In the wake of the global financial crisis of 2007–2008, advanced economies experienced heightened levels of uncertainty in macroeconomic policymaking. Against this backdrop, policymakers debated the domestic and global spillover implications of advanced-country policy uncertainty (e.g. IMF 2013). At the same time, the potential for monetary policy settings in advanced countries to spill over to emerging market economies (EMEs) via capital flows was hotly contested in both academic and policymaker circles (e.g. Fratzscher et al. 2013).
Policy uncertainty and capital flows
In this column, we present new evidence regarding the spillover impact of policy uncertainty in advanced economies on portfolio capital flows to EMEs (Gauvin et al. 2014). We distinguish between policy uncertainty in the US and the EU, and also between gross portfolio equity and bond inflows.
How might policy uncertainty affect capital flows?
- On the one hand, a less predictable political environment may hinder domestic growth prospects, decreasing the attractiveness of investing in a given country (Baker et al. 2013, Fernandez-Villaverde et al. 2011). Based on this we would, ceteris paribus, expect investors to shift more of their investment abroad given the reduced attractiveness of investing in the US or the EU.
- On the other hand, higher policy uncertainty may decrease the overall size of investors’ positions in relatively more risky investment funds, and impact advanced-economy investors’ willingness to take risk. This may lead to safe haven flows out of EMEs that are often perceived as ‘less safe’.
- Moreover, the effect of policy uncertainty on capital flows may be nonlinear – in periods when the equity risk premium is high, investors are more risk averse. This means that portfolio flows are likely to be more sensitive than otherwise to adverse shocks to news and growth expectations.
A first look at the data
Figure 1 shows the evolution of our measures of US and European policy uncertainty taken from Baker et al. (2013) over our sample period 2004–11. Following low levels of uncertainty in the period of the ‘Great Moderation’, policy uncertainty increased markedly in the period immediately following the crisis. While the levels of policy uncertainty in the US and the EU are highly correlated, changes in the indices – our main explanatory variable – are much more weakly correlated (0.31 over the sample period). To illustrate, policy uncertainty moderated somewhat in both the US and the EU after the crisis, but policy uncertainty in the US moved above European levels with the 2010 Midterm elections and then sharply so as the dispute about the US debt ceiling reached its climax in August 2011. But in the final three months of 2011, EU policy uncertainty moved above US levels, driven by a cut in Italy’s sovereign rating and the call for a referendum in Greece.
Figure 1. Policy uncertainty in the US and EU over time
Figure 2 shows the median of the rolling 12-monthly correlation between changes in US and European policy uncertainty and aggregate fund-level portfolio inflows across a worldwide sample of EMEs1 (blue and red lines respectively) together with a measure of global risk aversion (the VIX index, yellow dashed line). Notably, the correlation between portfolio inflows and policy uncertainty turned negative as risk aversion started to rise in early 2007. This suggests that the relation between policy uncertainty and portfolio inflows into EMEs may change as the level of global risk increases.
Figure 2. Policy uncertainty, aggregate portfolio flows, and global risk
Empirical evidence
We investigate whether changes in policy uncertainty in the US and Europe had significant effects on portfolio capital flows to emerging markets during the period from January 2004 to December 2011. The empirical analysis employs fund-level data on bond and equity inflows to a sample of 20 emerging market economies spread across the globe compiled by EPFR Global. We augment a standard ‘push-pull’ model of capital flows with (exogenous) changes in the policy uncertainty index of Baker et al. (2013), for both the US and Europe, as an additional global factor affecting capital flows. The analysis also employs a nonlinear regression framework that directly addresses the possibility that the marginal effects of policy uncertainty on portfolio flows to emerging markets may depend on both global (such as the VIX, the TED spread, or the level of policy uncertainty itself) and/or domestic (for example CDS spreads, equity market volatility, foreign exchange reserves, capital account policies) factors.
We find – using first a linear regression framework – that increases in policy uncertainty in the US tend to significantly reduce both bond and equity inflows into EMEs. Conversely, increases in EU policy uncertainty tend to have different effects on equity versus bond flows into EMEs – bond inflows into EMEs decrease, but equity flows to EMEs increase in response to increased EU policy uncertainty. This is consistent with the hypothesis that shocks to US policy uncertainty are associated with safe-haven equity flows out of EMEs, whereas the reduced attractiveness of investing in the EU following shocks to EU policy uncertainty appear to outweigh any safe-haven equity flows out of EMEs.
Nonlinearities play, however, an important role in the size and direction of spillover effects. First, we provide evidence for two structural breaks in the relationship between changes in policy uncertainty and capital flows. The first break coincides with the first large increases in the cost of insuring against mortgages of lower credit ratings (BBB- and BBB) in the US, providing evidence that the impact of the coming financial crisis was felt in portfolio flows slightly before the onset of funding illiquidity in the interbank market. The second break occurs in November/December 2010, coinciding with a significant expansion of QE2 by the US Federal Reserve in November 2010.
The level of global risk performs best in explaining nonlinearities. Increases in EU policy uncertainty have a significantly negative impact on bond inflows only in the high global risk regime and, pointing in the same direction, the spillover impact of EU policy uncertainty on equity inflows is less positive in the high global risk regime than the low global risk regime. Global risk (proxied by the VIX index in our baseline) appears therefore not only as an important determinant of capital flows on its own, but it also determines how other push/pull factors (including policy uncertainty) impact portfolio flows.
Turning to domestic factors, we find that the impact of policy uncertainty on bond inflows does not depend on domestic variables – changes to policy uncertainty have for example the same impact on bond inflows independent of a country’s level of sovereign risk or equity market returns. Conversely, the level of country-specific sovereign risk (as proxied by credit default swap spreads) does determine the magnitude of policy uncertainty spillovers via equity flows. Increased EU policy uncertainty pushes portfolio equity inflows into EMEs even if global risk is high, but only into countries with low sovereign default risk.
Domiciliation of funds
Importantly, portfolio flows from funds based in the US may show different reactions to EU policy uncertainty than those from funds based in the EU itself. In this respect, the degree of investor home bias may play a crucial role. And to the extent that policy uncertainty with regard to macroeconomic policies impacts variables such as investors’ wage income risk, it may also affect fund investors’ willingness to buy risky assets, including assets held in EMEs.
We repeat therefore our analysis by distinguishing between flows to EMEs from funds domiciled in the US and funds domiciled in the EU. The main results are robust – we observe positive spillover effects in both the low and high global risk regime even for equity flows originating from funds domiciled in the EU. These spillover effects are stronger for equity flows originating from funds domiciled in the US. In the high global risk regime, flows into EMEs from US-domiciled funds increase even into EMEs with high sovereign default risk, whereas, mirroring our aggregate results, flows from Europe-domiciled funds increase only into EMEs with low sovereign default risk.
Conclusion
This column highlights new evidence that uncertainty in macroeconomic policymaking in advanced countries spills over to emerging market economies via portfolio capital inflows, contributing to the ongoing spillover debate. The results suggest that the origin of policy uncertainty matters. The magnitude of these effects depends on the level of global risk aversion and, in the case of portfolio equity flows, the level of country-specific sovereign risk. Policies such as capital account openness and accumulation of foreign exchange reserves tend not to mitigate the impact of policy uncertainty shocks on portfolio inflows into EMEs. However, improving macroeconomic fundamentals and hence reducing sovereign risk premia would help EMEs with high external financing needs to benefit from more equity inflows in case policy uncertainty were to rise again.
References
Baker, S R, N Bloom, and S J Davis (2013), “Measuring economic policy uncertainty”, Working paper, Stanford University.
Fernández-Villaverde, J, P A Guerrón-Quintana, K Kuester, and J Rubio-Ramírez (2011), “Fiscal volatility shocks and economic activity”, NBER Working Paper 17317.
Fratzscher, M (2012), “Capital Flows, push versus pull factors and the global financial crisis”, Journal of International Economics 88(2): 341–356.
Fratzscher, M, M Lo Duca, and R Straub (2013), “A global monetary tsunami? On the international spillovers of US quantitative easing”, CEPR Discussion Paper 9195, ECB Working Paper 1557, DIW Discussion Paper 1304.
Gauvin, L, C McLoughlin, and D Reinhardt (2014), “Policy Uncertainty Spillovers to Emerging Markets: Evidence from Capital Flows”, Bank of England Working Paper 512.
IMF (2013), Hopes, Realities, Risks, IMF World Economic Outlook, April.
Endnotes
1 These countries are: Argentina, Brazil, Chile, Colombia, the Czech Republic, Hong Kong, Indonesia, India, South Korea, Mexico, Malaysia, Pakistan, Poland, Peru, the Philippines, Russia, Singapore, Thailand, Turkey, and South Africa.
