The Dynamics of Capital Flow Episodes

• Published date: 01 August 2020
• Author: CHRISTIAN FRIEDRICH,PIERRE GUÉRIN
• Date: 01 August 2020
• DOI: http://doi.org/10.1111/jmcb.12614

DOI: 10.1111/jmcb.12614

CHRISTIAN FRIEDRICH

PIERRE GU ´

ERIN

The Dynamics of Capital Flow Episodes

We first propose a novel methodology for identifying episodes of strong

equity and bond flows using estimates from a regime-switching model that

keeps context- and sample-specific assumptions to a minimum. We then

assess the impacts of U.S. stock market volatility (VIX) and U.S. monetary

policy shocks on equity and bond flow episodes. Our results indicate that

the impacts of both shocks differ across in- and outflow episodes and, based

on an assessment of equity flows, vary considerably over time. While VIX

shocks are mostly associated with asymmetric impacts acrossepisodes, U.S.

monetary policy shocks generate such asymmetries primarily over time.

JEL codes: F21, F32, G11

Keywords: capital flow episodes, Markov-switchingmodels, global

financial cycle.

FOLLOWING THE TRIAD OF EVENTS comprising the global finan-

cial crisis, unconventional monetary policies, and negative interest rates in many

advanced economies, the assessment of global capital flow dynamics has forcefully

re-entered the research agendas of policymakers and academics. In particular, these

recent experiences have renewed the interest in investigating and understanding the

determinants and consequences of international capital flows.

We contribute to this research agenda in two ways. First, building on the seminal

work of Forbes and Warnock (2012) and Ghosh et al. (2014), who classify episodes

The views expressed in this paper are those of the authors. No responsibility for them should be

attributed to the Bank of Canada, the OECD, or its member countries. We would like to thank Pok-sang

Lam and two anonymous referees, whose comments greatly improved the paper. We would also like to

thank Mahvash Qureshi, Yuriy Gorodnichenko, Michael Hutchison, Nelson Mark, Jean-Marie Dufour,

Oleksiy Kryvtsov, Gee Hee Hong, Garima Vasishtha, Gurnain Pasricha, Louphou Coulibaly, participants

of the 2017 Annual Meeting of the Central Bank Research Association in Ottawa, the 2016 Applied

Time Series Econometrics Workshop held at the Federal Reserve Bank of St. Louis, the 14th INFINITI

Conference on International Finance in Dublin, the 50th Annual Conference of the Canadian Economic

Association in Ottawa and seminar participants at the Bank of Canada for helpful comments. An earlier

version of this paper is available as Bank of Canada Staff Working Paper 2016-9.

CHRISTIAN FRIEDRICH is Principal Researcherat the International Economic Analysis Department,Bank

of Canada (E-mail: cfriedrich@bankofcanada.ca). PIERRE GU´

ERIN is an Economist at the Economics De-

partment, Organisation for Economic Co-operation and Development (E-mail: pierre.guerin@oecd.org).

Received July 13, 2016; and accepted in revised form December 27, 2018.

Journal of Money, Credit and Banking, Vol. 52, No. 5 (August 2020)

C

2019 The Ohio State University

970 :MONEY,CREDIT AND BANKING

of strong capital flows in quarterly and annual data, respectively, we employ a novel

methodology for identifying episodes of strong capital flows in high-frequency data

that is based on estimates from a regime-switching model.1A key advantage of

regime-switching models is that they allow us to determine the underlying regimes

endogenously, without the need for context- and sample-specific assumptions. We

apply our methodology to equity and bond flows into up to 80 different countries2

at the weekly frequency over the period 2000–14. We then convert the estimates

of the regime-switching models into capital flow episodes—that is, binary indicator

variables that mark the prolonged presence of a distinct pattern of strong capital

flows—also simply referred to as “episodes.”3

Second, we use two sets of structural vector-autoregressive(VAR) models to study,

at a monthly frequency, the dynamic interactions between capital flow episodes in

advanced and emerging-market economies on the one hand and global drivers of

capital flows, such as U.S. stock market volatility (VIX)4and U.S. monetary policy

shocks, on the other. Specifically, we document the impact of these shocks on the

share of sample countries in equity and bond flow episodes, and highlight potential

asymmetries in their responses. Using a set of linear VAR models for equity and

bond flow episodes, we first assess whether the responses of the share measure to the

same shock differ across in- and outflow episodes (i.e., an asymmetric impact across

episodes). Based on a set of time-varying parameter (TVP) VAR models for equity

flow episodes, we then analyze whether the responses of the share measure to the

same shock vary over time (i.e., an asymmetric impact overtime). Our results indicate

that the impacts of VIX and U.S. monetary policy shocks differ across in- and outflow

episodes, and vary considerably over time. In particular, we find that VIX shocks are

associated with a response of equity outflow episodes that is three times larger than

that of equity inflow episodes in emerging markets, and with a simultaneous increase

in bond in- and outflow episodes in all sample countries. Further, while the impacts

of both VIX and U.S. monetary policy shocks vary over time, the latter exhibits

an even stronger time-variation as its impact changes sign over our sample period.

While a tightening of U.S. monetary policy led to fewer equity outflow episodes and

more equity inflow episodes in the precrisis period, the pattern is reversed during the

postcrisis period. Overall, our findings suggest that VIX shocks are more often linked

with asymmetric impacts across capital flow episodes, and that U.S. monetary policy

shocks generate such asymmetries primarily over time.

Our first contribution relates to a growing literature that examines the

macroeconomic and financial stability implications of international capital flows. To

1. Earlier studies that have empirically identified episodes of strong capital flowsare Calvo, Izquierdo,

and Mej´

ıa (2004) for “sudden stop” episodes as well as Reinhart and Reinhart (2009) and Cardarelli,

Elekdag, and Kose (2010) for “surges.”

2. There are 65 countries in our equity sample and 66 countries in our bond sample. The notion of 80

different countries emerges since several countries appear in only one of the two samples.

3. We disaggregate capital flow episodes into “equity flow episodes” and “bond flow episodes.”

Further, episodes with strong capital inflows are referred to as “inflowepisodes” and episodes with strong

capital outflows are referred to as “outflow episodes.”

4. VIX refers to the CBOE index of implied volatility on S&P500 options.

CHRISTIAN FRIEDRICH AND PIERRE GU´

ERIN :971

separate extended periods of strong capital flows from regular fluctuations, this lit-

erature makes increasing use of episode classifications.5The decision to rely on

episodes instead of the underlying capital flow data is supported by two additional

reasons. First, since capital flows are oftentimes volatile (e.g., see Bluedorn et al.

2013), the aggregation of individual capital flow observations into episodes can pro-

vide a clearer pattern of the direction and the magnitude of flows. Second, since the

implications of capital flows have been shownto differ according to the level of flows

(e.g., see Abiad, Leigh, and Mody 2009), some of the macroeconomic or financial

effects of capital flows can only be observed when the level of capital flows reaches

certain magnitudes.

The corresponding classification of capital flow episodes has mainly been popu-

larized by Forbes and Warnock (2012). The authors divideepisodes of strong capital

flows into “surges” (inflowsof capital from nonresidents), “stops” (outflows of capital

from nonresidents), “retrenchments” (inflows of capital from residents), and “capi-

tal flights” (outflows of capital from residents). Based on a threshold approach that

identifies deviations from a long-term average as periods of strong capital flows, the

authors apply these categorizations to gross capital flows from the Balance of Pay-

ments (BoP) in a sample of 58 emerging and developed economies at the quarterly

frequency between 1980 and 2009. Ghosh et al. (2014) instead focus on surges of net

capital flows. The authors use a related, but differentlydefined, identification method-

ology than that in Forbes and Warnock (2012) and apply their episode definitions to

annual BoP data in a sample of 56 emerging markets between 1980 and 2011.

Complementing quarterly and annual classifications of capital flow episodes with

a classification for high-frequency data is desirable for at least two reasons. First,

from an academic point of view, it is important to better understand the transmission

of shocks across the global financial system, such as the impact of U.S. monetary

policy shocks and U.S. stock market volatility shocks on other countries. Amplified

by high levels of financial integration and the widespread use of the U.S. dollar,these

shocks can be transmitted rapidly into domestic financial systems with potentially

adverse implications for financial stability.Second, from a more practical perspective,

monitoring international capital flow dynamics in a timely manner is of considerable

importance for central banks and various other policy institutions. Since BoP data are

released at low frequencies and with substantial publication lags, the use of weekly

capital flow data provides timely information for monitoring emerging patterns more

thoroughly and gives policymakers additional time to respond.

Our second contribution, the subsequent analysis of aggregated capital flow dy-

namics, relates to a stream of literature that assesses the determinants of international

capital flows. Dating back at least to Calvo, Leiderman, and Reinhart (1993), who

introduced the distinction between international “push” and domestic “pull” factors,

a rich body of literature developed and culminated in a wealth of studies analyzing

5. Examples of studies that have recently worked with episode classifications are Caballero (2016),

Magud, Reinhart, and Vesperoni (2014), Benigno, Converse, and Fornaro (2015), and Eichengreen and

Gupta (2016).