Correction to: End of the sovereign-bank doom loop in the European Union? The Bank Recovery and Resolution Directive

Size: px

Start display at page:

Download "Correction to: End of the sovereign-bank doom loop in the European Union? The Bank Recovery and Resolution Directive"

Melissa Casey
5 years ago
Views:

JEvolEcon https://doi.org/10.1007/s00191-018-0577-1 CORRECTION Correction to: End of the sovereign-bank doom loop in the European Union?

1 JEvolEcon CORRECTION Correction to: End of the sovereign-bank doom loop in the European Union? The Bank Recovery and Resolution Directive Giovanni Covi 1,2 & Ulrich Eydam 3 # Springer-Verlag GmbH Germany, part of Springer Nature 2018 Correction to: J Evol Econ The publisher regrets its mistake in capturing Tables 1-7 incorrectly. The correct Tables 1, 2, 3, 4, 5, 6 and 7 are shown below: The original article has been corrected. The online version of the original article can be found at * Giovanni Covi giovanni.covi@univr.it; g.covi@sussex.ac.uk Ulrich Eydam ulrich.eydam@uni-potsdam.de Economics Department, University of Verona, Santa Marta, via Cantarane, Verona, Italy University of Sussex SPRU, Sussex House, Falmer - BN19RH, Brighton, UK Chair of Growth, Integration and Sustainable Development, University of Potsdam, Campus Griebnitzsee, House 1, Room , August-Bebel-Str, Potsdam, Germany

2 G. Covi, U. Eydam Table 1 List of banks by country and total assets Total assets AVG Country Banks Billion Euro Country Weight Austria Erste Group % Belgium KBC Bank % Dexia Credit Local % France BNP Paribas % Credit Agricole % Societe Generale % Germany Commerzbank % Deutsche Bank % IKB Industriebank % Bayerische Landesbk % Ireland Allied Irish bank % Bank of Ireland % Permanent Tsb % Italy Monte dei Paschi % Popolare di Milano % Intesa Sanpaolo % Unicredit Bank % Mediobanca % Unione di Banche % Netherlands ING Bank % Portugal Banco Comercial Portugues % Spain BBVA % Banco de Sabadell % Banco Popular Espanol % Banco Santander % Bankinter % Sweden Nordea Bank % Svenska Bank % Swedbank % Denmark Danske bank %

3 Correction to: End of the sovereign-bank doom loop in the European... Table 2 Summary statistics of bank and sovereign CDS Quantiles Full Sample (01/01/2012 to 31/03/2016) n Mean S.D. Min.25 Mdn.75 Max Bank CDS (bp) Sovereign CDS (bp) Δ Log( Bank CDS) % 3% -43% -1% 0% 0% 33% Δ Log( Sovereign CDS) % 5% -56% -1% 0% 1% 53% Δ Log( Bank Stock Price) % 3% -108% -1% 0% 1% 109% Δ Log( Foreign Exposure CDS) % 2% -34% -1% 0% 1% 20% Period 1 (01/01/2012 to 01/01/2013) Bank CDS (bp) Sovereign CDS (bp) Δ Log( Bank CDS) % 3% -43% -1% 0% 0% 17% Δ Log( Sovereign CDS) % 6% -56% -2% 0% 2% 50% Δ Log( Bank Stock Price) % 4% -34% -2% 0% 2% 43% Δ Log( Foreign Exposure CDS) % 2% -13% -2% 0% 1% 11% Period 2 (01/01/2013 to 01/01/2014) Bank CDS (bp) Sovereign CDS (bp) Δ Log( Bank CDS) % 2% -16% -1% 0% 0% 18% Δ Log( Sovereign CDS) % 5% -51% -1% 0% 1% 53% Δ Log( Bank Stock Price) % 3% -20% -1% 0% 1% 23% Δ Log( Foreign Exposure CDS) % 2% -12% -1% 0% 1% 12% Period 3 (01/01/2014 to 01/01/2015) Bank CDS (bp) Sovereign CDS (bp) Δ Log( Bank CDS) % 3% -29% -1% 0% 0% 25% Δ Log( Sovereign CDS) % 5% -39% -1% 0% 1% 41% Δ Log( Bank Stock Price) % 3% -35% -1% 0% 1% 34% Δ Log( Foreign Exposure CDS) % 2% -16% -1% 0% 1% 14% Period 4 (01/01/2015 to 01/01/2016) Bank CDS (bp) Sovereign CDS (bp) Δ Log( Bank CDS) % 3% -24% -1% 0% 1% 28% Δ Log( Sovereign CDS) % 6% -39% -1% 0% 0% 39% Δ Log( Bank Stock Price) % 3% -108% -1% 0% 1% 109% Δ Log( Foreign Exposure CDS) % 3% -34% -1% 0% 1% 20% Period 5 (01/01/2016 to 31/03/2016) Bank CDS (bp) Sovereign CDS (bp) Δ Log( Bank CDS) % 4% -24% 0% 0% 1% 33% Δ Log( Sovereign CDS) % 6% -34% -1% 0% 1% 43% Δ Log( Bank Stock Price) % 4% -36% -2% 0% 1% 36% Δ Log( Foreign Exposure CDS) % 3% -16% -1% 0% 1% 18%

4 G. Covi, U. Eydam Table 3 Bank CDS and changes in sovereign CDS explanatory power: time and bank fixed effects. This table shows the effect of sovereign credit risk on bank credit risk before, during and after the implementation of the BRRD. Standard errors are clustered at the bank level and bootstrapped Δ Log (BANK CDS) (1) Fixed effects (2) Fixed effects (3) Fixed effects (4) Fixed effects (5) Fixed effects (6) Fixed effects Variables Δ Log (SOV CDS) β * *** ** ** ( ) (0.0132) (0.0114) (0.0151) (0.0208) (0.0221) Lagged Δ Log (BANK CDS) ** *** ** * (0.0164) (0.0357) (0.0194) (0.0194) (0.0418) (0.0370) Δ Log (FX) ** *** *** (0.0234) (0.0633) (0.0401) (0.0291) (0.0222) (0.0484) Δ Log (SP) *** *** *** ** ** (0.0130) (0.0185) (0.0155) (0.0151) (0.0211) (0.0340) Δ Log (VDAX) 0.230*** 0.106* 0.367* *** (0.0417) (0.0631) (0.201) (0.193) (0.176) (0.0838) Δ Log (ITRAXEU) *** ** 1.079** 1.037* (0.228) (0.0862) (0.617) (0.499) (0.540) (0.248) Constant *** ** *** ( ) ( ) ( ) ( ) ( ) ( ) Observations 28,236 7,267 7,141 6,093 6,188 1,547 R-squared No. Banks Bank FE Yes Yes Yes Yes Yes Yes Time FE Yes Yes Yes Yes Yes Yes ***, **, and * indicates statistical significance at the 1%, 5%, and 10% level, respectively. The regression specification is given by: Δ log(bank CDSijt)=αi + ζt + βδ log(sov CDSjt)+ωΔ log(bank CDSit 1)+γΔ log(fxjt)+φδ log(spit)+δδ log (Xt)+εijt

5 Correction to: End of the sovereign-bank doom loop in the European... Table 4 Banks CDS: BRRD dummy and sovereign CDS interaction term. This table examines the robustness of our main results by reporting the coefficients after having included the BRRD dummy and the interaction term between the BRRD dummy and sovereign CDS, columns (2) and (3), respectively. The same specification is performed at monthly frequency columns (4) to (6), and the public sector purchase program of the ECB is added as further control variable. Standard errors are clustered at the bank level and bootstrapped Δ Log (Bank CDS) (1) Fixed effects (2) Fixed effects (3) Fixed effects (4) Fixed effects (5) Fixed effects (6) Fixed effects Variables Δ Log (SOV CDS) β * ( ) * ( ) ** (0.0108) 0.207*** (0.0379) 0.207*** (0.0379) 0.223*** (0.0407) Interaction Term: Θ BRRD dummy *** ( ) *** ( ) *** ( ) *** ( ) (BRRD dummy) * SOV CDS ** * (0.0632) Δ Log (PSPP) (0.941) (0.892) (0.859) Constant *** ( ) *** ( ) *** ( ) (0.0388) (0.0372) (0.0387) Observations 28,236 28,236 28,236 1,252 1,252 1,252 R-squared No. Banks Bank FE Yes Yes Yes Yes Yes Yes Time FE Yes Yes Yes Yes Yes Yes Lagged Δ Log (Bank CDS) Yes Yes Yes No No No Control variables Table 3 Table 3 Table 3 Table 3 Table 3 Table 3 Data frequency Daily Daily Daily Monthly Monthly Monthly ***, **, and * indicates statistical significance at the 1%, 5%, and 10% level, respectively. The regression specification is given by: Δlog BANK CDSijt ¼ α i þ ζ t þ βδlog SOV CDSjt þωδlog ð BANK CDS it 1 Þ þγδlog FX jt þ φδlog ð SP it ÞþδΔlog X t BRRD j * ΔlogðSOV CDSjt þ ε it ð Þ þθ1brrd j þ Θ2

6 G. Covi, U. Eydam Table 5 Sovereign CDS and financial sector distress: country and time fixed effects. This table shows the effect of banks credit risk on sovereign credit risk before and after the implementation of the BRRD. Standard errors are clustered at country level and bootstrapped by 200 replications. Δ Log (SOV CDS) (1) (2) (3) (4) (5) (6) Fixed Effects Fixed Effects Fixed Effects Fixed Effects Fixed Effects Fixed Effects Variables Δ Log (FSD) β 0.203*** 0.301*** 0.324*** 0.217** (0.0641) (0.0810) (0.123) (0.105) (0.155) (0.136) Lagged Δ Log (SOV CDS) *** *** *** *** *** *** (0.0283) (0.0727) (0.0489) (0.0329) (0.0394) (0.0430) Δ Log (VDAX) 0.246* *** (0.140) (0.218) (0.433) (0.105) (0.130) (0.297) Δ Log (ITRAXEU) ** ** (0.545) (0.324) (1.644) (0.237) (1.144) (1.627) Constant (0.0111) ( ) ( ) ( ) ( ) (0.0236) Observations 8,946 2,502 2,351 1,808 1, R-squared No. Countries Country FE Yes Yes Yes Yes Yes Yes Time FE Yes Yes Yes Yes Yes Yes ***, **, and * indicates statistical significance at the 1%, 5%, and 10% level, respectively. The regression specification is given by: Δ log(sov CDSit)=αi + ζt + βδ log(fsdit)+ωδ log(sov CDSit 1)+δΔ log (Xt)+εit

7 Correction to: End of the sovereign-bank doom loop in the European... Table 6 Sovereign CDS: BRRD dummy and financial sector distress interaction term. This table examines the robustness of our main results by reporting the coefficients after having included the BRRD dummy, and the interaction term between the BRRD dummy and financial sector distress, columns (2) and (3), respectively. The same specification is performed at monthly frequency columns (4) to (6), and further control variables are added: debt to GDP ratio column and the ECB s PSPP. Standard errors are clustered at country level and bootstrapped by 200 replications. Δ Log (BANK CDS) (1) Fixed effects (2) Fixed effects (3) Fixed effects (4) Fixed effects (5) Fixed effects (6) Fixed effects Variables Δ Log (FSD) β 0.203*** (0.0641) 0.203*** (0.0642) 0.253*** (0.0598) 0.215* (0.126) 0.215* (0.126) 0.265* (0.152) Interaction Term: τ BRD dummy ** ( ) * ( ) (0.0256) (0.0265) (BRD dummy) *FSD *** (0.128) (0.296) Δ Log (PSPP) (4.059) (4.162) (4.381) Constant (0.0111) (0.0119) (0.0117) (0.168) (0.169) (0.162) Observations 8,496 8,496 8, R-squared No. Countries Country FE Yes Yes Yes Yes Yes Yes Time FE Yes Yes Yes Yes Yes Yes Lagged Δ Log (BANK CDS) Yes Yes Yes No No No Δ Log (DEBT/ GDP) No No No Yes Yes Yes Control variables Table 5 Table 5 Table 5 Table 5 Table 5 Table 5 Data frequency Daily Daily Daily Monthly Monthly Monthly ***, **, and * indicates statistical significance at the 1%, 5%, and 10% level, respectively. The regression specification is given by: Δlog ðsov CDSit Þ ¼ αi þ ζ t þ βδlog ðfsdit Þ þωδlog ðsov CDSit 1 Þ þτ 1BRRD j þτ 2 BRRD j * ΔlogðFSDit þ δδlog ð X t Þþεit

8 G. Covi, U. Eydam Table 7 Bank CDS and sovereign CDS explanatory power: time and bank/country fixed effects. This table shows the effect of sovereign credit risk on bank credit risk and vice versa at monthly frequency, before and after the implementation of the BRRD Δ Log (Bank CDS) (1) Fixed effects (2) Fixed effects (3) Fixed effects Δ Log (SOV CDS) (4) Fixed effects (5) Fixed effects (6) Fixed effects Variables Variables Δ Log (SOV CDS) β 0.205* (0.0387) * (0.0567) 0.109** (0.0682) Δ Log (FSD) β 0.286* (0.158) *** (0.156) (0.203) Δ Log (PSPP) (0.791) Δ Log (PSPP) (4.108) Constant *** ( ) ***( ) (0.0530) Constant 0.104* (0.0538) 0.142*** (0.0266) (0.254) Observations 1, Observations R-squared R-squared No. Banks No. Countries Bank FE Yes Yes Yes Country FE Yes Yes Yes Time FE Yes Yes Yes Time FE Yes Yes Yes Δ Log (X) Yes Yes Yes Δ Log (X) Yes Yes Yes Δ Log (FX) Yes Yes Yes Δ Log (DEBT/GDP) Yes Yes Yes Δ Log (SP) Yes Yes Yes Channel Channel Data frequency Monthly Monthly Monthly Data frequency Monthly Monthly Monthly Standard errors are clustered at the bank/country level and bootstrapped. ***, **, and * indicates statistical significance at the 1%, 5%, and 10% level, respectively. The regression specification of columns (1) to (3) is given by Δ log(bank CDS ijt )=α i + ζ t + βδ log(sov CDS jt )+γδ log(fx jt )+φδ log(sp it )+δδ log (X t )+ε ijt. The regression specification of columns (4) to (6) is given by Δ log(sov CDS it )=α i + ζ t + βδ log(fsd it )+δδ log (X t )+ε it

A tale of two overhangs: the nexus of financial sector and sovereign credit risks

A tale of two overhangs: the nexus of financial sector and sovereign credit risks VIRAL V. ACHARYA Professor of Finance New York University Stern School of Business AMAR DRECHSLER Assistant Professor of