当银行退缩时：信用和流动性供应

When banks hold back: credit andliquidity provision Carlo Altavilla, Massimo Rostagno,Julian Schumacher Disclaimer:This paper should not be reported as representing the views of the European Central Bank(ECB). The views expressed are those of the authors and do not necessarily reflect those of the ECB. Abstract Banks are reluctant to tap central bank backup liquidity facilities and use the borrowed funds forloans to the real economy. We show that excessively parsimonious borrowing and lending can arisein a stigma-free model where the banking sector has an incentive to overissue deposits. Banks don’theed the central bank’s call for more credit to finance investment because they simply ignore thecollective gains from stronger activity in their atomistic decisions. Central banks can address thismarket failure by disintermediating market-based finance. A lender-of-last-resort (LOLR) system inwhich the central bank offers liquidity liberally but on non-concessionary conditions improves over apure laissez-faire arrangement, where asset liquidation in the marketplace is the only source ofemergency liquidity. But under LOLR banks remain reluctant to intermediate. Credit easing (CE) andquantitative easing (QE), instead, can stimulate bank borrowing and repair the broken nexus betweenliquidity provision and credit. Empirical analysis using bank-level and loan-by-loan data supports ourmodel predictions. We find no empirical connection between loans and borrowed reserves obtainedfrom conventional refinancing facilities. In contrast, there is a robust connection between loans andstructural sources of liquidity: reserves borrowed under a CE program or non-borrowed, i.e. acquiredfrom a QE injection. We also find that firms with greater exposure to banks borrowing in a CE programor holding larger volumes of non-borrowed reserves increase employment, sales, and investment. JEL Codes: E5; E43; G2 Keywords: Lending of last resort, credit easing, quantitative easing, loans, reserves Non-technical summary Over the past four decades, an influential body of economic research has highlightedbanks' inherent tendency to overextend credit, often fuelling financial booms that end incrises.This research has inspired a far-reaching regulatory reform agenda that hasstrengthened the global financial system, particularly banks. In parallel, however, analystshave long observed commercial banks borrowing from central banks only reluctantly, usingthe borrowed liquidity sparingly for credit creation. This paper investigates a factor behind banks’ reluctance to intermediate that has beenoverlooked in prior research. In a model of banking where borrowing from the central bankis stigma-free, we demonstrate that overly cautious borrowing and lending can emerge evenwhen banks have an underlying incentive to overissue deposits and overextend credit. Weshow that a lending of last resort facility can enhance welfare relative to a laissez-faire regimein which the only source of emergency liquidity for banks in the face of a deposit run is fire-sale asset liquidation. However, banks’ failure to internalise the superior social objective leadsto deficient bank intermediation and under-investment in a lender-of-last-resort regime ofliquidity provision. Credit easing and quantitative easing policies – changing the nature ofliquidity provision into something cheaper and more persistent – can nudge banks towardslevels of intermediation activity that match social optima. Using bank-level and loan-level data, the empirical analysis supports the predictions of ourmodel. We find no significant relationship between loans and reserves borrowed in standardrefinancing facilities. In contrast, there is a strong connection between loans and structuralsources of liquidity, such as reserves obtained through credit easing programs or thoseinjected through quantitative easing interventions. Furthermore, we show that firms withhigher exposure to banks participating in CE programs or holding larger volumes of non-borrowed reserves exhibit significant increases in employment, sales, and investment. 1 Introduction Over the past forty years, an influential strand of economic literature has pointed to banks’inherent tendency to overextend their exposures and occasionally fuel financial booms thatend badly. Among many other contributions, the literature on bank credit cycles and fragilebanking includes Bernanke and Gertler (1989), Kiyotaki and Moore (1997), Diamond andRajan (2001), Rajan (2006), Adrian and Shin (2008), Lorenzoni (2008), Shleifer and Vishny(2009), Stein (2012) and Acharya and Rajan (2022). This body of research deserves muchrecognition, because it arguably contributed to the rollout of an ambitious regulatory reformthat put the world financial system – and especially banks – on safer ground (see e.g. BCBS2015, BCBS 2016). At the same time, on a somewhat contrarian note, analysts have longobserved commercial banks borrowing from central banks only reluc