Part I. International Links
I. Summary of Foreign Entanglements: Measuring the Size and Source of Spillovers Across Industrial Countries
IMF Working Paper WP/07/182 by Tamim Bayoumi and Andrew Swiston
1. The size and sources of international spillovers of activity remain subject to significant uncertainty. One reason for this is the difficulty of separating the impact of global and regional shocks, given the high correlation of economic cycles.
2. This working paper uses a new approach to differentiating these effects using disturbances to a diverse group of small industrial countries as a proxy for global shocks. The logic is that a disturbance to an aggregate of growth in Australia, Canada, Denmark, Norway, New Zealand, Sweden, and Switzerland—countries that are varied in location and economic structure—is a good candidate for a global shock.
3. Vector autoregressions (VARs) are used to estimate the size and sources of spillovers. The size of spillovers is estimated using VARs of quarterly growth for four regions since the early 1970s—the United States, the euro area, Japan, and the small industrial countries. Next, the channels for spillovers are investigated by adding data on real net exports, commodity prices, and financial variables (short- and long-term interest rates and equity prices) to this baseline model.
4. Particular attention is placed on the robustness of results across alternative approaches to identifying the baseline VAR. The traditional approach to identification is to assume that contemporaneous spillovers are assumed to flow in only one direction (say from the United States to the euro area) and not in the other direction. In this paper, results are also reported for a new approach to identification in which spillovers flow in both directions.
5. The results from the baseline VAR suggest that shocks to the United States are significant for foreign activity. The spillovers to the euro area, Japan, and small industrial countries are roughly one-quarter to one-half as large as the disturbance in U.S. output. They are particularly large and statistically significant for the euro area and small industrial countries (Figure 1). On the other hand, while global shocks have some impact on U.S growth, spillovers from the euro area and Japan are small and insignificant. Similar results are found for the first and second half of the sample, except that the size of U.S. shocks diminishes considerably, reflecting the great moderation in U.S. macroeconomic volatility.
Figure 1.Responses to shocks to U.S. GDP
Source: IMF staff calculations.
6. The size of U.S. spillovers is robust to alternative orderings of the regions.Figure 1 reports the uncertainty around the impulse response functions based on potential volatility of the parameters of the VAR and the additional uncertainty reflecting alternative orderings of the variables in the identification scheme. As can be seen, the difference between the two estimates is small for U.S. spillovers. (This is less true of responses to global shocks, identified as disturbances to the aggregate of other industrial countries
7. To investigate the sources of spillovers, the baseline VAR is extended to encompass the main possible conduits—trade, commodity prices, and financial variables. More specifically, the impact of real net exports, real commodity prices, short-term interest rates, bond yields, and equity prices on spillovers were estimated by adding these variables to the VAR. To conserve degrees of freedom, each of these variables is included separately.
8. U.S. spillovers to foreign output are largely transmitted through financial channels (Figure 2). Trade explains only a limited amount of spillovers across regions and, while slightly larger, the role of commodity prices is also limited. Financial market conditions play the largest role in transmitting spillovers across regions, with the largest contribution coming from monetary policy but significant roles also played by equity prices and bond yields.
Figure 2.Decomposition of Responses to U.S. GDP
Source: IMF staff calculations.
9. The aggregate impact estimated from these separate sources corresponds reasonably closely to the overall impulses, providing a useful check on the results. Since the impact of each source of spillovers is estimated from a separate VAR, these effects can be added together to provide an alternative estimate of the size of the spillovers. As can be seen in Figure 2, there is a relatively close correspondence between the two approaches.
Conclusions and Policy Implications
10. To summarize our findings:
- The United States creates large spillovers to other regions. The impact of global shocks (identified in the approach using those experienced by smaller industrial countries) is also significant, although their size depends on the chosen specification. The euro area and Japan generally have limited spillovers on other parts of the world.
- The main source of spillovers is global financial conditions. Both monetary policy and financial conditions more generally matter for transferring activity across regions. By contrast, trade and commodity prices are less potent factors in this process.
- The great moderation in U.S. activity appears to have driven lower global output variability. Lower U.S. macroeconomic uncertainty, and the associated financial stability, appears to have been the main factor behind lower global output volatility.
11. These results help explain several features of global business cycle linkages. First, large macroeconomic models that capture trade linkages better than financial ones have consistently failed to find large spillovers across major global regions. Second, the impact of U.S. growth on the rest of the world is higher in recessions than mid-cycle slowdowns, consistent with the fact that U.S. financial conditions have typically become tighter in the former than in the latter. Finally, the global cycle is highly synchronized. The rapid propagation of shocks is consistent with a larger role in the transmission process for financial markets, given that they adjust more quickly to new information than do trade flows.