risk sharing and consumption smoothing are either less efficient or more complex than assumed in basic international real business cycle models. The persistence of this puzzle indicates that there are significant, unmodeled factors or frictions influencing global economic integration, particularly concerning the behavior of household consumption across borders.

Hypothetical Example

Consider two hypothetical countries, Alpha and Beta, that are highly integrated through international trade and financial markets. According to standard economic theory, if Alpha experiences a temporary negative shock to its output (e.g., a decline in productivity), its residents should be able to borrow from Beta to maintain their consumption levels. Similarly, if Beta experiences a positive shock, its residents would lend to Alpha, or Alpha's residents would borrow from Beta. This ability to borrow and lend globally should lead to highly synchronized consumption paths between Alpha and Beta, even if their production levels fluctuate independently.

For instance, if Alpha's annual output growth drops from 3% to 0% in a given year, while Beta's remains at 3%, Alpha's consumption growth might only slow from 3% to 2% because it borrows from Beta. This would make their consumption levels more correlated than their output levels. However, if empirical data reveals that Alpha's output growth and Beta's output growth move together by, say, 80% (high correlation), while their consumption growth only moves together by 40% (lower correlation), this would represent the Backus–Kehoe–Kydland puzzle in action. The lower observed consumption correlation suggests that despite economic linkages, residents of Alpha are not fully able to smooth their consumption by borrowing from Beta, or vice versa, to the extent predicted by the models.

Practical Applications

The Backus–Kehoe–Kydland puzzle has several practical applications in economics and finance. It is particularly relevant for policymakers and analysts seeking to understand and forecast global economic developments.

Firstly, it informs the development of macroeconomic models used by central banks and international organizations to analyze global linkages. Recognizing the puzzle's existence forces model builders to incorporate features like incomplete asset markets, transaction costs, or non-tradable goods, which can dampen the extent of international consumption smoothing. For example, some researchers have explored the role of real exchange rates and terms of trade in mediating international comovements.

Secondly, for investors, understanding the puzzle provides insights into the true extent of international diversification benefits. If consumption correlations are lower than output correlations, it implies that global financial markets might not be as effective at pooling country-specific risks as theoretical models suggest. This could influence strategies for international investment and asset allocation.

Finally, the puzzle is relevant for policymakers considering measures to enhance global financial stability and economic resilience. If observed consumption patterns differ significantly from theoretical predictions, it may signal underlying financial market imperfections or policy barriers that hinder efficient capital flows and risk sharing. Maximiliano A. Dvorkin of the Federal Reserve Bank of St. Louis, for instance, has explored factors influencing why "Why Consumption Moves Together Across Countries," contributing to the ongoing discussion of this puzzle.

Limitations and Criticisms

Des³pite its prominence, the Backus–Kehoe–Kydland puzzle faces several limitations and has drawn criticisms, primarily concerning the assumptions inherent in the economic models it scrutinizes. A key criticism is that the standard real business cycle models, while elegant, often simplify aspects of real-world financial markets and institutions. They typically assume complete markets, where agents can perfectly insure against all possible future states of the world, leading to a high degree of risk sharing and, consequently, high consumption correlation.

However, real financial markets are incomplete due to various frictions, such as information asymmetries, transaction costs, and limits to arbitrage. These imperfections restrict the ability of countries to fully smooth consumption in the face of idiosyncratic shocks. Therefore, some argue that the puzzle is not a flaw in the empirical data, but rather a limitation of the simplified theoretical frameworks that generate the "paradoxical" prediction.

Another critique relates to the measurement of output and consumption, particularly the detrending methods used to isolate cyclical components. Different detrending techniques can alter the observed correlation patterns. Furthermore, the role of fiscal policy and monetary policy in influencing international comovements, often abstracted away or simplified in early real business cycle models, is also a factor. The puzzle encourages ongoing research into how factors like non-tradable goods, different types of shocks, or varying degrees of financial integration might resolve the discrepancy. Glenn Rudebusch of the Federal Reserve Bank of San Francisco has noted the evolution of real business cycle theory over time, suggesting that models continually adapt to address empirical challenges.

Backus–Kehoe–Kydland Puzzle vs. Con²sumption Smoothing

The Backus–Kehoe–Kydland puzzle is intricately linked to the concept of consumption smoothing, but it describes a failure of theoretical models to match empirical observations regarding this phenomenon in an international context. Consumption smoothing refers to an individual's or country's desire to maintain a relatively stable level of consumption over time, even when income or output fluctuates. This is typically achieved by saving during periods of high income and borrowing during periods of low income.

In international macroeconomics, if countries can freely borrow and lend across borders, they should be able to smooth their national consumption paths more effectively than their national output paths. For example, a country experiencing a temporary negative productivity shock could borrow from abroad to keep its domestic consumption relatively stable, even as its output falls. This perfect international risk sharing implies that consumption across countries should be highly correlated, as global resources are reallocated to ensure similar consumption experiences. However, the Backus–Kehoe–Kydland puzzle highlights that empirical data contradict this, showing lower international consumption correlation than predicted by models, particularly when compared to output correlation. The puzzle suggests that barriers to perfect consumption smoothing, such as incomplete financial markets or non-tradable goods, are more significant in reality than standard models assume. The welfare costs of business cycles, as discussed in broader economic literature, often depend on the extent to which consumption can be smoothed.

FAQs

What does the Backus–Kehoe–Kydland pu¹zzle suggest about global financial markets?

The Backus–Kehoe–Kydland puzzle suggests that global financial markets may not facilitate perfect international risk sharing and consumption smoothing to the extent predicted by simplified theoretical models. It implies that there are significant frictions or unmodeled factors, such as capital controls, information asymmetries, or limits to international borrowing and lending, that prevent countries from fully diversifying their consumption risk.

Is the Backus–Kehoe–Kydland puzzle still relevant today?

Yes, the Backus–Kehoe–Kydland puzzle remains highly relevant in international macroeconomics. While much research has been conducted to address it, fully reconciling the theoretical predictions with empirical data remains an ongoing challenge. The puzzle continues to motivate new research into how global financial integration, international trade, and various economic frictions influence international business cycles.

How does the puzzle relate to international capital flows?

The Backus–Kehoe–Kydland puzzle is closely related to international capital flows. If capital markets were perfectly integrated and complete, countries could use cross-border borrowing and lending to achieve optimal consumption smoothing. This would lead to highly correlated consumption paths, even if national outputs differ. The fact that empirical consumption correlations are lower than predicted implies that capital flows may not be as effective in facilitating risk sharing as theoretical models assume, possibly due to financial market imperfections or other barriers.

What is the main difference between observed data and model predictions in the puzzle?

The main difference is that standard theoretical economic models predict that cross-country consumption should be more highly correlated than cross-country output. However, empirical data consistently show the opposite: cross-country output correlations are typically higher than consumption correlations. This discrepancy is the core of the Backus–Kehoe–Kydland puzzle.