Banque nationale de Belgique SA · EURONEXT:BNB

Banque nationale de Belgique SA — Investor Presentation (September 2015)

45 slides from Banque nationale de Belgique SA's investor presentation, Sep 10, 2015

1. Slide 1: Bank Exposures and Sovereign Stress Transmission

   Bank Exposures and Sovereign Stress Transmission Carlo Altavilla Marco Pagano Saverio Simonelli European Central Bank University of Naples University of Naples Federico II, CSEF and EIEF Federico II and CSEF Seminar at the National Bank of Belgium 10 September 2015 The opinions in this presentation

2. Slide 2: Basic idea

   Basic Rubric idea Sovereign Stress Sovereign stress Banks’ balance sheet (via exposures) banks’ risk Capital funding costs constraints bank loans Doubling up: lending rates exposures 2 www.ecb.europa.eu ©

3. Slide 3: Motivation: bank-sovereign nexus and domestic exposures

   Motivation: Rubric bank-sovereign nexus and domestic exposures Italy Spain Germany France Note: The red line is the 24-month rolling correlations between sovereign 3 and bank. The blue line is the average domestic www.ecb.europa.eu © sovereign exposures (of domestic banks) as % of average total asse

4. Slide 4: Motivation: sovereign exposures & lending in the periphery

   Motivation Rubric : sovereign exposures & lending in the periphery Note: Median of periphery countries (Greece, Ireland, Italy, Portugal and Spain) Sample: August 2007 - February 2015. 4 www.ecb.europa.eu ©

5. Slide 5: Sovereign exposures and lending: median core bank

   Sovereign exposures and lending: median core bank Rubric Note: The core countries are Austria, Belgium, France, Germany, Luxemburg and Netherland. Sample: August 2007 - February 2015. 5 www.ecb.europa.eu ©

6. Slide 6: Our contribution

   Our contribution Rubric Apply panel data estimation to euro-area bank-level data to assess: how much bank exposures contributed to transmitting sovereign stress to i. bank default risk ii. lending to firms how exposures themselves reacted to the return on sovereign debt Data are novel under three re

7. Slide 7: Preview of the results

   Preview Rubric of the results For periphery banks with larger sovereign exposures, increases in sovereign risk and in domestic sovereign yields translate into: – larger increase in bank solvency risk – stronger reduction of loans and a sharper increase in lending rates to firms Following increases i

8. Slide 8: Literature

   Literature Rubric We add to research on sovereign-bank nexus, and on the role of banks’ sovereign exposures: Risk transmission from sovereigns to banks Acharya, Drechsler & Schnabl (2014) (“post-bailout changes in sovereign CDS explain changes in bank CDS”) Sovereign shock transmission to bank loans

9. Slide 9: Outline

   Outline Rubric Data and sample structure Assessing the contribution of sovereign exposures on bank solvency risk loans and lending rates Doubling-up: effect of sovereign yields on exposures Conclusions 9 www.ecb.europa.eu ©

10. Slide 10: Outline

    Outline Rubric Data and sample structure Assessing the contribution of sovereign exposures to bank solvency risk loans and lending rates Doubling-up: effect of sovereign yields on exposures Conclusion 10 www.ecb.europa.eu ©

11. Slide 11: Data: variables and sources

    Data: Rubric variables and sources Individual bank data Bank-level balance sheet items (IBSI): main assets, sovereign exposures, loans to NFC & households. Source: ECB Bank-level interest rates (IMIR): on new business volumes. Source: ECB Bank-level capital ratio. Source: SNL Bank-level 5-year CDS p

12. Slide 12: Sample structure

    Sample Rubric structure Monetary Financial Institutions (MFIs) included in the sample, by ownership structure Private Public Foreign 70 65 60 50 37 40 26 26 30 20 13 10 11 10 9 10 7 6 6 5 4 5 4 5 3 0 Spain Belgium Portugal Austria Cyprus Estonia France Germany Greece Italy Latvia Malta Netherlands S

13. Slide 13: Sample structure

    Sample Rubric structure Sample representativeness: ratio of IBSI loans (our data) and syndicated loans to BSI (aggregate) loan data (in %), avg. 2012-15 Syndicated Loans (as % of total) IBSI (as % of total) 100 80 60 40 20 0 Euro Area Germany France Italy Spain 13 www.ecb.europa.eu ©

14. Slide 14: Descriptive statistics

    Descriptive Rubric statistics 14 www.ecb.europa.eu ©

15. Slide 15: Domestic sovereign holdings, by bank ownership

    Domestic Rubric sovereign holdings, by bank ownership Domestic DOMESTIC Public PUBLIC Foreign FOREIGN 18 18 18 15 15 15 12 12 12 CORE COUNTRIES 9 9 9 Core 6 6 6 3 3 3 0 0 0 2007m7 2009m1 2010m7 2012m1 2013m7 2015m1 2007m7 2009m1 2010m7 2012m1 2013m7 2015m1 2007m7 2009m1 2010m7 2012m1 2013m7 2015m1 1

16. Slide 16: Outline

    Outline Rubric Data and sample structure Assessing the contribution of sovereign exposures on bank solvency risk loans and lending rates Doubling-up: effect of sovereign yields on exposures Conclusion 16 www.ecb.europa.eu ©

17. Slide 17: Breakdown by exposures, periphery countries

    Breakdown Rubric by exposures, periphery countries Low-Exposure Banks High-Exposure Banks Note: Bank CDS are computed as the average of the CDS of the individual banks. Sovereign CDS is computed as the simple average of the sovereign CDS. The countries are Ireland, Italy, Portugal and Spain. The gro

18. Slide 18: Breakdown by exposures, core countries

    Breakdown Rubric by exposures, core countries Low-Exposure Banks High-Exposure Banks Note: Bank CDS are computed as the average of the CDS of the individual banks. Sovereign CDS is computed as the simple average of the sovereign CDS. The countries are Austria, France, Germany, and Netherlands. The g

19. Slide 19: Empirical methodology

    Empirical Rubric methodology • We estimate a panel data model: CDS ijt i t 1 Sov .CDS jt 2 Sov .Exp ijt 1 3 S ov.CDS Sov. E x pijt 1 4 X ijt 1 5Y j t 1 ijt The coefficient 3 captures the association between sovereign and bank credit risk due to sovereign exposures Controls: Time fixed effects contro

20. Slide 20: Domestic exposures and sovereign risk

    Domestic exposures and sovereign risk Rubric CDS ijt i t 1 Sov .CDS jt 2 Sov . Expijt 1 3 Sov .CDS So v . Expijt 1 4 X ijt 1 ijt Dependent variable: Bank CDS. The periphery countries are Cyprus, Ireland, Italy, Portugal and Spain. The core countries are Austria, Belgium, Finland, France, Germany and

21. Slide 21: Breaking down the effect of sovereign risk

    Breaking down the effect of sovereign risk Rubric Direct effect Sovereign- exposure related effect Baseline effect for stressed banks: 32 bps 100 bps Sov.CDS Exposure-related effect for the median bank in stressed countries: 30 bps 21 www.ecb.europa.eu ©

22. Slide 22: In-sample contribution of median banks' dom. exposure to its risk over time

    In-sample Rubric contribution of median banks’ dom. exposure to its risk over time 22 www.ecb.europa.eu ©

23. Slide 23: Robustness checks

    Robustness Rubric checks Check no. 1 (potential endogeneity of government CDS): Several papers argue that risk transmission can go (and actually went) both ways: Acharya et al. (2014), Cooper and Nikolov (2013) and Leonello (2014) Reverse causality should concern mainly SIFI, whose distress would pr

24. Slide 24: Outline

    Outline Rubric Data and sample structure Assessing the contribution of sovereign exposures on bank solvency risk loans and lending rates Doubling-up: effect of sovereign yields on exposures Conclusion 24 www.ecb.europa.eu ©

25. Slide 25: Sovereign exposures and lending: periphery banks

    Sovereign exposures and lending: periphery banks Rubric Note: Residuals of loan and sovereign exposure regressions on bank fixed effects and time dummies. The periphery countries are Greece, Ireland, Italy, Portugal and Spain. Sample: August 2007 - February 2015. 25 www.ecb.europa.eu ©

26. Slide 26: Sovereign exposures and lending: core banks

    Sovereign exposures and lending: core banks Rubric Note: Residuals of loan and sovereign exposure regressions on bank fixed effects and time dummies. The core countries are Austria, Belgium, France, Germany, Luxemburg and Netherland. Sample: August 2007 - February 2015. 26 www.ecb.europa.eu ©

27. Slide 27: Sovereign stress and bank lending to firms

    Sovereign Rubric stress and bank lending to firms Loansijt i t 1 So v .Yield jt 2 2 S ov . Expi jt 3 3 Sov .Yiel d jt 2 So v . Exp ijt 3 4 X ijt 2 ijt . Dependent variable: Total loans to NFC (% of asset). The periphery countries are Ireland, Italy, Portugal and Spain. The core countries are Austria

28. Slide 28: In-sample contribution of median banks' dom. exposure to its lending

    In-sample Rubric contribution of median banks’ dom. exposure to its lending www.ecb.europa.eu ©

29. Slide 29: Alternative measure of sovereign stress: surprises in yields

    Alternative Rubric measure of sovereign stress: surprises in yields 29 www.ecb.europa.eu ©

30. Slide 30: Endogeneity: effect of exposures on bank-level loan demand?

    Endogeneity: Rubric effect of exposures on bank-level loan demand? Dependent variable: proxy of bank-level demand for loans = BLS data on loan demand bank-level maturity composition of loans 30 www.ecb.europa.eu ©

31. Slide 31: Endogeneity: effect of sovereign exposures on bank-level NPLs?

    Endogeneity: Rubric effect of sovereign exposures on bank-level NPLs? Dependent variable: bank-level non-performing loans / gross loans all non-stressed stressed (1) (2) (3) (4) (5) (6) Sov.Yield10yt-2 0.89*** 0.21 0.47** (0.17) (0.34) (0.20) Sov.Exposurest-3 0.21* 0.15* 0.07 0.06 0.29** 0.35** (0.1

32. Slide 32: Sovereign stress and bank lending: lending rates

    Sovereign Rubric stress and bank lending: lending rates Lending.Rateijt i t 1Sov.Yield jt 2 Sov.Expijt 1 3Sov.Yield jt Sov.Expijt 1 4 X ijt 1 ijt . all non-stressed stressed (1) (2) (3) (4) (5) (6) Sov.Yield10yt 0.15*** -0.09 0.11*** (0.02) (0.09) (0.04) Sov.Exposures t-1 0.02 0.00 -0.02 0.00 -0.03*

33. Slide 33: Sovereign stress and bank lending: lending rates

    Sovereign Rubric stress and bank lending: lending rates Lending .Rateijt i t 1Sov.Yield jt 2 Sov.Expijt 1 3 Sov.Yield jt Sov.Expijt 1 4 X ijt 1 5Y jt 1 ijt . all non-stressed stressed (1) (2) (3) (4) (5) (6) Sov.Yield10y t 0.15*** -0.09 0.11*** (0.02) (0.09) (0.04) Sov.Exposurest-1 0.02 0.00 -0.02 0

34. Slide 34: In-sample contribution of median banks' dom. exposure to its loan rate

    In-sample Rubric contribution of median banks’ dom. exposure to its loan rate 34 www.ecb.europa.eu ©

35. Slide 35: Outline

    Outline Rubric Data and sample structure Assessing the contribution of sovereign exposures on bank solvency risk loans and lending rates Doubling-up: effect of sovereign yields on exposures Conclusion 35 www.ecb.europa.eu ©

36. Slide 36: Nexus between sovereign yields and exposures

    Nexus Rubric between sovereign yields and exposures “Carry trade” hypothesis: banks increase exposures when they expect high returns on sovereign debt especially if they are undercapitalized: “betting for resurrection” “Moral suasion” hypothesis: at times of sovereign stress banks should be more wil

37. Slide 37: Determinants of growth in domestic sovereign exposures

    Determinants Rubric of growth in domestic sovereign exposures Sov.Expijt i t jt 1Publici Sov.Yield jt 2 Foreigni Sov.Yield jt 3 Exp. Ret jt Levijt 1 4 X ijt 1 ijt . non-stressed stressed (1) (2) (3) (4) (5) (6) (7) Public x Sov.Yield10y t 0.56* 0.62** 1.59*** 1.07*** 1.06*** (0.30) (0.31) (0.42) (0.

38. Slide 38: Outline

    Outline Rubric Data and Sample structure Identifying the banks’ balance sheet channel Effects on bank solvency risk Effects on Loans and Lending rates Doubling-up: Effect of sovereign yields on exposures Conclusions 38 www.ecb.europa.eu ©

39. Slide 39: Conclusions

    Conclusions Rubric During and after the euro area debt crisis, the exposures of banks to domestic sovereign debt in periphery countries: amplified the impact of sovereign credit risk on the credit risk of those banks amplified the impact of sovereign stress on lending and loan rates High yields and

40. Slide 40: Thank you!

    Rubric Thank you! 40 www.ecb.europa.eu ©

41. Slide 41: Background slides

    Rubric Background slides 41 www.ecb.europa.eu ©

42. Slide 42: Sample structure

    Sample Rubric structure Monetary Financial Institutions (MFIs) included in the sample by ownership structure Total Domestic Foreign Private Public Austria 9 8 0 1 Belgium 10 2 1 7 Cyprus 5 3 1 1 Estonia 4 0 0 4 Finland 7 3 0 4 France 37 30 3 4 Germany 65 25 26 14 Greece 6 6 0 0 Ireland 13 2 3 8 Ital

43. Slide 43: Domestic sovereign holdings as % of main assets

    Domestic Rubric sovereign holdings as % of main assets Germany France 0.14 0.14 0.12 0.12 0.1 0.1 0.08 0.08 0.06 0.06 0.04 0.04 0.02 0.02 0 0 2010 2012 2015 2010 2012 2015 Italy Spain 0.14 0.14 0.12 0.12 0.1 0.1 0.08 0.08 0.06 0.06 0.04 0.04 0.02 0.02 0 0 2010 2012 2015 2010 2012 2015 Note: The shad

44. Slide 44: Sovereign exposures and yield on sovereign debt

    Sovereign Rubric exposures and yield on sovereign debt Dependent variable: domestic exposures. The variable "t1cer08" is the Tier1 capital ratio at the end of 2008. The periphery countries are Ireland, Italy, Portugal and Spain. The core countries are Austria, Belgium, Estonia, Finland, France, Germ

45. Slide 45: Empirical methodology

    Empirical Rubric methodology • We estimate a panel data model: Our coefficient of interest, 3, captures the association between sovereign and bank credit risk due to the banks balance sheet channel Controls: Time fixed effect control for aggregate determinants of credit spreads (demand channel, ulti

Banque Nationale De Belgique — Investor Presentation (September 2015) on 6ix