The Global Crisis has exposed the vulnerability of open economies to the vagaries of the global economy. It may even have induced profound changes in how international financial markets work. Lilley et al. (2020), for instance, show that since around 2007 there is a sustained increase in the explanatory power of global risk factors for the broad dollar exchange rate. The crisis has also triggered a lasting debate about how specific impulses are transmitted internationally (e.g. Rey 2013). Recently, Kalemli-Ozcan (2020) documents that emerging market economies are particularly vulnerable to international risk spillovers. This finding, in turn, is in line with the received wisdom that interest rate shocks are an important source of the business cycle in emerging economies (Neumeyer and Perri 2005, Uribe and Yue 2006).
Increased volatility of advanced economies’ country spreads after 2008
In recent work (Born et al. 2020), we provide new evidence that questions this received wisdom. We document what is – from the perspective of advanced economies – a rather unpleasant convergence in the behaviour of interest rate spreads across emerging and advanced economies.1 We measure spreads as the difference between foreign currency-denominated government or government-guaranteed bonds and risk-free bonds in the same currency. Changes in the spread therefore reflect changes in default risk and/or risk aversion (rather than expectations about inflation and/or expected currency depreciation). Our data set comprises of 3,200 quarterly observations for the period of the early 1990s to the end of 2018. It includes data for 21 advanced and 17 emerging economies. In our analysis, we distinguish the period before and after 2008 throughout.
Figure 1 Average country spread and country-specific standard deviations for advanced and emerging markets
Notes: Top panel: Average country spread for 21 advanced economies (blue dashed line), and 17 emerging economies (red dotted line). Shaded area denotes period since 2008Q1. Bottom panel: country-specific standard deviations for advanced (blue plus signs) and emerging (red circles) economies before 2008 (x-axis) and after 2008 (y-axis). Black line indicates 45-degree line.
The top panel of Figure 1 shows the average country spread for 21 advanced economies (blue dashed line) and 17 emerging economies (red dotted line). For the period before 2008 we observe that the average spread is very low and stable in advanced economies, but high and volatile in emerging economies. In contrast, the average spread behaves much more similarly across the two country groups in the period after 2008. The average standard deviation of the spread in emerging economies declines from 3.94% to 2.29% across the two sample periods. Instead, it increased by a factor of seven in advanced economies, from 0.32% to 2.22% – an unpleasant convergence indeed. Importantly, this convergence is broad-based and not just driven by individual (euro area) countries. The bottom panel of Figure 1 displays the standard deviation of the spread before 2008 (horizonal axis) against the spread after 2008 (vertical axis) on a country-by-country basis. Blue crosses (red circles) indicate observations for advanced (emerging) economies. We observe that the volatility of the spread has increased in all advanced economies but one and it has declined in most emerging economies.
Spread shocks before and after 2008
Against this background, we ask whether spread shocks impact emerging and advanced economies differently. The notion of spread shocks is economically meaningful: while the spread is certainly endogenous to the fundamentals of a country, it also varies for reasons that are exogenous to country-specific developments. One possibility is that global factors such as, for instance, changes in risk aversion or the global financial cycle cause the spread to vary. Another possibility is that spreads shift due to market sentiment or coordination failure, as a result of which changes in expectations may become self-fulfilling (Bocola and Dovis 2019, Lorenzoni and Werning 2019). We employ alternative strategies to identify spread shocks and robustly find that their transmission mechanism is fairly similar both across the two country groups and sample periods.
Figure 2 Impulse responses of output to a sovereign spread shock in period 0
Note: Solid blue and dashed red line represents deviation from pre-shock level for advanced and emerging economies, respectively. Shaded areas correspond to 90% confidence intervals based on clustered robust standard errors. Horizontal axis measures time after treatment in quarters.
Figure 2 shows the dynamic adjustment of output to a spread shock of on average 50 basis points. Output is measured in percentage deviation from trend against the vertical axis and the horizontal axis measures time in quarters. The top panel shows the response before 2008, the bottom panel the response after 2008. For each panel the solid blue line corresponds to the point estimate for advanced economies, while the dashed red line captures the estimate for emerging markets. Shaded areas indicate 90 percent confidence bounds. While the maximum output decline before 2008 is a bit larger than 0.2%, it reduces to about 0.15% for the period after 2008. We also find that the adjustment dynamics are similar across country groups and sample periods for a wide range of variables. However, we also uncover a couple of noteworthy exceptions, such as the response of policy rates, which are raised in response to a spread shock in emerging market economies but not in advanced economies.
Lastly, we compute the contribution of spread shocks to the fluctuations in the spread and in output, again for both sample periods and country groups. We report the average contribution to the forecast error variance over the first three years in Table 1.
Table 1 Average share of forecast error variance explained by spread shocks during the first three years
Note: In case of the country-specific shocks, the variance decomposition only refers to the country-specific variation in the dependent variable.
The first row captures the contribution of all spread shocks, while the second row only captures country-specific spread shocks. Our estimates show that spread shocks contribute significantly to emerging economy business cycles, both before and after 2008. However, because output is less volatile in emerging economies after 2008, the importance of spread shocks in absolute terms has declined somewhat. For advanced economies, the opposite holds true: the percentage contribution of spread shocks to the volatility of output has increased after 2008 – against the backdrop of an increased output volatility. Interestingly, we also find that the relative importance of spread shocks for the volatility of spreads has declined considerably in advanced economies. The increased volatility of spreads is therefore as much a result of shocks as of a stronger endogenous response in the transmission of other shocks.
Increased vulnerability and increased market discipline
In light of these findings, there is no doubt that the vulnerability of advanced economies to international financial markets has increased. On the one hand, this is due to an increased incidence of global and local spread shocks. This, one might say, is outside the control of domestic policymakers and rather a matter of bad luck. On the other hand, however, our results show that financial markets are now less lenient about economic fundamentals in advanced economies than before. This may be because fundamentals look less benign in advanced economies or simply because lending standards have gone up. Either way, policymakers in advanced economies are now exposed to more market discipline – for better or worse. Perhaps their peers in emerging market economies may teach them a lesson or two about how to operate under such conditions.
Bocola, L and A Dovis (2019), “Self-Fulfilling Debt Crises: A Quantitative Analysis”, American Economic Review 109(12): 4343–4377.
Born, B, G J Müller, J Pfeifer and S Wellmann (2020), “Different no more: country spreads in advanced and emerging economies”, CEPR Discussion paper 14392.
Kalemli-Ozcan, S (2020), “US monetary policy, international risk spillovers, and policy options”, VoxEU.org, 16 January.
Lilley, A, M Maggiori, B Neiman and J Schreger (2020), “Exchange rate reconnect”, VoxEU.org, 24 January.
Lorenzoni, G and I Werning (2019), “Slow Moving Debt Crises”, American Economic Review 109 (9): 3229-3263.
Neumeyer, P A and F Perri (2005), “Business cycles in emerging economies: the role of interest rates”, Journal of Monetary Economics 52(2): 345–380.
Rey, H (2013), “Dilemma not Trilemma: The Global Financial Cycle and Monetary Policy Independence”, Jackson Hole Symposium Proceedings.
Uribe, M and V Z Yue (2006), “Country spreads and emerging countries: who drives whom?”, Journal of International Economics 69(1): 6-36.
1 We thank Giancarlo Corsetti for suggesting the term ‘unpleasant convergence’ to us.