• No results found

The commitment of European governments to the Bank Recovery and Resolution Directive : an analysis of time-inconsistency in resolution and expected government support

N/A
N/A
Protected

Academic year: 2021

Share "The commitment of European governments to the Bank Recovery and Resolution Directive : an analysis of time-inconsistency in resolution and expected government support"

Copied!
11
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

This table shows the numerical values assigned to the Moody’s credit ratings and credit outlooks. The total numeric value assigned per bank rating is the sum of the numerical values of both panels represented below.

(2)

2

Event date raw returns of European banks

This table reports the event date raw returns for the event banks, and other European banks in the sample. The results are reported separately for the event-, core European- and peripheral European banks

Event Banks Deutsche Bank Event Banca Monte Dei Paschi Event Caixabank Event

Event Dates 26/09/2016 07/12/2016 10/03/2016

Deutsche Bank -7.79% 3.47% -0.61%

Banca Monte Dei Paschi 1.42% 10.25% 0.00%

Caixabank -2.06% 2.01% 2.89%

Core European Banks

BNP Paribas -3.32% 2.23% 0.96% ING Group -2.24% 0.00% 0.04% Deutsche Bank -7.79% 3.47% -0.61% Soiciete Generale -2.99% 3.65% -0.59% Gredit Agricole -2.78% 0.93% -0.16% KBC Group -1.78% -1.80% -0.22%

ABN Amro Group -0.86% 1.19% 1.41%

Natixis -2.07% 2.35% -2.48%

Erste Group Bank -1.83% -0.25% 0.37%

Commerzbank -4.16% 3.75% -0.66% Faiffeisen Bank -1.94% 0.22% 2.02% Observations 11 11 11 Average -2.89% 1.43% 0.01%

Periphiral European Banks

Banco Santander -2.07% 2.44% 1.06% BBV. Argentaria -3.24% 1.28% 1.35% Intesa Sanpaolo -1.16% 1.65% 1.11% Unicredit -3.70% 9.00% 2.30% Bankia -1.13% 3.18% 2.67% Banco de Sabadell -0.98% 0.15% 4.33% Bank of Ireland -5.53% 1.72% 0.00% Mediobanca -3.96% 4.69% 1.51% Bankinter R -1.20% -1.82% 1.35%

Banco Popular Espanol -2.55% 3.91% 4.53%

Banco BPM -2.36% 6.19% -1.63%

Unione Di Banche Italiane -2.31% 4.58% 1.25%

BPER Banca -1.62% 9.90% 0.89%

Banca Monte Dei Paschi 1.42% 10.25% 0.00%

Caixabank -2.06% 2.01% 2.89% Finecobank -2.05% 1.20% 3.57% Observations 16 16 16 Average -2.16% 3.77% 1.70% Indices

Stoxx Banking Index -2.35% 2.42% -0.52%

Stoxx European Market Index -1.56% 0.90% -1.68%

Total Observations 27 27 27

(3)

Balance sheet data of European banks

This table reports balance sheet data for the European banks in my sample, taken from DataStream. Deposits represent customer deposits, Interbank represents transactions between banks that are not classified as cash deposits from other banks, Equity represents shareholders’ equity, Debt represents debt securities in issue, all reported as a percentage of total liabilities. Total Capital represents the total investment in the company. The results are reported separately for the event-, core European- and peripheral European banks.

(4)

4

Expected government support of European banks

This table reports the foreign long-term deposit rating and the baseline credit assessment for the European banks in my sample, from Moody’s Investor Service taken from Bloomberg. This table also reports the measure of expected government support (Support), measured as the difference between a banks foreign long term deposit rate and its baseline credit assessment. The results are reported separately for the event-, core European- and peripheral European banks.

(5)

Abnormal Returns Event Banks

Panel A of this table reports the abnormal returns of the event banks (Deutsche Bank, Banca Monte Dei Paschi, Caixabank) at their respective event dates. SARi represents standardized abnormal returns based on the standard deviation of abnormal returns during the respective event’s 5-day event window. ESARi represents standardized abnormal returns based on the standard deviation of abnormal returns during the respective event’s 250-day estimation window. Panel B reports the CARi and SCARi defined in section 4.2. ***,**, and * denote the 1%, 5% and 10% levels, respectively; of the normal distribution. Significance of ARi is reported for both measures of standardized abnormal returns. SARi, ESARi and SCARi can be interpreted as t-values and are reported in parenthesis.

Event Banks Panel A Panel B

ARi SARi ESARi CARi SCARi

Deutsche Bank -5.02%* (-1.860) -4.98%** (-2.022)

-5.02%** (-2.573)

Banca Monte Dei Paschi 8.66% (1.082) 3.04% (0.531)

8.66% (1.616)

Caixabank 4.46%* (1.853) 4.43% (1.120)

4.46%*** (2.695)

Table 5

Cumulative Average Abnormal Returns – General Sector Contagion

Table 6 reports the cumulative abnormal returns and their respective test statistics for all events. ***,**, and * denote the 1%, 5% and 10% levels, respectively; of the normal distribution. CAAR represent total contagion in the banking system during the event dates. TS1 is defined as:   𝑁#$$%& ≈ 𝑁(0,1) and is reported in the parenthesis and can be interpreted as t-values.

Events CAAR TS1

Deutsche Bank -2.03%** (-2.031)

Banca Monte Dei Paschi 3.71%** (2.040)

(6)

6

Abnormal Returns all banks – Detailed Sector Contagion

Table 7 reports the Abnormal Returns of the 27 banks in the sample, on each event date. ***,**, and * denote the 1%, 5% and 10% levels, respectively. The maximum of SARi,andESARi is reported in the parenthesis, for each bank during each event. Significance is based on the maximum of SARi,and ESARi reported in the parenthesis,. SARi,andESARi can be interpreted as t-values.

Event Banks Deutsche Bank Monte Dei Paschi Caixabank

Deutsche Bank -5.02%** (-2.573) 2.07% (0.892) 1.64% (1.029) Monte Dei Paschi 5.52%* (1.769) 8.66% (1.616) 3.42% (0.954)

Caixabank 0.12% (0.058) 0.76% (0.373) 4.46%*** (2.695)     Core Europe     BNP Paribas -0.98%* (-1.812) 0.78% (0.574) 2.90%*** (2.644) ING Group 0.06% (0.101) -1.41% (-1.032) 2.08%* (1.743) Deutsche Bank -5.02%** (-2.573) 2.07% (0.892) 1.64% (1.029) Soiciete Generale -0.17% (-0.414) 1.95% (1.187) 1.73% (1.323) Gredit Agricole -0.38% (-0.460) -0.51% (-0.334) 1.78% (1.197) KBC Group 0.10% (0.117) -3.07%* (-1.887) 1.28% (1.138)

ABN Amro Group 0.71% (0.660) 0.15% (0.088) 2.75% (1.541)

Natixis 0.32% (0.287) 1.03% (0.906) -0.48% (-0.291)

Erste Group Bank -0.22% (-0.312) -1.33% (-0.742) 1.86% (1.246)

Commerzbank -1.66% (-1.248) 2.44% (1.311) 1.50% (1.230) Faiffeisen Bank -0.15% (-0.149) -0.84% (-0.456) 3.99%** (2.021) Periphiral Europe     Banco Santander 0.66% (0.662) 0.83% (0.493) 3.31%*** (2.599) BBV. Argentaria -0.88% (-1.467) -0.07% (-0.043) 3.27%*** (2.891) Intesa Sanpaolo 1.71% (1.206) 0.02% (0.010) 3.31%** (2.070) Unicredit -0.34% (-0.165) 7.38%** (2.462) 4.71%*** (2.710) Bankia 0.93% (0.745) 2.03% (0.911) 4.20%** (2.516) Banco de Sabadell 1.24% (0.782) -1.08% (-0.508) 6.01%*** (3.068) Bank of Ireland -2.73%* (-1.700) 0.29% (0.108) 1.67% (0.932) Mediobanca -1.15% (-0.856) 3.03% (1.408) 3.63%** (2.597) Bankinter R 0.20% (0.163) -2.79%** (-2.344) 2.71%** (2.145) Banco Popular Esp. -0.20% (-0.170) 3.07% (0.913) 6.25%*** (3.291)

Banco BPM 1.55% (1.253) 4.69% (1.207) 1.01% (0.441)

Unione Banche It. 0.96% (0.586) 3.15% (1.028) 3.78% (1.832)

BPER Banca 2.01% (1.038) 7.89%** (2.371) 3.38% (1.706)

Monte Dei Paschi 5.52%* (1.769) 8.66% (1.616) 3.42% (0.954)

Caixabank 0.12% (0.058) 0.76% (0.373) 4.46%*** (2.695)

(7)

The Determinants of Expected Government Support

The dependent variable is the ratings-based measure of expected government support, defined as the difference in rating notches between the baseline credit assessment and the long term foreign deposit rating. Liabilities / GDP is defined as the banks total liabilities divided by the home country GDP. Deficit / GDP is defined as the general government budget deficit divided by the country GDP. Interbank / Liabilities and Debt / Liabilities are defined as the banks interbank loans and total debt divided by total liabilities. Specification 1 includes entity fixed effects, specification 2 to 5 include entity and time fixed effects. Standard errors are clustered by bank-year. All specifications are estimated using OLS. ***,**, and * denote the 1%, 5% and 10% levels, respectively; of the normal distribution, t values are shown in italics.

(1) (2) (3) (4) (5) Liabilities / GDP -0.0006** -0.0001 -0.0002 -0.0002 -0.0002 -2.050 -0.630 -0.700 -0.890 -0.930 Deficit / GDP 1.7398*** 1.5016*** 1.5324*** 1.5337*** 2.2100*** 3.410 3.410 3.560 3.540 2.920 Interbank / Liabilities 1.9420 1.6944 -1.6789 1.280 0.940 -0.210 Debt / Liabilities -1.7296 -2.0518 -0.770 -0.750 Liabilities / GDP x Deficit / GDP -0.0004 -1.440

Interbank / Liabilities x Debt / Liabilities 10.9343

0.480

Constant 3.7122 3.1985 3.1414 3.6988 3.8160

Observations 12160 12160 12160 12160 12160

R-squared within 0.0282 0.1941 0.1959 0.2086 0.2109

Bank fixed effects Yes Yes Yes Yes Yes

(8)

8

The determinants of Contagion / Bank Interconnectedness

The dependent variable is the abnormal return observed for bank i during the event dates. Support is defined as the difference in rating notches between the baseline credit assessment and the long term foreign deposit rating, and represents expected government support. Interbank / Liabilities, Debt / Liabilities, Equity / Liabilities are defined as the banks interbank loans, total debt and total equity as a fraction of the respective liabilities. Deficit / GDP is defined as the general government budget deficit divided by the country GDP. Liabilities / GDP is defined as the banks total liabilities divided by the home country GDP. The dependent variable in Panel A is the observed abnormal return during the event dates. The dependent variable in Panel B is the absolute observed abnormal returns during the event dates. ***,**, and * denote the 1%, 5% and 10% levels, respectively; of the normal distribution, t values are shown in italics.

Panel A Panel B 1 2 3 4 1 2 3 4 Support 0.0034 0.0017 0.0005 -0.0005 0.0050* 0.0026 0.0022 0.0001 0.980 0.500 0.130 -0.100 1.820 1.210 0.940 0.020 Interbank / Liabilities -0.1029 -0.0601 -0.0439 -0.0555 -0.1403** -0.0832* -0.0803* -0.0830* -1.450 -0.880 -0.580 -0.770 -2.550 -1.900 -1.650 -1.710 Deficit / GDP -0.1015 -0.1262 -0.1741 -0.1916 -0.1155 -0.1485** -0.1822** -0.1857** -0.860 -1.140 -1.290 -1.480 -1.260 -2.080 -2.090 -2.130 D1 0.0271*** 0.0295** 0.0270** 0.0362*** 0.0381*** 0.0368*** 2.750 2.640 2.510 5.720 5.300 5.100 D2 -0.0068 -0.0105 -0.0044 -0.0050 -0.790 -1.240 -0.780 -0.880 Liabilities / GDP 0.0000 0.0000 0.0000 0.0000 -0.910 0.370 -0.510 -0.380 Equity / Liabilities 0.0143 -0.1765 0.050 -0.970 Debt / Liabilities 0.0919** 0.0300 2.440 1.190 Constant 0.0055 0.0045 0.0144 -0.0150 0.0081 0.0068 0.0107 0.0202 0.520 0.450 1.030 0.650 0.970 1.050 1.180 0.910 Observations 51 51 51 51 51 51 51 51 R-squared 0.054 0.188 0.211 0.310 0.143 0.499 0.508 0.533

(9)

Pricing expected government support - Data

Table 9 reports the following data for the DB and MPS events: Event date, share price one day prior to event date. Following data at event date: Amount of common share outstanding (reported in thousands), abnormal returns (absolute), foreign long-term credit assessment and baseline credit assessment (numerical and

alphanumerical data reported). Data is acquired from DataStream, Moody’s or calculated according to the event study methodology outlined in section 4.2.

Events DB MPS Event date 26/09/2016 07/12/2016 Pricet-1 10.146 18.9 Shares outstanding t=0 1545162 29321 ARi 5.02% 8.66%

Foreign long-term credit assessment 15 (A3) 7(B2)

Baseline credit assessment 11(Ba1) 2.5(Ca)

Table 10

Pricing expected government support – Historical Default Rates

Table 10 reports the historical default rates corresponding to the foreign long-term credit assessment and the baseline credit assessment for the respective event banks. ‘Difference’ denotes the historical default rate assigned to the value of expected government support divined as the difference in trading notches between the foreign long-term credit assessment and the baseline credit assessment. Historical default rates are reported as a percentage for a 1,5,10,15 and 20-year time window. In the parentheses the numerical values assigned to the Moody’s credit ratings are reported, according to the scale of Appendix 1.

Events Historical Default Rates (%)

n 1 5 10 15 20 DB FLTCA (15) 0.058 0.874 2.371 4.218 6.692 BCA (11) 0.47 5.47 10.355 14.666 20.31 Difference (4) 0.412 4.596 7.984 10.448 13.618 MPS FLTCA (7) 3.213 21.492 34.711 41.095 47.008 BCA (2.5) 30.042 52.763 56.696 58.944 58.944 Difference (4.5) 26.829 31.271 21.985 17.849 11.936

Pricing expected government support – The Equations

1)  

𝑉(𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑  𝑔𝑜𝑣𝑒𝑟𝑚𝑛𝑒𝑡  𝑠𝑢𝑝𝑝𝑜𝑟𝑡) = |𝐴𝑅

@ABC

| ∗ 𝑀𝑎𝑟𝑘𝑒𝑡  𝑉𝑎𝑙𝑢𝑒  

@ABC

2)  

𝑉(𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑  𝑔𝑜𝑣𝑒𝑟𝑚𝑛𝑒𝑡  𝑠𝑢𝑝𝑝𝑜𝑟𝑡) = 𝑓𝑜𝑟𝑒𝑖𝑔𝑛  𝑙𝑜𝑛𝑔  𝑡𝑒𝑟𝑚  𝑑𝑒𝑝𝑜𝑠𝑖𝑡  𝑟𝑎𝑡𝑖𝑛𝑔

@ABC

−   𝐵𝐶𝐴

@ ABC

 

3)  

|𝐴𝑅

@ABC

| ∗ 𝑀𝑎𝑟𝑘𝑒𝑡  𝑉𝑎𝑙𝑢𝑒  

@ABC

=   𝑓𝑜𝑟𝑒𝑖𝑔𝑛  𝑙𝑜𝑛𝑔  𝑡𝑒𝑟𝑚  𝑑𝑒𝑝𝑜𝑠𝑖𝑡  𝑟𝑎𝑡𝑖𝑛𝑔

@ABC

−   𝐵𝐶𝐴

@ABC

4)  

𝑓𝑜𝑟𝑒𝑖𝑔𝑛  𝑙𝑜𝑛𝑔  𝑡𝑒𝑟𝑚  𝑑𝑒𝑝𝑜𝑠𝑖𝑡  𝑟𝑎𝑡𝑖𝑛𝑔

@ABC

−   𝐵𝐶𝐴

@ABC

=   𝑃𝑟. 𝐷𝑒𝑓𝑎𝑢𝑙𝑡

QRASTABC

−   𝑃𝑟. 𝐷𝑒𝑓𝑎𝑢𝑙𝑡

U#$ABC

5)  

𝑓𝑜𝑟𝑒𝑖𝑔𝑛  𝑙𝑜𝑛𝑔  𝑡𝑒𝑟𝑚  𝑑𝑒𝑝𝑜𝑠𝑖𝑡  𝑟𝑎𝑡𝑖𝑛𝑔

@ABC

−   𝐵𝐶𝐴

@ ABC

=   |𝐴𝑅

@ ABC

| ∗ 𝑀𝑎𝑟𝑘𝑒𝑡  𝑉𝑎𝑙𝑢𝑒  

@ ABC

6)   |𝐴𝑅

@ABC

| ∗ 𝑀𝑎𝑟𝑘𝑒𝑡  𝑉𝑎𝑙𝑢𝑒  

@ ABC

=   𝑃𝑟. 𝐷𝑒𝑓𝑎𝑢𝑙𝑡

(10)

10

Expected government support – Pricing Probabilities of Default

Table 11 reports the value of expected government support defined in probabilities of default (difference in trading notches between the foreign long-term credit assessment and the baseline credit assessment) and the EUR value of a 1 percentage point decrease in the respective banks probability of default, for a 1,5,10,15 and 20-year time window. EUR values are reported in thousands.

Table 12

Expected government support – Average Values

Table 12 reports the average market value per Moody’s trading notch, Average EUR value per 1% decrease in probability of default and the average probability of default assigned per Moody’s rating notch. Average taken over a 20-year period and reported for Deutsche Bank and Banca Monte Dei Paschi. EUR values reported in thousands.

Events DB MPS

Average market value per Moody’s trading notch € 196,749 € 10,664

Average EUR value 1% decrease Pr. Default € 462,621 € 2,442

Average probability of default assigned per Moody’s trading notch 1.85% 4.88%

Table 13

Standard Deviations of Abnormal Returns

Table 12 reports the standard deviations of abnormal returns over the event window and estimation window used in the calculation of SARi and ESARi respectively, at all event dates. Standard deviations are reported separately for the event banks and the sample average. The sample average is calculated with exclusion of the event banks.

Event Bank Sample Average

Events Event Window Estimation Window Event Window Estimation Window

AR standard deviation AR standard deviation AR standard deviation AR standard deviation

Deutsche Bank 0.0270 0.0195 0.0133 0.0222

Banca Monte Dei Paschi 0.0801 0.0536 0.0338 0.0225

Caixabank 0.0241 0.0166 0.0238 0.0170

Average Std. Dev. 0.0437 0.0299 0.0236 0.0206

Average difference 1.38% 0.31%

Events n 1 5 10 15 20

DB Value expected government support (%) 0.412 4.596 7.984 10.448 13.618

EUR value 1% decrease Pr. Default € 1,910,184 € 171,235 € 98,571 € 75,325 € 57,790

MPS Value expected government support (%) 26.829 31.271 21.985 17.849 11.936

(11)

Random vs. Fixed effects – Hausman test

Table 14 reports 𝑞 = (𝛽YZ− 𝛽%Z) for all the independent variables and the result of testing 𝐻C:  𝑞 = (𝛽YZ−

𝛽%Z) = 0. . ***,**, and * denote the 1%, 5% and 10% levels, respectively; of the Chi2 distribution. ‘Effects’

represents the outcome of the Hausman test: the preferable specification for the regression on expected government support.

Variables Coefficients

Fixed Random Difference

Liabilities / GDP -0.0005677 -0.0005562 -0.0000115

Deficit / GDP 1.71065 1.714393 -0.0037429

Interbank / Liabilities -0.9721862 -0.8230343 -0.1491519 Debt / Liabilities -3.057152 -3.052044 -0.0051077

Hypothesis Hausman test

H0: Difference in

coefficients not systematic

Chi2 Prob. > Chi2 Effects

Referenties

GERELATEERDE DOCUMENTEN

H7: If a risk indifferent consumer expects energy prices to drop they will have a preference for (A) contracts with variable tariffs without contract duration and (B) low

The results confirmed the expected relation between the market value (measured using the market price to book ratio) and the credit rating, as well as relations between the CR

There are many MDD tools available, but they either lack meta-model support to struc- ture their models, proper modelling support to interact with the hardware of

This thesis shows that (a) when looking at the different types of power, spaces, levels and forms, there is an important distinction between formal and

We conducted a first study with the Giraff robot (http://www.giraff.org/) either administering a psychological treatment directly (the robot was presented as a coach with the

For this purpose we conducted a survey to find the general perception of people about crime and its possible causes especially to check the reliability and significance of

There are good evidence-based guidelines for platelet transfusion thresholds in different situations, but the limitations must be taken into account, namely that data are

Bo en behalwe die goeie gesindheid wat dit van die gemeente kan uitlok, kan die orrelis ‘n positiewe bydrae maak tot die kwaliteit van musiek en keuse van liedere wat in