The Role of US Monetary Policy in Banking Crises Across the World

Between the literature on how local monetary policy affects financial stability and the literature on a US-driven global financial cycle (Introduction, pp. 1-4). On the first side, monetary policy is known to move risk premiums on equity (Bernanke and Kuttner 2005) and corporate bonds (Gilchrist et al. 2015), and accommodative policy can build vulnerabilities through credit booms and leverage (Adrian and Liang 2018), with reversals raising crisis likelihood (Schularick and Taylor 2012; Baron and Xiong 2017; Danielsson et al. 2018) – “However, this literature does not address how external monetary policy decisions (e.g., monetary policy decisions in an hegemon country) could affect domestic financial stability.” On the second side, Rey (2015) argues for a global financial cycle driven by US policy, Gourinchas (2017) shows in an estimated DSGE model that the degree of financial spillovers matters for transmission, and Jorda et al. (2018) show US policy shapes risk appetite across global equity markets. The paper’s stated contribution is to test whether the spillover is uniform: “we study whether U.S. monetary policy is a uniform driver of financial vulnerabilities abroad or its effects are dependent on the country’s integration with the United States and the rest of the world.” The refinement of Rey’s thesis is stated directly: “we rather argue that U.S. monetary policy affects global financial stability only to the extent that foreign countries have direct exposure to the United States. Those countries with indirect exposure do not always face an increase in their financial stability risks” (pp. 4-5).

Because banking crises are rare, so statistical power requires a long panel – at the cost of thin data on almost everything else (Introduction, p. 2). “Since crises are rare events – a typical OECD country suffers a banking crisis once every 37 years on average according to the banking crisis database of Reinhart and Rogoff (2009) – focusing on long-time-series panel data helps to derive statistically meaningful relationships. However, using such historical data comes at the expense of limited data availability. The biggest challenge for this time period is to proxy U.S. monetary policy decisions.” That constraint drives the baseline design: the historical monetary policy proxy is simply the change in the US 3-month Treasury rate, with identified shock series reserved for a shorter, recent sample. It also drives an omission the authors flag: local monetary policy decisions and exchange rates are expected to matter, “however, historical coverage for both series for many of the countries are poor. When we include changes in the short-term interest rates and exchange rate in our baseline specification, the sample size shrinks by three quarters, hence we do not include these local variables in our baseline regressions” (fn. 2, p. 10).

Reinhart and Rogoff’s (2009) systemic banking crisis definition, coded as the start of a crisis (Section 2.1, p. 6). “A crisis is defined as an event with a closure, merger, or public takeover of one or more financial institutions or large scale government assistance of a systemically important financial institution. The unbalanced panel contains a binary indicator of whether a banking crisis starts in a given year and country and includes 239 distinct banking crises.” The United States itself is excluded from the sample – its own unconditional crisis probability appears in the descriptive figure “for comparison purposes only.” Unconditional annual crisis probabilities vary widely and, importantly, differ systematically between groups: among developed countries “Italy has the highest annual crisis probability at 6.38%; New Zealand has the lowest, 0.96%,” while among emerging countries the range runs “from 0% for Mauritius to 7.8% for Brazil” – which is why the paper later splits the sample.

By the change in US short-term rates historically, and by three published shock series for 1990-2010; the four proxies correlate between 0.52 and 0.83 (Section 2.2, pp. 7-8). The baseline uses the change in US short-term interest rates from the Jorda, Schularick and Taylor (2017) macrohistory database, implemented as changes in the 3-month Treasury rate. The three shock series are Romer and Romer (2004), who “narratively identify changes in the federal funds rate targets surrounding FOMC meetings” and regress them on the current rate and Greenbook forecasts of output growth and inflation for the next two quarters so as to “separate the natural policy response of the economy from the exogenous monetary policy surprise”; Gertler and Karadi (2015), using the change in interest-rate futures in a 30-minute window around a policy announcement, averaged monthly; and Rogers, Scotti and Wright (2014), applying the same idea to the fourth eurodollar futures contract over a window from 15 minutes before an FOMC announcement to an hour and 45 minutes after. On agreement: “all four measures of U.S. monetary policy proxies are significantly correlated with each other, with correlation ranging from 0.52 to 0.83. Of these, the Romer and Romer (2004) shocks have the most dispersion, while the Gertler and Karadi (2015) and Rogers et al. (2014) shocks have similar and relatively small standard deviations.”

Direct exposure by US trade intensity and dollar debt; indirect by overall trade openness and capital account openness – and emerging markets score higher on the direct measures, developed countries on the indirect ones (Section 2.3, pp. 8-9). Direct: “the ratio of a country’s trade with the United States to its total trade,” which “one can think of … as the country’s trade intensity with the United States,” plus, post-1990, “a country’s debt liabilities in USD (% of GDP)” measured net of dollar debt assets, built from the IMF’s CPIS and the BIS locational banking statistics following Lane and Shambaugh (2010) and Bénétrix et al. (2015). Indirect: trade openness (exports plus imports as a share of GDP) as a proxy for “the economic globalization of a country with the rest of the world,” plus, post-1970s, the Chinn-Ito (2006) capital account openness index. The contrast between groups is what makes the distinction bite: “Developed countries on average are more globally integrated than emerging countries, irrespective of the way we measure. In contrast, emerging economies, on average, hold higher U.S.-denominated debt and have higher trade intensity with the U.S. compared to developed economies.”

Because trade itself responds to macroeconomic and financial conditions, so gravity estimates built from geography stand in for it (Section 2.3, pp. 9-10; Introduction, p. 3). “Both trade intensity and trade openness can affect the macroeconomic outlook and financial stability of a country, suggesting a possible endogeneity problem,” and as Frankel and Romer (1999) argued, “the effects of trade on income or crises is expected to be endogenous and hence, one can question the causality. Gravity instruments, however, are derived via countries’ geographic characteristics. Such characteristics are expected to be correlated with trade as they have important effects on trade and are plausibly uncorrelated with other determinants of economic and financial stability measures.” The trade-intensity instrument regresses the log of a country’s share of trade with the US on log weighted distance, US population, a common-language dummy, a shared-border dummy, the log product of the two countries’ land areas, a landlocked dummy and a colonial-link dummy; the openness instrument follows Cavallo and Frankel (2008) with the analogous bilateral specification summed across partners. Instruments are the exponentials of the fitted values. The gravity estimates for equations (1) and (2) are not reported “for the sake of brevity,” so the instruments’ first-stage strength cannot be checked from the paper itself.

A panel logit of the crisis indicator on exposure, exposure interacted with US policy (contemporaneous and lagged), controls, and country and time fixed effects – with the interaction structure forced by the fact that US policy is common across countries (Section 3, pp. 11-12). The reason is stated explicitly: “Using an interaction term is also necessary from an econometric standpoint. U.S. monetary policy shocks do not vary by country, so including is as a stand-alone variable in a panel regression would be akin to adding time-series fixed effects” (fn. 3, p. 11). This is worth carrying as a scope condition: the paper can identify how the sensitivity to US policy varies with exposure, not the average level effect of US policy, which is absorbed by the time effects. Controls are per-capita GDP growth, inflation (annual change in the CPI, following Demirguc-Kunt and Detragiache 1998) and institutional quality proxied by the POLCOMP variable from Polity IV (following Cerra and Saxena 2008), entered contemporaneously and lagged. Standard errors are “dually cluster[ed] … both at the country and year levels to address possible time-series and cross-country correlation of residuals.”

Tightening raises crisis probability where the exposure is direct, by 1.0 to 6.8 percent for a given level of exposure per 1 percent of tightening (Section 4.1, pp. 12-13). “U.S. monetary policy tightening has a positive and statistically significant effect on the probability of a banking crisis for those countries that have direct trade linkages with the United States,” and replacing trade intensity with its gravity instrument leaves the contemporaneous interaction positive while making it “more significant both statistically and economically.” The marginal effect: “a 1 % tightening in monetary policy increases the probability of a crisis by 1.0-6.8% for a given level of direct exposure to the United States.” (The Introduction rounds this to “about 1% to 7%,” and there describes the outcome as a “default probability,” which is a slip – the dependent variable throughout is the onset of a banking crisis, not sovereign or corporate default.) Controls behave as expected: “Higher GDP growth has a negative coefficient, suggesting higher growth reduces the probability of a banking crisis. Higher institutional quality of a country (POLCOMP) lowers the probability of a banking crisis (albeit not significant).”

The effect is ambiguous, and the contemporaneous estimate actually points the other way (Section 4.1, pp. 12-13). “For those countries without direct exposure to the United States, the role of U.S. monetary policy is ambiguous.” With trade openness as the exposure measure, “the coefficient for the contemporaneous interaction term is negative and statistically significant, suggesting that the contemporaneous rate changes for these countries might decrease the probability of a banking crisis” – note the hedge, “might.” Three readings are offered without being separated empirically: “openness helps with diversification”; “these countries might be the immediate beneficiaries of the funds flowing from those other countries which have direct exposure with the United States”; and “even if these countries are not the direct beneficiaries …, a more orderly reversal of capital flows might help correct imbalances that might have accumulated in the run-up period.” Crucially, the negative result is not robust to the instrument: “the impact diminishes when we correct for the endogeneity.” So the safe reading is the paper’s own – ambiguity for indirect exposure, not protection.

Capital flow reversals, and the sudden-stop story wins over the leaning-against-the-wind story (Introduction, pp. 3-4; Section 4.2, pp. 13-15). The two candidates are set out symmetrically. Loosening can produce a credit boom abroad as capital leaves the US chasing yield; during booms “loan quality decreases” (Greenwood and Hanson 2013) and excessive lending worsens crisis risk (Schularick and Taylor 2012; Baron and Xiong 2017) – so on this view “a tight U.S. monetary policy could help rein in excesses and reduce the probability of a crisis (e.g., leaning-against-the wind channel).” Against that, tightening “might lead to a sudden reversal of capital flows” (Neumeyer and Perri 2005; Uribe and Yue 2006). The verdict: “We find evidence that the latter effect dominates. In particular, an increase in U.S. monetary policy rates significantly reduces capital flows to foreign economies for those countries with direct exposure to the United States. When the adjustment becomes disorderly, the crisis probability indeed increases.” The test regresses the change in total portfolio investment flows as a share of GDP on the same exposure interactions plus lagged flows and the full control set, now additionally including the change in domestic interest rates, over a sample beginning in the 1970s. “The interaction term is negative and significant for the direct exposure measures but insignificant for indirect exposure measures.” Note that the mechanism is established in two separate regressions – flows respond to tightening, and crises respond to tightening – rather than by a single mediation test.

Because a US tightening raises its cost twice over (Section 4.2, p. 14). “We consider countries with dollar-denominated liabilities as having direct exposures to U.S. monetary policy, since changes in U.S. monetary policy directly affect the debt servicing costs.” Citing the liability-dollarization literature (Calvo 2002; Choi and Cook 2004; Mendoza 2002), the paper notes that “when dollar denominated liabilities are financed by income derived in local currency, any changes in exchange rate fluctuations could make the debt servicing cost higher,” and specifies the two routes: “First, the rate at which the borrowers roll over their debt would be higher. Second, a higher U.S. monetary policy rate would drive up the value of the dollar relative to other currencies.” Conversely, “we consider countries with more open capital accounts as being globally integrated but not necessarily having direct exposure with the United States. Therefore, we consider capital account openness as an indirect exposure measure.”

Yes on the direct-exposure side, and the indirect-exposure results become statistically stronger without becoming less ambiguous (Section 4.3, pp. 15-16). Restricting to 1990-2010, where the shock series exist, “U.S. monetary policy shocks increase the probability of a banking crisis for countries with direct trade links with the United States … or countries that hold more dollar-denominated liabilities.” For the globally integrated, “results are again ambiguous but statistically stronger compared to using monetary policy stances. U.S. monetary policy shocks decrease the probability of a banking crisis for countries that have strong trade links globally, even controlled for the endogeneity.” But the other indirect measure does not cooperate: “we failed to document a robust relationship between the policy surprises and banking crises probability using the Chinn-Ito index as the exposure variable.” The authors’ summary is appropriately hedged: the effect on crisis probability is positive “only for those countries with a direct exposure to the U.S. For other countries with indirect exposure, the effect of monetary policy shocks is ambiguous. The results in these cases point to a reduction in probability of a banking crisis with some exposure measures and point to an ambiguous effect with some other exposure measures.”

Yes – the developed-country subsample is the one place the main interaction loses significance (Section 4.4, p. 16). The subsamples tested are the post-World War II period; a sample excluding the Great Depression, the Great Recession and both world wars; emerging markets only; developed countries only; and a specification controlling for countries that anchor to the dollar. “In all these sub-samples, the interaction variable for U.S. monetary policy and exposure variable remains positive and statistically significant, with the exception of the sample for developed countries. This finding suggests that the effect of U.S. monetary policy on the probability of a banking crisis (due to increased risk of a sudden stop) is mainly an emerging market phenomena.” This also echoes the earlier channel result, where the authors report that “the increase in the probability of a crisis due to disorderly adjustments in capital flows or sudden stops is an emerging market phenomenon” (Introduction, p. 4).

It does not remove the effect but does amplify it (Section 4.4, p. 16). Adding contemporaneous and lagged interactions of US policy with a dummy for countries anchoring their exchange rate to the dollar, “we find that U.S. monetary policy has a positive effect on the probability of a banking crisis for those countries with direct exposure to the United States, regardless of their anchor policy. However, the impact is economically more meaningful for the countries that anchor their exchange rate to the dollar.” This is a useful boundary on the trilemma reading: the direct-exposure channel operates whether or not the exchange rate is fixed, but fixing makes it bite harder.

OLS, probit, local policy and exchange rate controls, an alternative merged crisis database, and no winsorization – all qualitatively similar (Section 4.4, pp. 16-17; fn. 1, p. 1). The paper is explicit about why the crisis database needs testing: “Our motivation in doing this to see if our results are sensitive to the critiques raised in the literature (see, e.g., Romer and Romer, 2015) regarding Reinhart and Rogoff (2009).” Following Danielsson et al. (2018) it merges the banking-crisis databases of Bordo et al. (2001), Laeven and Valencia (2012), Gourinchas and Obstfeld (2012) and Schularick and Taylor (2012) with Reinhart and Rogoff’s using consistent definitions, and re-estimates. Local monetary policy changes – proxied by changes in short-term local interest rates – and exchange rate changes are added back in one column. The verdict is stated with a mild caveat rather than as a clean pass: “Overall, we find that the results are qualitatively similar under the various robustness checks. There are small changes in different specifications, but the main conclusions hold.” Note that robustness is reported only for the paper’s main direct-exposure column, the gravity instrument for trade intensity, “and leave robustness for all the other columns to an online appendix.”

It reframes the debate by splitting integration into two kinds (Introduction, pp. 5-6). The paper lays out both sides. Integration may raise crisis risk through propagation, since the country is more exposed to foreign shocks (Stiglitz 2010), and Azzimonti et al. (2014) show government debt rising with economic integration so that “integration increases the vulnerability to a crisis.” Or integration may reduce vulnerability (Ayhan et al. 2006; Cavallo and Frankel 2008), because “countries that rely more on trade would be less prone to default, as they are heavily incentivized to maintain trade. Hence, international investors would be less likely to pull out of countries with high trade integration,” and because “trade integration helps countries better absorb shocks.” The paper’s contribution is not to pick a side but to change the question: “We contribute to this literature by distinguishing the integration with the center country and globally.” The concluding statement keeps the two apart: “positive U.S. monetary policy shocks leads to an increase in the probability of a banking crisis for those countries with direct linkages to the U.S., either in the form of trade links or significant share of USD-denominated liabilities. However, if a country is integrated globally, rather than having a direct exposure to the U.S., the effects of U.S. monetary policy shocks are ambiguous” (Section 5, p. 17).