Globalization of financial risk: How ETFs make the local market less local

ETFs have been an important development in fund management. ETF assets have grown faster than the industry, accounting for over 20% of NYSE market capitalization. This is especially true for funds that invest in emerging markets (figure 1). Over the past 15 years, equity inflows have steadily increased to global risk. These two trends are closely linked, which is the heart of our argument. Are there any causal connections between them?

Figure 1: ETF Market Share, Emerging Markets’ Exposure To Global Financial Shock

In a new paper (Converse et al. A recent report (Converse et al. 2022) shows that ETF investor flows are more responsive than traditional mutual funds to local risk factors. The greater the country’s market capitalization that ETFs hold, the more sensitive portfolio flows are to changes in global financial stress.

We will argue this point more clearly by following two steps. To prove this more clearly, we first look at monthly fund-level data regarding investor flows to equity, bond mutual funds, and ETFs (sourced via EPFR Global) and then run a regression against one global factor (the St Louis Fed Financial Stress Index, which measures global risk conditions), and another local factor (the median growth in industrial production across all countries). We also find that ETFs have a significantly greater negative relationship with global risk than mutual funds investing in emerging market markets.

We also analyze the possible mechanisms that explain this excessive investor flow sensitivity towards ETFs. Our analysis shows that investors with shorter trading horizons tend to hold more ETFs that are more sensitive to global risk factors and to trade less frequently in response to shocks.

There is sufficient evidence to suggest that what happens abroad might be more important than what happens in the country. Here is the second part of our analysis.

From country implications to investor behavior

We calculate a country-level sensitivity parameter (the coefficient from a regression of (a) portfolio equity inflows and (b) stock market returns on our global risks factor) for each country. Then we compare that with the country’s percentage of local equities owned by ETFs (average over the period). Figure 2 illustrates the results: Global shocks strongly influence capital flows (b = more negative), and the greater the ETF share, the better.

Figure 2: Country Betas, ETF Share of Market Capitalization

We can see that portfolio equity inflows are 2.5 times more sensitive to global risks when ETFs have a more significant share of the equity market capitalization of a country. A similar increase in stock market prices is associated with an almost 1.4-fold greater exposure to global factors.

Many variables could lead to increased exposure to global financial stress or a higher ETF share at the country level. To strengthen our understanding of the results, we address these possible concerns in three different ways:

• We also include control variables that affect financial integration, such as the local equity share held by traditional mutual funds and the external liabilities at the country level. When compared with the global risk factor, we find that the ETF share is the most significant.
• To identify changes in mutual fund and ETF shares less closely related to country conditions, we use Vanguard’s funds or ETFs change from the MSCI index towards the FTSE index. Unlike mutual fund shares, ETF share changes due to an exogenous event are related to greater exposure to global financial stress.
• We have linked our country-level and fund-level estimates and found that ETFs have a dollar flow sensibility that is similar to mutual funds.

These findings are consistent with the hypothesis that the rise of passive benchmarked tools contributes to cross-market movement and capital flow synchronicity, at the expense of local fundamentals. This is in line with Levy-Yeyati, Williams (2012), Raddatz, et al. (2017).

These results, along with the growing popularity of ETFs all over the globe, pose challenges to emerging market policymakers to the extent that they increase the “dilemma, not trilemma”, concerns for small, open economies that allow cross-border flows. (Rey 2013, These restrictions may be amplified by ETF penetration, which could strengthen the case for macro-prudential or capital controls.

Refers

Converse, N, E LevyYeyati and T Williams (2022), “How ETFs amplify the Global Financial Cycle In Emerging Markets,” forthcoming Review of Financial Studies.

E. Levy-Yeyati and T Williams (2012), “Emerging economies in the 2000s : Real decoupling and financial recoupling,” Journal of International Money and Finance 31(8), 2102-2126.

Raddatz C, S Schmukler and T Williams (2017). “International Asset Allocation & Capital Flows : The Benchmark Effect,” Journal of International Economics 108, 413-430.

H. Rey (2013), “Dilemma not trilemma: Global Financial Cycles and Monetary Policy Independence”, Federal Reserve Bank of Kansas City Economic Policy Symposium.