EBA Stress Test Methodology - Key changes

Differences in approaches to

the 2014 exercise by risk type

Malta, January 2016

EBA/ECB Stress test 2016

Content

1. General information on EBA Stress Test

2. Methodological changes

a) FOCUS 1: Migration Matrices

b) FOCUS 2: Stressed LGDpit

c) FOCUS 3: FX Lending

d) FOCUS 4: Conduct Risk and OpRisk

e) FOCUS 5: Net Interest Income

f) FOCUS 6: Other Changes

3. Technical impact (Templates)

4. Lessons Learned from 2014

Contents

© 2016 Deloitte 2

General information on EBA Stress Test

After the global financial crises the use of stress

tests has become a common supervisory tool

further stress tests executed by the FSA or rather BoE/PRA (UK) plus the IMF in Japan

2010 EU

CEBS – Capital

Adequancy EU Stress

Test

2009 USA

FED – Supervisory

Capital Assessment

Program (SCAP)

2009 EU

CEBS – EU-wide stress

testing exercise

2011 USA

FED – Comprehensive

Capital Analysis and

Review (CCAR)

2011 EU

EBA – Solvency Stress

Test

2014 USA



Review (CCAR)

2012 USA



Review (CCAR)

2013 USA



Review (CCAR)

2014 EU

EBA – Stress Test (Euro

Area Compre-hensive

Assessment)

2016 EU

EBA Stresstesting

exercise is announced

for 2016

© 2016 Deloitte 3

Europe (EBA banks) SSM/European Monetary Union (ECB banks)


EBA and ECB organise the 2016 stress-testing

exercise for 53 banks in the European Union

53 EBA banks

EBA banks to submit

mandatory templates

Other SSM Banks

Other banks under the

SSM are subject to ECB

scrutinty, but details yet

unpublished

Goal SREP Input

In contrast to 2014, no

hurdle rates are specified

but stress test will inform

SREP.

2 scenarios

Each bank has to submit

data from stress-testing

calculations for 2016-18

© 2016 Deloitte 4

General EBA stress test approach

General information on EBA stress test

© 2016 Deloitte

Background and Objectives

Coordination and scope

• Investigating the resilience of the capital position of European banks against adverse macro-economic developments

• Conducting an assessment of the systemic risk within the EU financial system

• Comparing and contrasting EU-banks under adverse market decision

• Informing the SREP exercise with regard to the sufficiency of the capital position of individual banks

• All significant SSM supervised institutions (assets of more than €30bn) are in scope

• All other significant institutions will be asked by the ECB to conduct their own stress test, consistent with the EBA methodology, in

order to inform their SREP submission

Methodological focus

• The following risks are in scope: Credit Risk, Market Risk, Counterparty Credit Risk, CVA Risk, Operational Risk, Net Interest

Income/Expenses, Non Interest Income

• Compared to the 2014 exercise, there is an increased focus on conduct risk and FX risk, as well as methodological enhancements

particularly to the calculation of credit risk and interest risk income


6 © 2016 Deloitte

The EBA stress test 2016 follows a tighter

schedule than the EBA stress test 2014

Timeline EBA stress test 2014

Project kick-off

and on-site set-

up

Submission of

templates

Transparency

Exercise

Calculation

and

submission of

results

Finalisation of

quality

assurance by

end of June

Launch of EU-wide

stress tests and

publication of final

methodology and

templates as well as

scenarios

Outcome of stress

test including banks’

individual results

September October November December January February March April May June July

2015 2016

2016 EU-wide stress-test 2015 EU-wide

Transparency Exercise

EBA stress-test

2015/16

10-11/2015 02/2016 03-04/2016 06/2016 07/2016 12/2015

August

Expedited

publication

designed to align

the finalisation

with the cycle of

the annual SREP

08-09/2016

Announce-

ment of new round of

stress tests

Discussion with banks on

preliminary draft

methodology and

templates

EBA publishes

details of stress

test scenarios

EBA/ECB

receive results

data from banks

End of three

consultation and

Q&A periods

January February March April May June July August September October November

2014

NCAs receive

preliminary

results

Publication of

results three

weeks prior to

start of SSM

01-02/2014 03/2014 04-05/2014 06-07/2014 07-08/2014 09-10/2014 11/2014

Timeline EBA stress test 2016

Duration from launch to

publication of results

2016: about 5 month

2014: about 8 month

Duration


• The 2016 EBA stress test is an integrated

stress test. It forecasts all relevant P&L items as

well as the capital position to re-calculate

performance metrics.

• In order to forecast P&L, there has to be a

forecast of income and costs including losses

from various risk types.

• Likewise, the capital requirements for various

risk types needs to be forecast to calculate

capital ratios.

Elements new to stress test 2016

Elements already present in 2014 stress test

Risk elements covered by EBA Stress Test 2016

Risk elements

Market Risk incl. CVA

and FX Risk

Non Interest Income

Conduct Risk

Operational Risk

Net Interest Income

Credit Risk incl. CCR

and FX Risk

© 2016 Deloitte 7

Integrated Stress Test

Joint influence on RWA and P&L figures


Perfor-mance metrics

Leverage ratio

Total capital ratio

Tier 1 capital ratio

CET 1 ratio

Four performance metrics replace CET1 hurdle

rates while focus remains on capital position

Common Equity Tier 1 Total Risk

Exposure

CET 1 capital Total Risk

Exposure

Total capital (Tier 1 + Tier 2) Total Risk

Exposure

4,5% (Art. 92 1 CRR)

CET 1 capital ratio

6% (Art. 92 1 CRR)

Tier 1 capital ratio

8% (Art. 92 1 CRR)

Total capital ratio

3% (§ 49 to 96 of

Basel III framework)

Leverage ratio Total

Exposure Tier 1 capital

8

new surveyed key elements

previous key elements

• The 2016 stress test does not set hurdle rates as the 5.5%/8.0% in the EBA stress test 2014.

• As the 2016 exercise informs each bank’s SREP process, the assessment of the above performance

metrics will take place in the light of the bank’s capital requirements incl. transitional adjusments.

Performance metrics inform SREP

© 2016 Deloitte

Contents

9

Content











© 2016 Deloitte

Credit Risk

New requirements: Migration Contribution PDpit

Minimum criteria for

migration matrices/PDpit

How does the bank

create stressed migration

matrices?

Probability of moving from

one grade to another has to

be adjusted according to

the scenario

The PDpit for each grade

is adjusted appropriately to

reflect the scenario

What is the relationship

between grades and

default rates over the

economic cycle?

10

Definition

• Migration Contribution PDpit is the difference between the exposure-weighted PD average of non-defaulted Stock pre

migration versus post migration under the same scenario.

© 2016 Deloitte

Credit Risk

PDpit estimation on grade level

Parameters on portfolio level are obtained as weighted average of the respective

buckets

Calculation of point-in-time migration matrices is needed

Methodology for stressed point-in-time migration

matrices?

PDpit estimation on portfolio level

Explanation of how grade migrations are considered is

must be given

Some assumptions on the relationship between grades

and default rates may have to be formulated

1 2

Options for PDpit calculation

11

It is obvious that there is an increased focus on the use of migration matrices, although the

methodology stops short of requiring their use

No formal requirement to calculate migration matrices

© 2016 Deloitte

Credit Risk

Cyclicality of rating approaches as determining

factor The size of the migration contribution will depend on the rating approach

Reactive Rating Approach Stable Rating Approach

• Assignment of grades moves with

economic cycle

• Many Migrations

• Relatively stable relationship

between grades and PDs

• Assignment of grades stable across

economic cycle

• Few migrations

• Strong shift in default rate/PDpit for

a given grade

High migration contribution Low migration contribution

Default rate/migration matrix

A high migration contribution is indicative of a Point-in-Time rating approach, although Point-

in-Time PDs can also be converted to stable ratings

12 © 2016 Deloitte

Credit Risk

Methods for creating stressed PIT migration

matrices

Internal ratings

External

ratings (TTC)

A B C D

A 97% 2% 1% 0%

B 2% 91% 4% 3%

C 0% 1% 84% 15%

D 0% 0% 0% 0%

Mathematical

and statistical

methods

Migration matrix

Defa

ult

rate

Dependent on macroeconomic factors for the givens scenarios

(CPI, GDP) the stressed PDpit has to be determined

macroeconomic

factors

Aggregate

stressed-

PDpit

Parameter

models

PDpit=f(rate, oil)

Satellite

models

Oil=h(GDP,CPI)

Determination of stressed migration matrix that produces

PDpit in line with macro-economic projection

• Example: PDs are

converted into distance to

default (DD). Changes in

DD are associated with

upgrades / downgrades.

Migration matrices are

stressed by assuming a

uniform change in DD.

Internal data

Adjustments

for cyclicality

First Step Second Step

13 © 2016 Deloitte

Contents

14

Content











© 2016 Deloitte

Credit Risk

New guidance for stressed LGDpit calculation

splits defaulted assets into new and old defaults

LGDpit

LGDpit_new LGDpit_old

Weighted by

new defaulted

Stock

Weighted by

old defaulted

Stock

Cure rate

new

Cure rate

old

Consideration of LGD

grade migration

Consideration of migration

between default

categories

15

Calculating stressed LGDpit

• LGDpit are split into LGDpit

new for assets defaulted

within the year and LGDpit

old for LGDpit of assets

which defaulted previously

• The respective default stock

is used for weighting

• Cure rates have to be

calculated for each LGDpit

bucket separately

• Different levels of LGDs for

previously and newly

defaulted assets have to be

considered

© 2016 Deloitte

Credit Risk

Two approaches for creating LGD grades refer

to either loan-to-value grades or LGD bands LTV grade approach LGD band approach

LGD band Exposure

% (t0)

0-10% 25%

11%-20% 15%

21%-30% 5%

31%-40% 15%

… …

81%-90% 10%

91%-100% 5%

Exposure

% (t1)

20%

10%

2%

10%

…

20%

15%

LTV grade Exposure

% (t0)

50% 35%

51%-60% 25%

61%-70% 15%

71%-80% 10%

… …

100%+secured 5%

125% 4%

Exposure

% (t1)

20%

10%

2%

10%

…

10%

10%

LGD grades are defined via LTV percentage. Stressing

collateral values leads to grade migrations. Assumes

knowledge of loss rates by LTV bucket.

16

LGD grades are simply a banded version of LGD.

Migrations are based on capability to stress LGD,

which also yields stressed LGD by LGD grade.

© 2016 Deloitte

Credit Risk

Modelling approach for cure rates – assessing

cure rate old

Non-performing Grade

i.e. 90dpd

Recovery &

Resolution

No migration

Cure (Return

to satisfactory)

90 dpd Cure Rec. &

Res.

90 dpd 30% 40% 30%

Cure 5% 93% 2%

Rec. &

Res.

0% 30% 70%

Migration between default categories Migration matrix for default grade migration

30% 40%

30%

The green shaded cells drive

cure rate old

17 © 2016 Deloitte

Credit Risk

A practical approach for stressed cure rates lies

in stressing migration between default states

A B D

90dpd Cure R&R

A 97% 2% 1% 0% 0% 0%

B 2% 91% 3% 1% 1% 1%

C 0% 1% 83% 4% 4% 4%

D

90dpd

0%

30% 40% 30%

Cure 5% 93% 2%

R&R 0% 30% 70%

18

• The migration matrix can be extended for the different default grades.

• Stressing the migration matrix automatically leads to stressed cure rates

Stressed cure rate approach

Rescaling default classes

• The light green cells need to

be re-scaled to calculate

cure rate new

• The dark green cells reflect

cure rate old

• Stressing the migration

matrix will lead to stressed

cure rates

• Defaulted assets are split

into three buckets

• Each bucket receive

individual stress

cure rate migration out of

default bucket

is not allowed

© 2016 Deloitte

Contents

19

Content











© 2016 Deloitte

FX Lending

20

Starting point and stressed credit risk

parameters for FX lending have to be provided

FX lending

Starting point credit risk parameter Stressed credit risk parameter

FX breakdown for EUR/USD/CHF

• Impairment loss - Old defaulted assets

• PDpit

• LGDpit_new/old

• Non-defaulted exposure - share of FX

lending Exp

All FX Lending exposure has to be considered

where the currency of the credit facility is different

from the local currency of the borrower

No FX breakdown

• Impairment loss - Old defaulted assets for FX

lending-all currencies

FX exposure that meets a specific threshold and

the currency of the credit facility is different from

the local currency of the borrower has to be

considered

© 2016 Deloitte

FX Lending

A decision tree helps in determining whether or

not exposure has to be reported as FX lending Meets threshold

on asset class

level (5%)?

Currency of

Exposure is exposed to Bank A

Bank A

Germany

EUR

Banks in Czech

Republic

(Czech Koruna)

Banks in USA

(USD)

Banks in Slovakia

(EUR)

EUR

USD

EUR

Exposure has to

be reported?

No

No

Yes

No No

Yes

Classification of FX lending

• FX lending items refer to counterparts for which FX risk is not visible from a currency mismatch in the

bank’s books, i.e. counterparts who face FX risk which may increase their PD

21 © 2016 Deloitte

Contents

22

Content











© 2016 Deloitte

Operational Risk has been considerably

extended compared to the 2014 implementation

Operational Risk

Operational Risk

Other Operational Risk Conduct Risk

Qualitative

Approach

Quantitative

Approach

OpRisk framework 2014

OpRisk framework 2016

23 © 2016 Deloitte

Conduct Risk is introduced as a new item in the

calculation of operational risk as of 2016

Operational Risk

Current or prospective risk of

losses to an institution arising from

an inappropriate supply of financial

services including cases of willful

or negligent misconduct

Conduct Risk

Applicable if the institution must

not apply the qualitative approach

Quantitative Approach

Applicable if

• Material conduct risk event from

2011-2015

• Competent Authority decision

Qualitative Approach

Applicable for non-material

conduct risk events for the 3 years

Floor

Losses arising from conduct

risk

24 © 2016 Deloitte

Reporting requirements for qualitative and

quantitative approach

Operational Risk

Conduct Risk

Institutions use their own methods to calculate

the impact on the P&L over the three year time

horizon.

Projections shall take into account scenario

dependent information if applicable

Quantitative Approach

Institutions have to report during the 3 years:

• Historical data on incurred losses

• Projections for non-material events

• Historical material events and estimated

losses for the scenarios

• Projections for new material events

Qualitative Approach

Average of historical losses over 5 years

(2011‐2015) for non‐material conduct risk

events only.

Stress multiplier applied for the adverse

scenario to the average.

Floor

25 © 2016 Deloitte

An approach to determine conduct losses may be

based on a stressed quantitative method

Operational Risk

Apply stress on loss

frequency modelling Determine stressed distribution

Additional stress on

key risk indicators

Combined total stressed distribution

Stressed

Conduct

Risk

Multiply

conduct loss

with a scaling

factor derived

by key risk

indicators

(e.g. ratio of

stressed and

unstressed

results)

Apply stress

parameters

on Key Risk

Indicators

Determine conduct loss

based on the combined

total distribution

Loss frequency modeling

stressed

Loss severity modeling

unstressed

26 © 2016 Deloitte

Projecting future conduct risk losses under the

qualitative approach

Operational Risk

Existing treatment of misconduct issue

Accounting provision already raised Accounting provision not yet raised, but a

significant settlement cost possible

High certainty over

eventual cost

Evidence sufficient to

determine range of

settlement outcomes

High uncertainty

over eventual cost

Projection estimate =

existing provision

Projection estimate >

existing provision

Evidence insufficient

to determine range of

settlement outcomes

Make multiple settlement

estimates and assign

probabilities to each one;

maximize confidence that

actual loss does not

exceed the loss estimate

Exercise expert judgment

to determine a settlement

estimate and attempt to

settle at or below it

27 © 2016 Deloitte

Treatment of other operational risk

Operational Risk

According to CRR but excluding

all conduct related losses

Other operational risk

Average of historical losses

over last 5 years (2011-2015)

Stress multiplier applied in

adverse scenario

Floor

Losses arising from conduct

risk

Usage of bank’s own methods

Projection of losses

Considering the 90th percentile

of aggregated amount of losses

Adverse scenario

Considering the 50th percentile

of aggregated amount of losses

Baseline scenario

28 © 2016 Deloitte

Contents

29

Content











© 2016 Deloitte

Key changes in determining Net Interest Income /

Expenses compared to 2014

Net Interest Income / Expenses

1 Differentiation between Reference Rate and Margin

for interest rates

2 Exposure is separated into existing, maturing and

new exposure

3 Discount unwinding can be included in interest

income

4 Introduction of caps and floors for the Margin

component

5 Additional reporting requirements for country /

currency pairs

30 © 2016 Deloitte

Initial state

2015 2017 2018

Volume (average

over 2015, in M

EUR)

Total amount

(in M EUR)

With original

maturity <1 (in

M EUR)

With original

maturity <2 (in

M EUR)

Total amount

(in M EUR)

Total amount

(in M EUR)

Margin

Ref Rate

EIR

component

2016

Maturity schedule of the total portfolio at the cut-off date, split by original

maturity

Calculation for each component is subject to

specific constrains


• Banks have to provide information about:

• Notional

• Residual maturity

• Original maturity

• Rounded maturity

• EBA will provide templates for automatic

calculation of future exposures

Step 1: Calculation of existing, maturing and new exposure

Exis

ting

Matu

ring

New

Exis

ting

Matu

ring

New

Exis

ting

Matu

ring

New

Margin

Ref Rate

EIR

component

2016 2017 2018

Total volumes (in M EUR)

31

Projections of interest income from

discount undwinding

Step 2: Calculation of income from discount unwinding

Init

ial

sta

te

Pro

jec

tio

n

Constrains under the adverse scenario

© 2016 Deloitte

+ +

Boundaries for projected margins and final

calculation of Net Interest Income


Step 3: Calculating boundaries of margin for repriced liabilities and assets

Floor on the margin paid on new liabilities

• Sovereign bond spreads (country of exposure) and

idiosyncratic risk of the bank

• Type of new liabilities

• Historical margin changes

Cap on the margin earned on new assets

• Sovereign bond spreads (country of exposure) for

financial corporations and derivative positons

• Historical margin changes

Step 4: Determining defaulted assets for each year

Net Interest Income

Net Interest Income

from future volumes

and margin / ref rate

projections

Income from discount

unwinding (according

to certain constrains)

Impact resulting from

defaulted positions =

32

Step 5: Impact on Net Interest Income

© 2016 Deloitte

Additional reporting requirements in case of large

amounts of foreign business


Compute for each country /

currency pair the maximum

between total assets and

liabilities

Rank the country / currency

couples according to their

volumes

Banks are requested to report the country / currency

breakdown, either:

1. Up to 90% coverage of sum of all country / currency

volumes or

2. Up to 15 country / currency couples

Domestic banks are not requested to report any country /

currency breakdown

33

Collecting relevant information Thresholds for additional reporting requirements

© 2016 Deloitte

Contents

34

Content











© 2016 Deloitte

The horizon for NTI projections in market risk is

extended to five years alongside other changes

Other Changes

• Additional to the focused changes stress methodology for other risk types have been also adjusted compared to 2014

• In general the Stress Test 2016 requires banks to deliver more granular data then before with stronger constrains on calculation

methods for the stress results

• The following changes with a higher workload or impact on banks key figures can be observed as of 2016:

Changes of methodology for other risks

• Information about economic hedges have to provided if they should be recognized

• The time series for calculating NTI values have increased from three to five years

• New approaches to calculate shocks for additional risk factors which are not provided by the EBA

have been introduced

• Banks using additional risk factors are requested to calculate and report a measure to quality

assure additional risk factors

• Both measure are floored according to the changes in the IRB portfolio.

• Introduction of a scaling factor for economic hedges

• Changes in calculation of initial NTI and the application of losses within the projected years

• Banks using a comprehensive approach have to include a VaR based portfolio specific scaling

factor

Ne

t T

rad

ing

Inco

me

Ad

dit

ion

al

ris

k f

acto

rs

CV

A a

nd

IRC

calc

ula

tio

n

Additional

data

requirements

Changes in the

methodology Market Risk

35 © 2016 Deloitte

CVA shocks are included while NII includes

projections and five years of historical data

Other Changes

• All banks have to use a generalized approach to calculate the income from dividends and fees

and commissions

• Administrative expenses, profit or loss from discontinued operations and other operating

expenses is floored at the 2015 value but can be adjusted due to one-off effects which have to

be permitted by the competent authorities

• Dividend payments have to be calculated with a uniform methodology if no dividend policy

exist

• Collateral is shocked according to the scenario

• CVA hedges are recognized but no adjustments are possible Cre

dit

Va

lua

tio

n

Ad

justm

en

t

No

n I

nte

res

t In

co

me

Additional

data

requirements

Changes in the

methodology CVA

Non Interest Income

• Banks have to provide five years of historical data together with their projections

Additional

data

requirements

Changes in the

methodology

36 © 2016 Deloitte

The 2014 logic of credit risk is unchanged with

tweaks to impairment losses and segmentation

Other Changes

Credit Risk

Technical effort: Impact on results:

A separate category is newly introduced for the impairment flow of sovereign exposure

Risk exposure amount for counterparty credit risk is kept constant for credit purposes and only adjusted for PD and LGD

purposes

Impairment losses on new defaulted assets:

• The calculation of the new defined “Gross Imp flow new“ is based on LGDpit new. “Net imp flow old” is the same as in

2014.

Impairment losses on old defaulted assets:

• The new EBA-methodology does not contain a DPC factor, the estimation of the old defaulted assets is based on the

“Def Stock“ instead.

Starting point-in-time-parameters:

• For bank with non-approved internal models, which are used regularly by the bank for internal purposes, the competent

authority has to be satisfied with the using them for the purpose of the EU-wide stress test.

• If no internal model is in place banks ere expected to approximate PDpit and LGDpit via Default and Loss rates.

Therefore only adjustments that results in a more conservative starting point s are permitted

37 © 2016 Deloitte

Exclude exposures not

within the scope of the

largest counterparty default

(e.g. central government,

central banks)

Apply stress factors defined in

the market risk scenario to all

traded positions for macro-

economic adverse scenario,

Subprime Crisis, and European

Sovereign Crisis.

Assuming any collateral to

be called beyond what is

currently held considering

only positive exposures

Considering the change in

the mark‐to‐market

exposure to the

counterparties as well as the

revaluation of the collateral

In each scenario only the top

10 counterparties are

considered according to

their stressed exposure

Stressed current exposure

multiplied by the respective

stressed LGD netting the

CVA impact on the P&L

before application of stress

Summing up the impact of

the two most vulnerable

counterparties that default

for each particular scenario

The final P&L impact will be

the maximum between the

overall impact across the

three scenarios

A new approach for calculation of counterparty

credit risk is introduced

Other Changes

Determine relevant

exposure

Apply stress factors on

the exposure

Calculate stressed

current exposure

Ranking of counterparties

according to stressed

current exposure

Consider the top 10

counterparties

Calculate the impact of

the default

Determine the overall

impact for each scenario

Calculate the final P&L

impact

In addition to the Counterparty Credit Risk exposure banks are asked to calculate losses from the jump‐to‐default (net profit / loss

resulting from an issuer’s instantaneous default) of the direct credit exposure to this counterparty in all accounting portfolios.

(Indirect exposures to the issuer (i.e. CDS) should be included, as this corresponds to the default of the reference entity)

Counterparty credit risk methodology

1 2

3

5

4

7 8 6

38 © 2016 Deloitte

Contents

39

Content











© 2016 Deloitte

Credit risk

Parameters are reported at higher granularity

and PiT starting values have to be provided

Release of provisions

from newly def. assets

Gross Impairment loss –

New defaulted assets

LGD PiT new/old EUR (%)

LGD PiT new/old USD (%)

LGD PiT new/old CHF (%) Cure rate – new/old

defaulted assets (%)

LGD PiT new/old –

provision coverage (%)

Cures in 2015 from

new/old defaults in 2015

LGD PiT new/old 2015

Scenario (%)

PD PiT EUR (%)

PD PiT USD (%)

PD PiT CHF (%)

Due to assumed changes in LGL

Due to assumed re-defaults on cures

Impairment loss – Old

defaulted assets

PD PiT

Non-defaulted exposure

– end of year Share of FX lending Exp (%)

Share of EUR Exp (%)

Share of USD Exp (%)

Share of CHF Exp (%)

• Stress test 2016 involves

several changes in

reporting:

− Asset classes are more

granular

− No additional templates

any more - all templates

have to be filled

mandatorily by all banks

− Starting parameters

(i.e. PD, LGD) have to

be provided for different

currencies (EUR, USD,

CHF)

− Some calculation

formulas are provided in

the templates

Comparison of EBA Stress

Test 2014 and 2016

Migration Contribution

PD PiT (ppt)

Implied stress factor

from LGD PiT (%)

Imp Flow FX

Net Impairment from EUR

Net Impairment from USD

Net Impairment from CHF

• Light Blue items are just new for starting parameters

• Green items are just new for the evolutions

• Dark blue items are new in general

40 © 2016 Deloitte

Technical

IRB asset classes STA asset classes

The STA asset classes must not be mapped anymore to the IRB asset classes; the new categorization follows the CRR classification in Articles

112 and 147.Moreover, several asset classes which were optional in 2014 are mandatory in 2016 ( marked fat blue) and there are also asset

classes in 2016 which were not considered in 2014 at all (Marked fat black).

Credit Risk has no direct mapping between IRB

asset classes and STA ones any more

41

Central banks and central governments

Institutions

Corporates

Specialised Lending

Real Estate Related, Non Real Estate Related

SME


Other


Retail

Secured on real estate property

SME

(OPTIONAL) of which: subject to SME-supporting factor

Non SME

(OPTIONAL) of which: Owner Occupier

(OPTIONAL) of which: Buy to let

(OPTIONAL) of which: Other secured by real estate

Qualifying Revolving

Other

SME


Non SME

Equity, Securitisation

Other non-credit obligation assets

Central governments or central banks

Regional governments or local authorities

Public sector entities, Multilateral Development Banks

International Organisations

Institutions

Corporates

SME


Non SME

Retail

SME


Non SME

Secured by mortgages on immovable property

SME

Non SME

(OPTIONAL) of which: Owner Occupier

(OPTIONAL) of which: Buy to let

(OPTIONAL) of which: Other secured by real estate

Items associated with particularly high risk

Covered bonds

Claims on institutions and corporates with a ST credit assessment

Collective investments undertakings (CIU)

Equity, Securitisation, Other exposures

© 2016 Deloitte

Technical

42

Market Risk, Counterparty Credit Risk and CVA

Detailed Information for

postions in AFS and FVO

Reporting of

counterparties

Provide additional information about loss / gain stemming from delta, gamma or vega effects via

sensitivity analysis

Reporting of 10 largest counterparties for the adverse and both historical scenarios according to the

described calculation approach. Additional information about total loss and jump to default loss

Banks are requested to give detailed information about non-sovereign AFS positions within hedge

accounting and FVO and AFS positions except hedge accounting with respect to interest rates and a

breakdown on gains and losses for the adverse scenario

Explanation of loss / gain

Additional reporting requirements for Market

Risk and Net Interest Income (1/2)

Ref Rate and Margin

exposure

Reporting of exposures split into margin and reference rate exposure for future exposure calculation.

Additional information about projections of margin and reference rate necessary

Net Interest Income

© 2016 Deloitte

Technical

Operational Risk

Risk exposure

Loss recovery

Banks have to report for both conduct and other operational risk the number of risk categories

and the occurred losses in different loss intervalls

For both conduct and other operational risk loss recovery should be reported

Banks are requested to include the risk exposure for the last 5 years (2011-2015) and the projections

for the 3 years (2016-2018) covered in the exercise

Loss events

Additional reporting requirements for

Operational Risk (2/2)

Material conduct risk

losses

For conduct risk banks are additionally requested to report material conduct risk losses with respect to

the mapping of article 324 CRR

43 © 2016 Deloitte

Contents

44

Content











© 2016 Deloitte

Deloitte’s clients’ findings from the 2014 EBA

stress test concern data issues in three areas

Stresstesting lessons learned

Data storage and

data

reconciliation

System

rationalisation

Definition of key

processes with

counterparties

Up-to-date data,

segmentation and

data definition

Functional

documentation

Internal

processes and

coordination

Data governance

and coherence

Standardised

stresstest models

Modes of

communication

Template Population

A dedicated Template

Population Tool

addresses all three

identified areas and

supports the mitigation

of problems in five out

of nine fields

Data

qu

ali

ty

IT s

ys

tem

s

Pro

ce

sse

s

45

In scope of tooling Out of scope issue

© 2016 Deloitte

Data and calculation were identified as main

sources of additional effort


Data availability

Data preparation

Scenario calculation

• Large gap between data requirements and data availability

• Differences exist in terms of segmentation, point-in-time availability,

identification and mapping, data sources, and formatting

• Definition of generic databases for various stress-testing applications

• Preparation of data is a time-consuming process

• Lag until first complete calculation of results is too long leaving too little

time for data validation and consistency checks

• Regular updating of generic database and cross-checking of data

• The implementation of new methodologies is a laborious process, as is

the identification of levers with significant impact on the final result

• Calculation on unique database does not permit validation and

comparison thereby preventing a „grasp“ on the data

• Regular calculation of stress-tests using generic scenarios

46 © 2016 Deloitte

Preparation of templates and communication

within the bank were regarded as non-standard


Template preparation

External

Communication

• Poor template quality and lack of clear data derivations

• Mapping data and final output requires exact definitions and is prone to

errors when updates to templates and data specifications occur

• Use of a universal data tool allowing for versioning and versatility

• Communication with competent authorities is time-consuming

• Preparation of messages and and getting C-level approval under time

pressure clashes with last-minute updates and data checks

• Creation of secured data spaces with restricted access to results;

ensuring high data quality and leaving more time for discussions with

senior management

47 © 2016 Deloitte

Feedback from Deloitte Stress Testing Survey

2015 sees further scope for centralisation


48

Deloitte Stress-Test Survey 2015: Firm-Wide Stress tests remain challenge particularly for smaller banks

© 2016 Deloitte

Our Banking Team

Dimitrios Goranitis

Banking Leader

Tel: +356 9909 7707

Email: [email protected]

Location: Malta

© 2015. For information, contact Deloitte Malta. 50

Background:

Dimitrios joined Deloitte Financial Advisory Services during 2012. Prior

to Deloitte, Dimitrios was a manager in Transaction Advisory Services

with Ernst and Young (Greece and Eastern Europe), an assistant

director of M&A with Altium Capital Group (UK HQ, Eastern Europe

desk), a senior manager with UBS Financial Services (New York) and

VP of Business Strategy with Bear Stearns - JP Morgan (New York).

Dimitrios has gained substantial AQR, stress test and NPL advisory

experience by leading AQR and stress test engagements in Greece,

Cyprus and Eastern Europe, and participating in similar exercises in

Spain and Portugal.

Dimitrios has reviewed the models and methodologies for Greek banks

and he has been involved in various Basel II,III gap analysis and Basel

II,III implementation projects (e.g. Pillar II) for banks in Greece and

CSE.

He has also participated as project manager in various international

projects related to advanced risk management and numerous Recovery

and Resolution Plans for Banks located in Malta, Greece and Cyprus.

Relevant experience:

• AQR, Stress Test and Recovery Plan PMO and Quality

Assurance support to Bulgaria National Bank - Full scope PMO

and QA of the country wide project in relation to 22 licensed banks in

the country (in progress).

• SREP Governance transformation of a systemic bank: A

governance and entity level controls transformation exercise of a

domestically systemic bank, to assist the bank with their compliance

with the regulators’ new SREP supervisory methodology.

• Recovery Plan according to BRRD for a systemic lender and a

less significant institution in Malta. Included several stress test

scenarios and an analysis of a number of capital and liquidity

specific issues.

• AQR on Mediterranean Bank plc – Full scope AQR in 2015 on

behalf of ECB. The process included QA on risk management

practices of IAS 39 rules.

• AQR PMO and QA support to MFSA - Full scope PMO and QA of

the country wide project in relation to three systemic banks in the

country. The process included QA on risk management practices of

IAS 39 rules.

• AQR and stress test of five tier 1 banks in Greece - Full scope

AQR of five tier one banks in Greece on behalf of Blackrock Advisory

and European Central Bank. The review extended to their

international holdings in six countries in Eastern Europe,

Switzerland, UK, and Turkey. The process included the quality

assessment of the internal credit rating models using Basel II’s

internal-ratings-based approach.

• Due Diligence of two tier 1 banks in Greece - Full scope financial

due diligence of two tier 1 banks in Greece on behalf of European

Stability Fund (ESF) and Hellenic Financial Stability Fund (HFSF) in

advance of their recapitalization. The process included the review of

models and advanced risk management techniques under

operational, market, credit risk and other risk.

Deloitte Malta banking credentials

Mark Micallef

Senior Manager – Banking Team

Tel: +356 7929 8274


Location: Malta


Background:

Mark joined Deloitte Malta in 2004 and has extensive experience in

audit, regulatory and compliance, and risk and controls assignments of

firms within the financial services industry, in particular banks and

investment services providers. Mark’s area of specialisation is banking

regulation and he has managed a local engagement to support the

Maltese Financial Services Authority and ECB in the Comprehensive

Assessment of the three systemic banks in Malta. Mark has also carried

out statutory audits of numerous banks at Deloitte Malta and Deloitte

London, including the audit of a domestically systemic Maltese bank.

He was recently seconded to one of the smaller banks in Malta where

he was responsible for the design and implementation of the finance

and regulatory reporting, and client on-boarding processes.


• AQR and Stress Test support to Malta Financial Services

Authority/ECB - Full scope PMO and Quality Assurance of the

country wide project in relation to three systemic banks in the

country.

• AQR on Mediterranean Bank plc – PMO for a full scope AQR in

2015 on behalf of ECB. We were engaged to carry out the role of

NCA Bank Team.

• AQR PMO support to BNB – Currently engaged to support the

Bulgarian National Bank with the first round of AQR, stress testing,

and recovery planning exercises in Bulgaria.

• SREP Governance transformation of a systemic bank: A

governance and entity level controls transformation exercise of a

domestically systemic bank, to assist the bank with their compliance

with the regulators’ new SREP supervisory methodology.

• Secondment to a non-systemic Maltese bank - Designing and

implementing the finance and regulatory reporting, and client on-

boarding processes of the bank including the design and testing

phases of the core banking system.

• Auditor of Bank of Valletta plc. Mark was a key part of the audit

team of Bank of Valletta plc for a number of years, focusing on the

audit of the credit portfolio, financial markets section, and regulatory

disclosures.

• Secondment to the B&S audit department at Deloitte London.

Mark worked for a number of months in the Banking and Securities

audit department in London where he was involved on the audit of

Abbey National Bank

• .


Andreas Karameros

Manager – Banking team

Tel: +356 2343 2335


Location: Malta


Background:

Andreas joined Deloitte Financial Advisory in 2015. Prior to Deloitte,

Andreas was a manager in Transaction Advisory Services with Ernst

and Young (Greece, Southeast Europe and Germany), and an assistant

manager in Project & Structured Finance with Piraeus Bank Group

(leading bank in Greece).

Andreas has been specializing in Financial Modeling (including the

development of credit and operational risk tools) and Valuations for

more than 7 years and has developed industry expertise in Financial

Services, Energy and Infrastructure by leading a number of international

assignments.






• Governance, controls, and transformation of the ICAAP and

ILAAP processes of a systemic bank into advanced risk

management tools, including how they feed into the risk

management requirements and capital calculations.

• Financial modelling and project financing on behalf of Piraeus

Bank Group, in relation to the financing, construction and operation

various motorways, gas-fired installations and RES projects.

• Risk management framework set-up and assessment of

Operational and market risk, on behalf of Piraeus Bank Group in

Greece.

• Valuation of derivatives and structured products for various

funds in Malta.

• Collateral valuation of all systemic banks in Greece of SME and

SBP loan portfolio of four systemic Greek Banks on behalf of

Blackrock Advisory and European Central Bank.

• Development of a Credit Risk model, in the context of transfer

pricing documentation prepared for a major bank in Greece.

• Development of Risk-based pricing and Credit models for a

major bank in Greece.

• Collateral valuation of the Bank of Cyprus on behalf of Central

Bank of Cyprus.


Nick Lavdis

Manager – Banking Team

Tel: +356 7902 2412


Location: Malta


Background:

Nick joined Deloitte Financial Advisory in 2015. Prior to Deloitte, Nick

was an assistant manager in Transaction Advisory Services with Ernst

and Young (Greece and South-Eastern Europe), a senior financial

analyst with Johnson & Johnson (EMEA) and a loan officer with Alpha

bank (Greece). Nick has gained substantial AQR, stress testing and risk

management experience by leading AQR and financial risk related

engagements in Greece and Malta and participating in similar exercises

in Cyprus.

Nick supported several systemic banks in Greece and abroad (i.e.

Malta, Cyprus, Serbia) on various risk management engagements and

the development of their Recovery and Resolution Plans in line with

relative regulatory requirements (‘BRRD’).

Nick holds a Masters Degree in Applied Economics and Finance from

Athens University of Economics and Business and a Bachelor in

Economics from Aristotle University of Thessaloniki.






• Governance, controls, and transformation of the ICAAP and

ILAAP processes of a systemic bank into advanced risk

management tools, including how they feed into the risk

management requirements.

• AQR on Mediterranean Bank plc – Full scope AQR in 2015 on

behalf of ECB. The process included QA on risk management

practices of IAS 39 rules practices.

• AQR PMO and QA support to Bank of Greece - Full scope PMO

and QA support to Bank of Greece in relation to 2014 AQR of all

systemic banks in Greece. The process included QA on risk

management practices of IAS 39 rules.

• Recovery Plan according to BRRD for a systemic lender and a

less significant institution in Malta. The process included several

stress test scenarios, an analysis of a number of capital and liquidity

specific issues. The detailed critical review of the operational,

market, credit, reputational and other risk was also part of the

process.

• AQR of all systemic banks in Greece - Quality assessment of

Corporate, SME and SBP loan portfolio of four systemic Greek

Banks on behalf of Blackrock Advisory and European Central Bank.

The process included the quality assessment of the Bank’s internal

credit rating models using Basel II’s internal-ratings-based approach.

• Due diligence of two systemic banks in Greece - Full scope

financial due diligence of two systemic banks in Greece on behalf of

European Stability Fund (ESF) and Hellenic Financial Stability Fund

(HFSF) in advance of their recapitalisation. The process included the

review of Bank’s internal models and advanced risk management

techniques under operational, market, credit risk and other risk.


Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”), its network of member firms, and their

related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as “Deloitte Global”) does not provide

services to clients. Please see www.deloitte.com/de/UeberUns for a more detailed description of DTTL and its member firms.

Deloitte provides audit, tax, consulting and financial advisory services to public and private clients spanning multiple industries; legal advisory services in Germany

are provided by Deloitte Legal. With a globally connected network of member firms in more than 150 countries and territories, Deloitte brings world-class capabilities

and high-quality service to clients, delivering the insights they need to address their most complex business challenges. Deloitte’s more than 210,000 professionals

are committed to making an impact that matters.

This communication contains general information only not suitable for addressing the particular circumstances of any individual case and is not intended to be used

as a basis for commercial decisions or decisions of any other kind. None of Deloitte & Touche GmbH Wirtschaftsprüfungsgesellschaft or Deloitte Touche Tohmatsu

Limited, its member firms, or their related entities (collectively, the “Deloitte network”) is, by means of this communication, rendering professional advice or services.

No entity in the Deloitte network shall be responsible for any loss whatsoever sustained by any person who relies on this communication.

© 2015 Deloitte & Touche GmbH Wirtschaftsprüfungsgesellschaft

Documents

EBA Stress Test Methodology - Key changes