Anusha Chari | Macro Paper Warehouse

Capital Flows in Risky Times: Risk-on/Risk-off and Emerging Market Tail Risk

Research on cross-border capital flows has concentrated on the first moment – what moves average flows and average returns. This paper asks instead how shifts in global risk appetite reshape the whole distribution of emerging-market portfolio flows and asset returns, and in particular the left tail. It measures “risk-on/risk-off” (RORO) two ways: a statistical index built as the first principal component of daily changes across advanced-economy credit spreads, equity returns and implied volatility, funding-liquidity spreads, and the dollar and gold; and, to separate the price of risk from the quantity of risk, the model-based risk aversion series of Bekaert, Engstrom and Xu (2020). Both measures are right-skewed and fat-tailed, spiking in the global financial crisis, the European debt crisis, the taper tantrum and COVID-19. Outcomes are weekly EPFR country flows (scaled by the previous week’s allocation) and daily total returns from the EMBI, a local-currency bond index, and MSCI local-currency and dollar equity indices, for 22 emerging markets, beginning 7 January 2004 and running to April 2020, with push and pull controls and country and time fixed effects. Estimating panel quantile regressions in the manner of Machado and Santos Silva (2019), the paper finds that a risk-off shock lowers flows and returns across the distribution, and that “in nearly every case we consider” the fifth-percentile realisation moves more than the median while the ninety-fifth moves less – so the distribution shifts left and lengthens. The exceptions and the asymmetries are the interesting part. Bond-fund flows shift left with tails pulled apart; equity-fund flows shift left with tails modestly pulled in, a pattern the paper traces specifically to sensitivity to risk aversion rather than to physical risk. Among returns, equity is far more sensitive than fixed income – more than fivefold on the statistical measure – and within each asset class dollar-denominated indices react more than local-currency ones. Decomposing the index, corporate spreads supply much of the leftward shift and funding liquidity much of the bond funds’ tail-lengthening; the risk-aversion component dominated the global financial crisis while the quantity of risk dominated COVID-19. The paper’s conclusion is methodological as much as substantive: a focus on central tendency is “incomplete,” because the tail responses it documents would be masked by conditional means and variances.

Do finite horizons matter? The welfare consequences of capital account liberalization

Cross-sectional regressions have repeatedly failed to find robust effects of capital account liberalization on investment or GDP per capita, and Gourinchas and Jeanne (2006) found that the welfare gain from integration, measured over an infinite consumption stream, is small. This paper argues both findings are what the neo-classical model should predict and that neither settles whether liberalization is worth doing, because the theory predicts a temporary growth effect with a permanent level effect, and because the standard welfare measure spreads a gain concentrated in the first years over an infinite horizon on which the two consumption paths eventually coincide. Working in a calibrated infinite-horizon Ramsey model with Cobb-Douglas production (beta = 0.96, capital share 0.3, depreciation 0.06, output growth 1.012, population growth 1.022, log utility), the authors assume the liberalizing economy faces the world interest rate immediately and so jumps to the integrated steady state at once – an assumption made explicitly “to abstract from speed of convergence issues” – while the autarkic economy climbs there gradually from the same starting point, the population-weighted capital stock of 1995 (1.96, against a steady state of 3.97). They then compute the Hicksian consumption-equivalent gain over horizons from five years to infinity for 81 non-OECD countries. The timing result is the paper’s core empirical claim: “95% of the increase in annual consumption from capital account liberalization accrues in the first 10-15 years after the opening.” Consequently the same welfare difference reads very differently depending on the normalising horizon. In the baseline, where the economy pays interest on the inflow in perpetuity and never repays principal, the average infinite-horizon gain is minus 3.46 percent of annual consumption – financing costs outweigh the gain – while the five-year figure is 19.02 percent, and the numbers decline monotonically with the horizon, crossing zero around 35 to 40 years. The gain scales with initial capital scarcity: 51.47 percent at five years for the most capital-scarce quartile of countries, against minus 0.90 percent for the most capital-abundant quartile. The robustness exercises mostly preserve the horizon result while showing how sensitive the level is to the debt contract: allowing conditional convergence with country-specific earnings-price ratios gives 9.89 percent at five years and minus 8.35 percent at infinity; imposing the Obstfeld-Rogoff transversality condition, so principal and interest are settled only at infinity, gives 29.32 percent at five years and a positive 4.97 percent at infinity; a 50-year amortising contract gives 18.22 percent and minus 4.12 percent; starting from 1960 rather than 1995 capital stocks, when gaps were wider, gives 32.03 percent at five years. Adding a normally distributed technology shock (mean 1, standard deviation 0.03, described by the authors as “not a realistic shock … used for illustrative purposes”) makes the contract form decisive: with a non-contingent debt contract the computed gains turn sharply negative at every horizon, whereas an equity-like contract that suspends payments in bad states “comes close to the case of an economy that liberalizes and does not need to make repayment.” The authors’ conclusion is methodological and carefully bounded: they “do not claim that policies that lead to permanent effects on TFP and growth are not important,” only that finite-horizon evaluation “may be more appropriate and policy-relevant” for policies with temporary growth and permanent level effects.

Spillovers at the extremes: The macroprudential stance and vulnerability to the global financial cycle

The existing evidence says macroprudential regulation does little to portfolio capital flows; this paper argues that finding is an artefact of looking only at averages. It links two literatures – one on the leakages and spillovers from macroprudential policy, one on extreme events in capital flows – and asks whether a country’s ex-ante macroprudential stance changes how sensitive its bond and equity portfolio flows are to the global financial cycle. The answer is yes, in the tails and not at the mean. Tighter prior regulation amplifies risk shocks in both directions: bigger inflows during risk-on episodes, bigger outflows during risk-off ones. Four innovations make the result visible. Rather than dummy variables for recent policy changes, the paper builds four measures of the regulatory stance that incorporate intensity, combining Bank for International Settlements and European Systemic Risk Board data on countercyclical capital buffer levels with the IMF’s iMaPP database, including its quantitative loan-to-value ratios. Risk is measured by the RORO index of Chari, Dilts-Stedman and Lundblad (2020), the first principal component of daily changes in advanced-economy credit spreads, equity returns and volatilities, funding-liquidity spreads, the dollar and gold, whose distribution is skewed toward risk-off. Flows come from weekly EPFR data covering over 14,000 equity funds and 7,000 bond funds with more than $8 trillion under management, cleansed of valuation effects. And reverse causality – policymakers tightening because flows surged – is handled with a policy-shocks approach that regresses the stance on eighteen crisis, credit, growth and institutional variables and uses the residual, with first-stage F-statistics around 100. On a sample of 65 countries excluding the United States, Japan and Switzerland, the second stage reproduces the two known facts: a one-unit rise in RORO cuts weekly bond flows by 0.09 to 0.10 percent, about $2.3 to $2.4 billion, while the macroprudential stance on its own is insignificant. The interaction is where the new result sits: negative and usually significant, but modest at the mean, worth only $151 to $543 million of extra bond outflow. Evaluated across the risk distribution it grows sharply – for the preferred Broad Intensity Index, a one-unit tighter stance adds nothing at median risk but -$636 million, -$1,529 million and -$2,076 million at the 95th, 99th and 99.5th percentiles, against an unconditional risk effect of about -$2 billion, and +$631 million to +$1,215 million at the 5th to 0.5th percentiles. At a 99th-percentile shock (a RORO of 3.49, reached in 2008-09, 2011 and 2020), the amplification of bond outflows is 22 to 87 percent across all four measures and 45 to 67 percent for the two preferred ones. The pattern holds for equities with smaller coefficients but larger dollar amounts, and it is driven by tools that target specific exposures – LTV ratios, FX measures, bank credit supply – while the countercyclical capital buffer and demand-side measures show the same sign but no significance. The authors are explicit about the inference they are not making: “we do not suggest that macroprudential policies render the broader economy less resilient or more sensitive to risk shocks – as the increased resilience of banks may outweigh the greater sensitivity of non-bank financial intermediation.”

Taper Tantrums: Quantitative Easing, Its Aftermath, and Emerging Market Capital Flows

Identifying US monetary policy shocks from daily moves in five-year Treasury futures around FOMC announcements, this paper shows that during the unconventional-policy years those shocks largely represent revisions to required risk compensation rather than to the expected short-rate path, and that their effects on emerging market portfolio positions run mainly through valuations rather than physical flows – with by far the largest effects during the taper period. The shock measure follows Rogers, Scotti and Wright (2014): the daily change in the implied yield of the five-year Treasury futures contract on FOMC announcement dates, plus the additional policy events in Gagnon et al. (2011) and the taper-tantrum date of 22 May 2013. Its average value is a fall of 2.0 basis points during the QE period and a rise of 1.6 basis points during the taper period, against minus 0.6 for the full sample and minus 0.5 pre-crisis, with the period differences statistically significant. Feeding the shock through the Kim and Wright (2005) affine term structure decomposition shows it moves both the expected short rate and the term premium in the conventional period, but that in the unconventional periods the largest effects are on term premia and those effects rise monotonically with maturity – a one-standard-deviation shock raises the ten-year yield by 4.7 basis points pre-crisis but 12.2 basis points during QE, against unconditional daily ten-year standard deviations of 5.8 and 7.1 basis points respectively. The capital-flow analysis uses Bertaut-Tryon and Bertaut-Judson monthly estimates built from US Treasury International Capital data, covering 15 emerging markets monthly from 1994 to 2014, with positions, flows and valuation changes for debt and equity separately scaled by annual GDP, estimated in a random-effects panel with lagged dependent variables, an extensive set of lagged push and pull controls, and country-clustered robust standard errors. Three kinds of heterogeneity emerge. Flows versus prices: “in nearly every specification, the effect of monetary policy shocks on asset returns is larger than that for physical flows,” which the authors read as consistent with the shocks capturing revisions in required risk compensation. Debt versus equity: during QE the coefficient on equity valuations is ten times that on debt valuations, and during the taper period equity effects are double or triple debt effects. QE versus tapering: during QE the significant responses are confined to debt and equity valuations and equity positions, whereas after tapering was first mentioned the coefficients are inversely signed and significant at the 1 percent level across essentially every variable, and an order of magnitude larger than pre-crisis for debt positions, debt valuations and equity flows. Because the shock has a magnitude, the paper can price these effects: a mean-sized QE shock corresponds to roughly a $153.5 million monthly increase in US emerging market equity positions per country, a mean-sized taper shock to a $144.1 million monthly outflow, with one-standard-deviation ranges of roughly minus $672 million to plus $979 million during QE. The paper is explicit that its estimates are associations from a controlled panel regression rather than structural effects: coefficients are described throughout as correlations, the shock’s channel is inferred from coefficient signs rather than separately identified, and the exchange rate results, while statistically significant, are reported as economically modest against an unconditional monthly bilateral exchange rate standard deviation of 3.57 percent.

The real effects of capital controls: Firm-level evidence from a policy experiment

After the 2008-09 crisis, very low interest rates in advanced economies pushed capital into emerging markets, and several governments taxed the inflows; in December 2012 the IMF endorsed limited use of capital controls. This paper asks what those controls cost the firms in the country imposing them, using Brazil – “seen as a poster child for the recent policy changes” – and a feature of Brazilian tax law that turns the policy into something close to an experiment. The Imposto Sobre Operacoes Financeiras (IOF) is set by decree rather than statute, so it needs no Congressional approval and “the Finance Ministry can overnight change the IOF tax that becomes effective immediately from its enactment date”; investor interviews in Forbes, Fratzscher, Kostka and Straub (2016) confirm investors did not anticipate the changes. The authors collect the announcement dates for Brazil’s IOF changes between 2008 and 2013, along with which instruments each covered, and run an event study on listed Brazilian firms, matching Datastream prices and Worldscope financials to proprietary export data from Brazil’s trade secretariat (Secex). The theoretical prediction comes from Black (1974) and Stulz (1981): a discriminatory tax on foreign investors segments markets and “drives up the expected return relative to the benchmark return under full integration,” so prices should fall and cumulative abnormal returns should be negative. They are. Two-day CARs computed against a market model with Scholes-Williams betas fall about 0.28 percent on average, significant at 1 percent; controlling for firm size the average effect rises an order of magnitude to -2.66 percent, while size itself enters positively – so the average masks large heterogeneity. Fitted CARs rise monotonically with size, stay negative through the 75th percentile, and turn positive at the 90th and above; exporter status is positive and significant at 5 percent, concentrated in the larger export-revenue bins; and external finance dependence, measured as the Rajan-Zingales gap between capital expenditure and internal cash flow, is negative and significant at 1 percent. Controls on equity inflows hit harder than controls on debt, with the equity-event dummy negative and significant at 5 percent, which the authors attribute either to surprise – Brazil had previously taxed only debt flows, extending the IOF to equity for the first time in October 2009 – or to the market viewing debt controls as a legitimate macroprudential response to systemic risk. On mechanism, five-year market interest rates rise 11.8 basis points around the announcements (significant at 5 percent), “against the backdrop of quantitative easing in the US and other developed countries that put downward pressure on the world interest rate,” and an implied cost of capital computed from IBES forecasts via the Easton (2004) modified PEG ratio rises significantly at the 10 percent level in short windows. The exchange rate moves in the direction the policy intended – depreciation – but insignificantly, so “the lack of statistical significance precludes us from drawing robust inference.” Two scope conditions matter most. Only listed firms are observable, so the smallest firms are missing entirely and the authors treat their estimates as “a lower bound estimate of the adverse impact.” And the design measures announcement-window asset prices, not realized investment: the source text reports no regression of firm capital expenditure on the controls, despite the paper’s title and abstract framing.

The Return to Capital in Capital-Scarce Countries

The recent resolution of the Lucas paradox has been that the marginal product of capital is not actually higher in poor countries once measurement is done properly, so there was never much incentive for capital to flow there. This paper reopens the question by measuring both quantities that the neoclassical first-order condition links – the marginal product of capital and the financial return – on the same firms, using Worldscope accounting and stock-market data for listed firms in MSCI developed and emerging countries from 1997 to 2014 (334,471 firm-years across 42 countries). The marginal product is proxied by earnings before interest, tax, depreciation and amortisation over the previous year’s market value of assets (debt at book plus equity at market); the financial return is that plus the capital gain net of new investment, following Fama and French’s internal-rate-of-return-on-value construction; both are inflation-adjusted. The results split the two apart. Consistent with the neoclassical prediction, firm-level return on assets is significantly negatively related to GDP per capita, and this holds after firm, industry and time controls, in 40 of 44 non-financial Fama-French industries, in every single year of the sample, in the post-crisis window, among IFRS adopters, in the EU subsample, using output per worker or per hour instead of per capita, and after adjusting income for corporate tax. The internal rate of return shows nothing of the kind: the coefficient on GDP per capita is statistically insignificant in the main specification and in every robustness variant, insignificant in 42 of 44 industries, and insignificant or positive in 10 of 18 years. Averaged across the sample, return on assets is 9.2 percent and the internal rate of return 8.3 percent, with emerging markets showing higher return on assets but lower internal rates of return than developed markets, in means and medians alike. Quantile regressions sharpen the point: the negative relation with income is strongest for the most profitable firms, yet even those firms show no corresponding advantage in realised returns – “even the best-performing firms within emerging countries cannot successfully translate their higher marginal products of capital to higher investment returns.” The proposed mechanism is a capital accumulation friction: adding a quadratic adjustment term to the accumulation equation breaks the constant-depreciation link, and a firm-level test finds the squared investment-to-capital ratio significantly related to the growth of capital at market prices, so the linear accumulation process implicit in perpetual-inventory capital stocks needs modification. The implication the paper draws is a redirection rather than a solution: “a key explanation for the pattern of international capital flows may indeed be domestic rather than international frictions.” Its own stated limits are firm: the sample is listed firms only, so “our conclusions about the Lucas paradox are restricted to the sample of public firms,” and firm data say nothing about the self-employed or informal sector that “make up a large part of the economy in developing countries.”