Log in
Show password
Forgot password ?
Become a member for free
Sign up
Sign up
New member
Sign up for FREE
New customer
Discover our services
Dynamic quotes 
News: Latest News
Latest NewsCompaniesMarketsEconomy & ForexCommoditiesInterest RatesBusiness LeadersFinance Pro.CalendarSectors

Banking networks and economic growth: from idiosyncratic shocks to aggregate fluctuations

12/01/2021 | 10:01am EST

Working Paper Series

No 128 / December 2021

Banking networks and economic growth: from idiosyncratic shocks to aggregate fluctuations


Shohini Kundu Nishant Vats


This paper explores the transmission of non-capital shocks through banking networks. We develop a methodology to construct non-capital (idiosyncratic) shocks, using labor productivity shocks to large firms. We document a change in the relationship between foreign idiosyncratic shocks and domestic economic growth between 1978 and 2000. Contemporaneous changes in banking integration drive this phenomenon as geographically diversified banks divert funds away from economies experiencing negative shocks towards other unaffected economies. Our GIV estimates suggest that a 1% increase in bank loan supply is associated with a 0.05-0.26 pp increase in economic growth. Lastly, this can potentially explain the Great Moderation.

Keywords: financial intermediation; growth; deregulation; cross-border spillovers; idiosyncratic shocks; credit; the Great Moderation

JEL classification: E32; E44; F36; G21; G28; O47; R11; R12


  • Introduction

The objective of this paper is to explore the transmission of non-capital (real) shocks through banking linkages. Understanding how shocks that materialize inside and outside of the banking sector transmit across geographies is critical to deepening our understanding of how the typology of shocks is a key determinant of macroeconomic consequences. In both standard international real business cycle models and banking models, greater financial integration can result in the synchronization of business cycles when banking shocks are the prime source of aggregate fluctuations, and desynchronization of business cycles when non-banking shocks are the prime source of aggregate fluctuations.1 While a large body of empirical work has studied the transmission of bank capital shocks though banking networks, it has yet to address how non-capital shocks propagate through banking networks.2 In this paper, we provide empirical evidence that geographically diversified banks divert funds away from economies experiencing negative non-capital shocks, and towards other unaffected economies. This suggests that the transmission of non-capital shocks through banking networks results in negative comovement of business cycles, consistent with theoretical predictions. Thus, we present a mechanism through which banks act as potential aggregators of idiosyncratic shocks to understand the origins of aggregate fluctuations.

Specifically, we develop and empirically implement a test of how non-capital shocks, idiosyncratic shocks, hereafter, are transmitted through banking linkages. We develop a simple statistical model that links foreign idiosyncratic shocks with domestic economic growth through banking networks. Idiosyncratic shocks can affect future returns on capital, but do not affect bank capital contemporaneously. We use this model to derive an empirically testable relation between foreign idiosyncratic shocks, the strength of banking networks, and domestic economic growth. The basic insight in the model comes from the distinction between bank capital shocks and non- capital shocks. While bank capital shocks directly affect the aggregate amount of loanable funds, non-capital shocks affect the relative lending share across geographies, keeping the total stock

1In their theoretical work, Perri and Quadrini (2018) show that with banking integration, endogenous shocks related to the banking sector may result in the synchronization of business cycles, whereas exogenous country-specific shocks originating outside of the banking sector may cause desynchronization of business cycles among interconnected economies. Other related works that highlight the two competing mechanisms include Holmstrom and Tirole (1997), Morgan, Rime, and Strahan (2004), Kalemli-Ozcan,Papaioannou, and Perri (2013), and Kalemli-Ozcan,Papaioannou, and Peydró (2013).

2Several papers have exploited periods of macroeconomic downturns to understand the transmission of bank capital shocks through banking networks. Such works among others include Peek and Rosengren (2000); Khwaja and Mian (2008); Schnabl (2012); Chodorow-Reich (2014); Huber (2018).


of funds fixed. Specifically, if a banking network spans two economies, domestic and foreign, foreign negative idiosyncratic shocks may boost the domestic loan supply and subsequent domestic economic growth. This implies that geographic diversification of banks in the presence of non- capital shocks reduces the covariance of business cycle fluctuations across geographies. However, banking networks make domestic and foreign economies more vulnerable to foreign idiosyncratic shocks, increasing the variance of business cycle fluctuations in both economies. Ultimately, if the reduction in covariance dominates the increase in variance, aggregate volatility declines.

The cleanest natural experiment to test the transmission of shocks through bank networks requires an exogenous shock to the banking network and measurement of non-capital shocks. Dissolution of regulatory barriers to geographic expansion of banks in United States from 1980s through 1990s provides such an environment with plausibly exogenous shocks to the banking network. State-level fluctuations are constructed using labor productivity shocks to large firms headquartered in that state after partialling out industry-wide labor productivity shocks as in Gabaix (2011). We focus on state-level fluctuations, constructed from labor productivity shocks to large firms for two reasons. First, these shocks are geographically isolated, lack temporal dynamics, and are firm-specific events. Second, state-level fluctuations that are constructed from labor productivity shocks to large firms are unlikely to be related to bank-capital shocks, as large firms are less reliant on banks as a source of external financing (Gertler and Gilchrist (1994); Kashyap, Lamont, and Stein (1994)). In addition, idiosyncratic shocks may alter banks' expectations of future economic growth of the state. Hence, the geographic isolation, lack of temporal dynamics, orthogonality to contemporaneous bank capital and ability to predict future economic growth make idiosyncratic shocks prime candidates for measuring non-capital shocks.

For illustration of the mechanism that connects idiosyncratic shocks to economic growth through banking network, consider the following microcosm of our empirical setting. There are only two states in the economy: Illinois and Indiana. Prior to deregulation, Illinois and Indiana are connected through a non-banking channel, namely, an exports/imports channel. Suppose that in this fictionalized world, Illinois' greatest exports are free-market economists and Indiana's greatest exports are conservative politicians. If a new bill is passed allowing banks operating exclusively in Illinois to operate in Indiana and vice versa, the two states will now be connected by a banking channel in addition to the existing non-banking channel. Our focus is on how the transmission


of shocks between Illinois and Indiana changes upon passage of this new bill. For example, if a localized fire destroys all economics textbooks in the largest printing house in the state of Illinois, how will Indiana's economy be affected in the presence of banking linkages? We hypothesize this negative idiosyncratic shock will hurt Illinois' labor productivity as economists may need to reinvent several basic theories for their work and lose easy access to existing research. Banks will note that due to reduced labor productivity, returns to capital in Illinois will be lower as economists will use a portion of their capital to reinvent knowledge. As a result, banks will divert their loan supply to Indiana, thereby increasing investment in Indiana and fostering positive economic growth. This reductive example is intended to illustrate the mechanism that connects foreign (Illinois) shocks to domestic (Indiana) growth in the presence of a banking linkage between the two entities.

We begin with aggregate analysis showing that idiosyncratic shocks in state 9 were positively correlated with economic growth in state 8 during the late 1970s and early 1980s. This implies that a good (bad) news for state 9 was also a good (bad) news for state 8, suggesting that states behaved as complements during that period. However, the relation monotonically reversed post 1984, i.e., good (bad) news for state 9 became bad (good) news for state 8, suggesting that states behaved as substitutes after this period. We attribute this changing relation between idiosyncratic shocks in state 9 and economic growth in state 8 to banking integration between the two states.

In a difference-in-differences (DID) framework, combining the state pairwise banking integration natural experiment with the measurement of non-capital shocks, we show that a one standard deviation negative idiosyncratic shock, 9-C1 , in state 9 increases economic growth in state 8 by 0.05-0.19 pp after the state pair "8- 9" is integrated via a banking linkage.3 This estimation is based on the assumption that the linkages between states are equally strong across all state-pairs. Taking into account the strength and the direction of real linkages between states by considering imports and exports, we find that a one standard deviation negative 9-C1 increases economic growth in state 8 by 0.13-0.19 pp post banking integration.

The effect of idiosyncratic shocks in state 8 on economic growth in state 9 operates via changes in bank loan supply. We employ an instrumental variable (IV) strategy similar in spirit to the granular IV methodology presented in Gabaix and Koijen (2020). Using idiosyncratic shock, 9-C1 , in state 9 combined with banking integration as an instrument for bank lending in state 8,

3The DID estimator is relative to the pre-integration economic growth level.


This is an excerpt of the original content. To continue reading it, access the original document here.


ESRB - European Systemic Risk Board published this content on 01 December 2021 and is solely responsible for the information contained therein. Distributed by Public, unedited and unaltered, on 01 December 2021 15:00:07 UTC.

ę Publicnow 2021
Latest news "Companies"
05:59pWebster Financial Corporation Declares Common and Preferred Dividends
05:57pALERT : Rowley Law PLLC is Investigating Proposed Acquisition of Activision Blizzard Inc.
05:57pEllington Financial Announces Estimated Book Value Per Common Share as of December 31, 2021
05:57pBHP Metallurgical Coal Operations Disrupted by Australia Covid-19 Surge -- Update
05:56pBoston Properties Announces 2021 Tax Treatment of Its Distributions
05:56ponsemi to Announce Fourth Quarter and 2021 Annual Financial Results
05:56pNOVA MINERALS LIMITED (ASX : NVA) Geophysics Confirms Extension of RPM High-Grade Gold Deposit
05:55pSilver Spike Investment Corp. Announces Proposed Initial Public Offering
05:53pVintage Energy Ltd Share Purchase Plan Completed
05:52pYOJEE : Quarterly Activities Report & Appendix 4C - Dec 2021
Latest news "Companies"