Asset Quality Review and Stress Test: Recent
Experiences and Potential Implications for the
MENA Region
Dubai, 7th May 2014

© Oliver Wyman LON-FSP22401-197

Agenda

1. AQR and stress test: setting a new standard for banking supervision
2. Applying the new methodologies to the prevailing context in MENA region: potential scenarios…. 3. … And implications for MENA Banks
4. Concluding remarks

© Oliver Wyman LON-FSP22401-197

1

1 AQR and stress test: setting a new standard for banking supervision

Since the start of the Eurozone crisis a number of AQRs and stress tests have been carried out in Europe with relevant impact on the Banks
Greece – ’11

Ireland – ‘10
Spain – ‘12
• Economy: ~2% GDP EU
• Asset Quality
Review
• Credit Loss Projections 
• Loss Absorption

Capacity

•
•
•
•


Capital shortfall ~€24mld

Economy: ~12% GDP
EU
Asset Quality

Review
Credit Loss Projections

Loss Absorption
Capacity

•
•
•
•

Economy: ~2% GDP EU
Asset Quality
Review

Credit Loss Projections

Loss Absorption Capacity

Capital shortfall ~€50mld





Capital shortfall
~€60mld

Cyprus – ’12
•
•

Portugal – ’11
•
•
•
•

•

Economy: ~2% GDP EU
Asset Quality
Review

Credit Loss Projections

Loss Absorption Capacity

Capital shortfall ~€7mld

•

Economy: ~0.2% GDP EU
Asset Quality

Review
Credit Loss

Projections
Loss Absorption Capacity 

Capital shortfall ~€6mld


Slovenia – ’13
•

Economy: ~0.4 % GDP
EU

Capital shortfall ~€4.8mld

© Oliver Wyman LON-FSP22401-197

3

A new, Eurozone-wide Asset Quality Review and stress test has recently been undertaken by ECB
European Baking Union outlook

Three Pillars of the European Banking
Union and the Comprehensive Assessment

• Pillar 1: ECB will keep direct responsibility for the largest banks, but will have power to deal with small banksi f necessary (SSM)

Iceland
Sweden

Finland

• Pillar 2: Common resolution fund & mechanisms for failing banks

Norway

• Pillar 3: Common deposit- insurance schemet o prevent bank runs

Estonia
Latvia
Denmark

Lithuania
Belarus

Ireland

Uk

Neth.
Belgiu
m
Lux.

Czech

Ukraine
Slovakia

France

Austria
Switz.
Italy

Portugal

Poland

Germany

Spain

Moldova
Hungary
Sloveni
Romania
Croatia a Bosnia
Serbia
Bulgaria
Montenegro
Macedon ia Albania
Greece

Eurozone

EU members not using Euro

© Oliver Wyman LON-FSP22401-197

Non EU members

Georgia
Armenia
Turkey

• ECB - in close cooperation with the
National Competent Authorities (NCAs)
– started a “Comprehensive
Assessment (CA)” of the 128 most relevant European Banks , covering around 85% of total assets of the banking system as a preparatory phase of the SSM set up

EU Candidate countries

4

Comprehensive Assessment is made up of 3 elements, the Asset Quality
Review, is the one currently ongoing
Comprehensive assessment

1
2
3

Risk Assessment RAS)
• Supervisory judgements on key risk factors, such as liquidity, leverage and funding
• Quantitative and qualitative analysis

Asset Quality Review (AQR)
• Assessment of data quality, asset valuations, classifications of non-performing exposures, collateral valuation and provisions
• Covering credit and market exposures, following a risk-based, targeted approach

Goals of the Comprehensive Assessment
• Transparency - enhancing the quality of information available on the condition of banks • Repair - identifying and implementing any necessary corrective actions
• Confidence building – assuring all stakeholders that banks are fundamentally sound and trustworthy

Stress Test (ST)
• Forward-looking view of banks’ shockabsorption capacity under stress
• Conducted in collaboration with the European
Banking Authority

© Oliver Wyman LON-FSP22401-197

5

The exercise has started in October 2013 and will be completed by
November 2014, when there will be the transfer of power to the SSM
Illustrative timeline

2013

Comprehensive Assessment

Nov

2014
Dec

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sept

Oct

Nov

Dec

Supervisory
Risk
Assessment
Asset Quality
Review

Stress Test

Start SSM

Data collection

© Oliver Wyman LON-FSP22401-197

Live activities

On-site exercise

End of the Comprehensive Assessment exercise, publication of results and start of corrective actions

6

Asset Quality Review (AQR) and Stress Test (ST) parts of the exercises follow different, but complementary objectives
Asset Quality Review (AQR)
• Definition: an asset quality review considers the key attributes of the different asset portfolios of the bank to evaluate the “quality” of a bank’s assets
– Loan tape provides an overall characterisation of each portfolio – Individual file review allows to get an on-the-ground understanding of the real portfolio quality
– Supplemented by independent real estate appraisals of top and random real estate collaterals by specialist firms
• Objective: for each portfolio the objective is to

Stress Testing (ST)
• Definition: a stress test quantifies and projects future losses under a baseline and an adverse macroeconomic scenario
– Bottom-up stress tests draw on detailed data at the individual bank level
- Individual calculations designed in detail and subsequently aggregated to system-level results
- Generally more accurate, precise and reliable since they use individualised, loan-by-loan data
– Top-down stress tests draw on system-level data
(some breakdown at bank-level but not in-depth)

– Understand portfolio asset mix and quality

- Generally apply techniques uniformly to all banks

– Assess need for adjustments to categorisation

- Designed with the help of system-level assumptions – AQR outputs will serve as the basis for following stress tests
• Main elements:
Data completeness and reconciliation

• Objective: provide a loan-level loss estimate for each portfolio to estimate the capital required to maintain a specific capital ratio under said scenarios
• Elements:

Data Integrity Verification (DIV)

Loss forecasting

Individual loan file review

Loss absorption capacity

Real estate collateral review

Capital needs (under the specified scenario)

© Oliver Wyman LON-FSP22401-197

7

1.1 Asset Quality Review

Asset Quality Review is split into a preparatory and three main phases

July –
August ‘14

February –
June ‘14

December ‘13 –
February ‘14

November ‘13 –
January ‘14

Phases of the AQR process

Phase 0
AQR
preparation

Phase 1
Portfolio
selection1

Phase 2
Performance of
AQR at national level • The AQR is a risk-based assessment and focuses on those elements of individual banks’ balance sheets that are believed to be most risky or non-transparent
• Specific objectives of the AQR are the
– assessment of adequate provisioning for credit exposures
– determination of the appropriate valuation of collateral for credit exposures
– assessment of the valuation of complex instruments and high-risk assets on banks’ balance sheets
• The AQR is divided into a preparatory and three main phases, which started in December 2013 and will last until
October 2014
• Detailed indications on the methodology for phase 2 are provided in a comprehensive manual + series of appendices
(> 300 paes)

Phase 3
Report writing
& disclosure

The complete methodology manual has been made public by ECB
1. Partially overlapping with phase 0

© Oliver Wyman LON-FSP22401-197

9

The core of the AQR consists of various work blocks, which started in
February and last until the end of July
Timeline for the key workblocks in Phase 2 of the AQR
February
1

March

April

May

June

July

PP & A¹ review

2 Loan tape creation

3

DIV²

4

Sampling

Projections on findings of CFR

5

Credit file review (CFR)

6

Collateral & RE valuation
Collective provisioning analysis

7
8

Level III Fair Value exposure review

CET1% adjustment calculation

9 Delivery/ check of CET1% breakdowns
Quality assurance

General remarks on the methodology for Phase 2 of the AQR
All methodological standards are set centrally by the ECB program management
The methodological details have been consulted with the NCAs
The majority of analyses will be performed based on templates and output forms defined by the ECB, that will be populated by Audit firms
Source: AQR Phase 2 Manual Version 1.0
1. Processes, policies & accounting review
© Oliver Wyman LON-FSP22401-197

2. Data integrity verification

10

2 The loan tape provides the data basis for the AQR which makes the loan

tape generation one of the essential steps in the overall process

Purpose of the loan tape

Scope of the loan tape

• Provides the basis for Data Integrity Validation and the evaluation of collective provisioning practices

• All credit exposures of the portfolios selected during
Phase 1 of the AQR and the portfolios connected to these • Constitutes the universe for random sampling for the
Credit File Review

• Information about all loans and advances and debt securities (excl. securitisation) which are held at amortised cost for the above portfolios
• Off-balance sheet exposure (excl. derivatives)

•

Loan tape
Structure of the loan tape
• The loan tape consists of 3 sub-tapes:
– Facilities tape
– Collateral tape
– Debtor tape
• Unique keys are requested such that facilities and collaterals can be aggregated up to the debtor level

Snapshots required for the loan tape
• Two complete snapshots of all three data tapes are required:
i. End of year – Dec 2012 ii. End of year – Dec 2013
• The 2012 snapshot is required for the evaluation of collective provisioning for retail exposures

EOY
2013
EOY
2012

Loan tape

Source: ECB Phase 2 Manual
© Oliver Wyman LON-FSP22401-197

11

4 The sampling process will be two-staged with an early identification of

top exposures followed by the selection of the full sample

Files from the selected in-scope portfolios for Phase 2 will be sampled in order to carry out the Credit File Review
General principles:
• Only portfolios selected for Phase 2 will be sampled
• No sampling of Retail exposures (with the exception of RRE)
• One sample will be selected for each portfolio
• Sampling process is designed to focus on areas of the portfolios with the greatest uncertainty (and therefore excludes significant parts of the portfolio)

1

Define basis for sampling

• Some parts will be excluded from sampling (see above) 2

Stratify portfolio • Stratification of the portfolios by risk and exposure

3

Select priority group

• Select top debtors by exposure size and riskiness bucket Begin Credit File
Review of priority groups 4

Select main sample

• Calculate sample sizes and designate specific borrowers 5

Select reserve sample

• Calculate sample sizes and designate specific borrowers Begin Credit File
Review for remaining
(main) sample

The sampling will be performed with no direct interaction with banks
Source: ECB Phase 2 Manual 1.0
© Oliver Wyman LON-FSP22401-197

12

5 The detailed Credit File Review (CFR) will provide information about

misclassification and under-/ overprovisioning on sampled exposures

Key steps followed by Audit Firms as part of the CFR

1

Credit File Review data preparation

• Banks collect and verify the completeness of the information necessary to complete the classification review • Timely data delivery required

2

Review of the classification of the exposure • Audit firms to review correct classification of the exposures in banks’ systems 3

Review of individual impairment and provisioning • For non-Retail exposures only: Audit firms to analyse appropriate provisions given debtor status

Required input / priorities for banks • Banks are required to deliver all information to perform (e.g. credit file documentation) the loan file review on time
• Open communication and interaction with Audit firms throughout on-site reviews is expected • Banks can start preparation with regards to identify locations of key documentation (incl. foreign subsidiaries) and key contact persons Source: ECB Phase 2 Manual

© Oliver Wyman LON-FSP22401-197

13

Credit file review - example
Assessment of Non Performing Exposures

1

Decide on going or gone concern approach

Going concern approach when:
• Operating cash flows continue and can service debt partially
• Collateral value is low and/or is not essential to the debtor’s cash flows

Gone concern approach when:
• Low or negative cash flows estimated going forward
• High collateral value and collateral is essential to the debtor’s cash flows

2

Perform gone or going concern approach & calculate provisioning Going concern approach
i. Steady-state cash flows: 1 period sustainable cashflows (EBITDA + Cash flow adjustment + Sustainability adjustment) x Multiple ii. Two-step DCF: Annual cash flow forecast and terminal value discounted by EIR1
Gone concern approach
i. Estimate the liquidation time frame required for the assets ii. Estimate the cash flows by years including
a. Income generated from the asset during liquidation window
b. Proceeds from asset sale based on the market prices (discounted by EIR)
i.
All physical assets should be valued on basis of market value in line with Blue Book or other international standards ii. An existing appraisal in last 12 months from external appraisers qualify (should be indexed to current value) iii. Estimate the cost of liquidation
i.
Market price haircut (based on liquidity of market and liquidation strategy) ii. Cost of legal enforcement and other liquidation costs iv. Analyze the lien structure to estimate the accessible share of collateral value

1. Effective interest rate of impaired facility
Source: ECB AQR Phase II Manual (public)
© Oliver Wyman LON-FSP22401-197

14

6 Collateral valuations will be used to identify if physical assets valuations

are correctly reflected in provisions or carrying values

Overview of the key elements of the approach

Required input / priorities for banks Complete sample selection for loan file review

Determine the collateral items for revaluation Perform review based on ECB standards • Identify the credit files under review based on sampling

• Determine the collateral sample for review – All items >€ 1 MM
– >90% of collateral items (in value) per debtor – Has not been revalued on market value since 1/1/13

• Real estate will be valued in line with
European Standards
EVS-2012 (Blue
Book) and other international standards (e.g. RICS guidelines) unless more conservative local valuation standards are used

• Identify the sampled exposures with collateral items attached • Banks are required to provide basic information on collateral items (e.g. address for real estate) to facilitate the valuation – failure to deliver will result in a value of 0
• Banks can start preparation by checking availability of key information on their collateral databases • Gather basic information required to perform valuations

Source: ECB Phase 2 Manual
© Oliver Wyman LON-FSP22401-197

15

7 Collective provision analysis challenge reported provisioning levels on

portfolios that would typically be impaired on a collective basis

Overview of the key elements of the approach
The collective provision analyses are conducted by comparing the results of a simple model developed by the NCA team A , with bank internal provisions B and analysing resulting buffers or gaps C
C

Potential buffer / gap analysis

Required input / priorities for banks • Banks are asked to provide the output of their own collective provisioning model

Challenger Model

Bank provisions

• Banks are asked to support NCA
Bank Teams in retrieving the variables require to parameterise the Challenger Model

Estimate collective provisions based on a 1 year point in time expected loss

In place bank provisions as reported • If existent, banks will be requested to provide the IRB validation reports A

B

SME
Corporate
Real Estate related
Retail Mortgages

Challenger
Bank
Model
Provisions

• Banks are asked to cooperate in the investigation of drivers for a potential buffer or gap between the Challenger Model outcome and Bank provisions

Retail Other

Source: ECB Phase 2 Manual 1.0
© Oliver Wyman LON-FSP22401-197

16

1.2 Stress test: recent experiences from selected European countries

The bottom-up stress testing exercise estimates the potential capital need by comparing entities’ projected losses against their loss absorption capacity
Stress test: Oliver Wyman methodology

Projected loss calculation
(loss forecasting)
• Bottom-up loan-level
“economic loss” valuation of key assets/portfolios
• With particular emphasis on high risk areas
• 2 scenarios (base vs adverse stress) considered
I

I

Illustrative

Projected loss calculation
(Loss forecasting)

III
Potential
capital needs Capital buffer II Loss absorption capacity

• Detailed and accurate preprovision profit generation forecasting (structural P&L component analysis)
1. Existing provisions currently in BS

PPP

II
Loss
absorption capacity 1. Bottom-up modelling – PD
2. Embedding of AQR results

2. Accumulated pre-provision profit:
(PPP = NII - Fees - Costs)
3. Capital buffer: Credit deleveraging

Provisions

3. Macroeconomic modelling
III Potential capital needs
4. LGD modelling
2013

2014

2015

5. Collateral asset valuation review
6. Exposure projections

© Oliver Wyman LON-FSP22401-197

New book
Performing loans
Non-performing loans

• Capital excess/deficit under baseline and adverse scenario
• Serves as a basis to derive final requirements 18

Illustrative

I

Loss forecasting built combining Macroeconomic models with micro analytics on specific risk factors (e.g., Default Rate – DR)
Data input

Stationarity analysis Projections for macroeconomic variables
Source: IOs

© Oliver Wyman LON-FSP22401-197

3. Final
Credit Quality
Indicator
projections

Households
(Retail)

Households
(Retail)

Households
(Retail)

Corporate

Corporate

Small business Small business Small business RED

RED

RED

Filter 1
Filtering based on statistical significance and economic intuition
(R-squared, P-stat, signs)

Macro data
Source:
Country A
Central Bank,
OxEcon,
Eurostat

2. Final model identification Corporate

DR and NPL data Source:
Country A
Central Bank

Stationarity analysis and transformation Macroeconomic model
1. Univariate and multivariate regressions for variable and model shortlisting

Filter 2
Filtering based on goodness of fit and economic sense
(Shape, credibility of projections) 19

II

Loan loss provisions, pre-provision profit and the free capital buffer are the three components of Loss Absorption Capacity
Real case example – pre SSM exercise supported by OW in a european country Loss absorption capacity component break-down

Recap of cornerstones

1
2

• Clear rules and constraints to apply LLPs to absorb losses
• Confirmation of eoy 2012 baseline LLPs
• Analyse breakdown into general/ generic
(performing) and specific (non-performing) LLPs
• Scenario definition
• Confirmation of assumptions and methodological questions • Business plan reconciliation
• Benchmarking of PPP drivers
• Normalisation of PPP to BoS base case
• Stressing of normalized PPP to BoS adverse case 1

In-place loan loss provisions
(LLPs) 2012

2

Pre-provision profit (PPP)
2013-2015

3

Capital buffer / shortfall 2015

Total Loss
Absorption
Capacity
2015

3

• Confirmation of 2012 Core Tier 1 capital baseline
• Confirmation of 2012 Regulatory Capital
Requirement (RWAs) and breakdown
• Projection of Regulatory Capital Requirement
(RWAs) over 2013-2015
• Projection of 2015 capital buffer in excess of minimum capital requirements

Notes: LAC = Loss Absorption Capacity; LLPs = Loan Loss Provisions; PPP = Pre-Provision Profit; ST = Stress Testing
Source: XYZ Stress Test and Asset Quality Review documentation
© Oliver Wyman LON-FSP22401-197

20

Illustrative

III

The capital ratios are projected based on current available CT1 capital and forward projected RWAs/capital requirement

Pro-forma capital adequacy calculations
Pro-forma baseline

3 years pro-forma projection
D RWA projection

Ops risk Market risk Capital buffer x 8%1

Ops risk Market risk x 8%1

Capital shortfall Credit risk Credit risk A

B

B

Current available CT1 capital

Current
RWA

Current ST capital requirement

C
Current
capital buffer / shortfall A

B

Future proFuture proforma available forma RWA
CT1 capital

B

C

Future proforma ST capital requirement

Pro-forma future capital buffer/ shortfall

1. 8% for base case; 5.5% in adverse case
Notes: LAC = Loss Absorption Capacity; LLPs = Loan Loss Provisions; PPP = Pre-Provision Profit; ST = Stress Testing
© Oliver Wyman LON-FSP22401-197

21

2 Applying the new methodologies to the prevailing context in MENA region: potential scenarios

Risk factors that could determine shocks for Banks in MENA region (1/2)
High volatility of real estate prices are common in the region
House prices in Dubai
% change y/y

Dubai is a case study for House prices in the region

50%

After the drop in 2008, house prices have been fairly stable even during troubled times

40%
30%
20%
10%
0%
-10%

Q1'08

Q1'09

Q1'10

Q1'11

-20%
-30%
-40%

Q1'12

Q1'13

• House prices recovered from the 20082009 severe drop
• Real Estate prices have been resilient since the 2009 drop, they have not been affected by
– Troubled times in the region (in 2010)
– and in Europe (in 2011)
• Dubai area market outlook for is positive because of economic growth and positive impact linked to International events (e.g.
Expo 2020)

-50%

Another big financial crisis could hit house prices as in 2008, and the shock could spill over on banks’ asset quality (increasing NPLs)
Source: Colliers International. Oliver Wyman analysis.
© Oliver Wyman

23

Risk factors that could determine shocks for Banks in MENA region (2/2)
Oil price and GDP shock scenario
Oil price and MENA GDP
% change y/y

Oil price plays a big role in MENA region economy 100%
After 2009, GDP of main MENA countries is highly correlated to oil price

80%
60%
40%

• MENA region economy is heavily dependent on the oil sector
• Regional GDP is correlated with the oil prices; this is more evident after the crisis and can clearly be seen following 2009

20%
0%
2003

2007

2011

-20%
-40%
-60%

During financial crises
GDP of MENA countries shrank

• In crises period the effect is less evident and external facotrs rather than oil price can effect GDP of MENA economies (e.g.
2008 – 2009)

Oil price
MENA region GDP
Saudi Arabia GDP
UAE GDP

• A large and prolonged decline in the Oil price could lead to negative effects on Nonoil growth, hence on Corporate and
Banking sectors

An oil price shock could spill over on banks’ assets quality (increase in NPLs)
Source: Datastream, Oxford Economics. Oliver Wyman analysis.
© Oliver Wyman

24

3 Implications for MENA Banks

The trend towards increasing use of AQR and stress test in banking supervision will generate important implications both at Regulators and at
Bank level

Regulator level

Bank level –
“technical”

Bank level –
“strategic”

© Oliver Wyman LON-FSP22401-197

1
2
3
4
5
6

• Adapt regulatory approach in AQR/Stress Test definition to recent best practices • Introduce rigorous Recovery and Resolution Planning requirements

• Ensure availability of relevant data
• Introduce advanced methodologies and tools to allow appropriate stress simulations • Preserve capital adequacy under stress scenarios
• Steer business decisions with the objective to maintain required robustness under severe assumptions

26

1 Adapt regulatory approach to recent best practices methodology to a
Regulators should move from the an ad-hoc sporadic structured exercise conducted on a regular basis
Case Study: Saudi Arabia Stress Test result
Assessment of banking sector’s soundness in 2011

Banks are well capitalized

However, a severe shock could undermine stability

IMF recommendations

• Top-down stress tests were used to assess resilience of the banking sector • All banks could withstand a severe shock (e.g. uniform nonperforming loan increase, default of their largest debtor, etc.)

• Banking sector could suffer from prolonged oil price decline
• Deposit runs could affect banks funding • Small and domestic banks are most vulnerable

1

Stress testing and Financial System policies should be further strengthened • Incorporate lessons into prudential rules and supervisory actions
• Develop a formal policy framework

2

Regulators should increase the range of shocks included in stress tests
• Include other emerging risks such as oil prices • Formalize systemic liquidity management, financial safety nets and resolution frameworks

3

Regulators should conduct stress tests on a wider range of shocks on a regular basis Source: IMF Country Report No. 12/92 - Saudi Arabia: Financial System Stability Assessment—Update
© Oliver Wyman

27

Planning (RRP)
2 Introduce rigorous Recovery and ResolutionGlobal SIFI, RRP requirements
Originally introduced only with reference to requirements are being progressively extended to all banks
Recovery vs. Resolution Plans

Recovery plans Resolution plans To prevent bank from failing
• Defining management actions and triggers, with the aim of restoring the capital and liquidity position, and continuing operations
• Tied to stress testing and scenario analysis (“reverse stress testing”)
• Overseen by regulators’ supervisory staff

To prepare for orderly bank failure and transition to a bridge bank under government or 3rd party ownership • Defining key operations and assets to enable an orderly transition
• Identifying potential roadblocks to an orderly resolution and appropriate mitigating plans
• Overseen by regulators’ resolution staff

© Oliver Wyman LON-FSP22401-197

The authority's objectives in recovery and resolution planning 1. Identify initiatives in order to improve credibility and feasibility of recovery actions and level of
“resolvability” of the institution
2. Improve quality and effectiveness of RRP and decreasing the cost of failure

Potential focus areas for recommendations
• Operating model (e.g. decouple/create redundancy in the shared services in order to improve capability of the bank to break up in an orderly fashion)
• Business model (e.g. reduce volumes/ riskiness on trading and derivatives positions)
• Company set-up (e.g. segregate risky activities from deposits, hard ring-fence, or create a pure holding company)
• Capital constraints and limitations to financial resources circulation
(e.g. adopt additional discretionary capital add-on or intra-group exposures, free circulation of capital and liquidity)

28

relevant data
3 Ensure availability ofcan lead to huge data challenges
Loan tapes creation
Category

Examples

# Fields

ID contract
Legal Entity
Product type
Undrawn/ drawn amount

•
•
•
•

Maturity
EAD/ LGD
Provision
…

1

Contract information

•
•
•
•

2

Counterparty information (legal entity)

•
•
•
•

ID counterparty segment Financials
PD

•
•
•
•

Rating
Length of relationship
Geography
…

~65

3

Counterparty information (individual)

•
•
•
•

Name
Geography
Client segmentation
Income

•
•
•
•

PD
Length of relationship
Employment status
…

~45

4

Collateral &
Guarantees

•
•
•
•

ID Collateral/Guarantee
Type of collateral
Value of collateral
Valuation date

•
•
•
•

Seniority of guarantee
Lien rank
Guarantor
…

~20

Underlying assets

•
•
•
•

ID
Type
Valuation
Date of Valuation

•
•
•
•

Dimensions
Appraisal
Location
…

~85

Cash flows

•
•
•
•

Type of cash flow
Product type
Projections 2013- 2030
…

5
6

TOTAL

© Oliver Wyman LON-FSP22401-197

~125

~20
~360

29

4 Introduce advanced methodologies and tools to allow appropriate stress simulations A granular modelling approach allows to leverage a broad spectrum of information sources, generating detailed and accurate results
Inputs & Parameters

Scenarios

Scenario impacts

Modelling

Modular projections

Scenario results

Macroeconomic

Top-down planning and Stress Testing
Toolkit

Risk parameters (PD, LGD, EAD)

Probability of default
Bank strategy

50%

40%

PD (%)

40%

36%

20%

34%

10%

Inputs

LGD (%)

38%

30%

32%

0%

30%
2012

Secured

2013

2014

Base

2012

Adverse

2013
Base

2014
Adverse

Portfolio data
LGD

Parameters

Unsecured

Expected Loss
Balances EL (€)

PD models

PD

EL (%)

2011

Stock

Total

Non-performing loans 2011

EL
12-14

EL
11-14

EL
12-14

EL
11-14

(ALL) - by DRC segment

Cure rates

Mortgages

Cured balances flows 45.847

RED

12,64%

17,8%

25,4%

1.874

12,2%

23,7%

5.578

723

280

1.004

8,5%

18,0%

5.917

742

314

1.056

8,1%

17,8%

5.770

704

369

1.073

8,4%

18,6%

143.290

17.663

11.481

29.144

12,5%

20,3%

Total

Defaulted balances flows EAD

5,9%

18.340

556

Consumer Finance

Collateral haircuts

5.796

8.229

1.317

SME

LGD

1.731

10.111

7.900

Large Corporates

LGD models

4.065

72.278

Public works

Adjustments

Balances evolution (performing, non-performing, substandard)
200
160

EAD

Repayment & prepayment CCF models

© Oliver Wyman LON-FSP22401-197

200

Portfolio composition (€MM)
Base case

180

160
140

120

120

100

Amortization/ prepayment Portfolio composition (€MM)
Stress case

180

140

100

80

80

60

60

40

40

20

20

0
2011

2012
Substandard

2013
Performing loans

2014
NPLs

0
2011

2012
Substandard

2013
Performing loans

2014
NPLs

30

(1/2)
5 Preserve capital adequacy under different scenariosmean safety in a
Capital adequacy in a normal environment does not stressed situation
…but the situation could change dramatically in stressed conditions
Results of stress test on Saudi Arabia banks in 20113

Capital of top Mena banks in normal environment is above regulatory requirements…
Top 5 MENA banks1 Tier 1 ratio and Capital Adequacy
Ratio (CAR)2, 2013 data
Tier 1
19.6%

15.6% 15.6%

16.2%

17.1%

CAR

19.6%
18.5%
18.2%
16.5%

15.3%

• Scenario defined
– Oil price decline to historical average of $40 a barrel
– Evolution of System’s Capital Adequacy
Ratio

Evolution of System’s Capital Adequacy Ratio
17.7%

16.2%

14.8%
10.8%
5.2%

National
Qatar
National Bank Commercial
Bank (The)

Total assets $ bn

Emirates National Bank Al Rajhi Bank
NBD PJSC of Abu Dhabi

-1.9%
2010

122

101

93

89

2011

2012

2013

2014

2015

75

Source: Annual reports. Banking Sector: structure, vulnerability and stress test, September 2011. Oliver Wyman analysis.
1. Oliver Wyman on 2012 assets 2. Total regulatory capital 3. The exercise used 2010 data.
© Banks ranked

31

under different scenarios
5 Preserve capital adequacyPlanning will become a key(2/2)
Recovery and Resolution priority Examples of recovery measures
Example measures

Restructuring and sale of assets Sale of business activities

Capital increase
© Oliver Wyman LON-FSP22401-197

• Reduction of structural costs,
e.g. bonus or staff
• Sale of non-core assets
• Sale and lease back of self used real estate
• Sale of other real estate and property • Sale and outsourcing of back office functions
• Sale of minority holdings
• Buy backs of hybrid instruments
• Sale of subsidiaries
• Sale of foreign branches
• Sale of group servicing activities • Portfolio sales
• Issue of new capital
• CoCoCos

Liquidation/ deposit pay-off

Closed bank

“Business as usual” managerial levers

• Cost savings, e.g. marketing, legal, IT
• Sale of liquid trading book assets • Hedging of credit portfolios
• Restructuring of asset-portfolios
• Other funding measures
(e.g. tighter collateral management) Closed
Purchase &
Assumptions
(P&A)

Bridge bank • Failed bank’s credit machinery stop, and deposit accounts cease
• Deposit-guarantee fund pays off insured depositors
• Purchase of all or some assets and assuming liabilities by a healthy institution
i.e. a merger, but for a closed bank
• Remaining assets sold/wound-down
• Temporary entity used to maintain essential services and preserve franchise value
• Institution sold or liquidated in an orderly way in the mid- to long-term

Direct government aid

Open bank

Category

Potential long list of resolution options

Assisted merger Private sector solutions • Bank issues new equity to the government
(common/preferred stock)
• Usually accompanied by liquidity assistance from the central bank
• Merger with another healthy financial institution with government assistance in the form of guarantees or direct support

• Capital infusion from existing shareholders or other interested parties
• Merger with another healthy financial institution

32

with the objective
6 Steer business decisions assumptions (1/2) to maintain required robustness under severe
The experience of repeated stress tests required many banks to reduce exposure towards some vulnerable sectors
Group of UAE’s banks1
% total loans2, 2013 data

Regulators might require the bank to reduce exposure to RE sector

27% of total loans in 2012

Other
Real Estate & Construction

22%

28%

Mostly Energy (7%)
Manufacturing (5%)

• Real Estate and construction loans represent a relevant share (28%) of the bank’s loans portfolio
• Another house price shock could affect asset quality (increasing NPLs)

Personal loans

• Data from 2009 house price decline could be used for stress tests purposes

18%
16%

Banks & FI
16%
Sovereign

Banks should carefully manage concentration risk and modify their business decisions based on stress test results
Source: Annual reports. Oliver Wyman analysis.
1. Oliver Wyman
Group composed of NBAD, Union National Bank, ADCB. Note that for ADCB loans to hospitality sector were also considered in Real Estate sector. 2. Loans gross of the allowance for
©
impairment and interest suspended.

33

with the objective
6 Steer business decisions assumptions (2/2) to maintain required robustness under severe
Banks might consider to adopt capital optimization actions which may result more effective in stress conditions
General types of levers

Typical impact on RWA (*)
• Other things being equal the impact on RWA should be not material 1

Increase portfolio exposures in less cyclical sector

2

• It can lead to an increase of
Increase geographic diversification RWA in case the portfolio is of the portfolio diversified in non EU countries with bad credit quality

3

• If collaterals are eligible, have the same face value and are of the same category, the impact on RWA should be negligible

Use collaterals with less volatile values • It can impact RWA if the bank use internal rating models that include the industry as one of the information used to assign the rating

Typical impact in case of stress
• Non cyclical sectors will be less impacted by an economic downturn and will produce a lower level of expected losses

• Countries less correlated with the national economy will be less penalized by a downturn scenario
• Collaterals with a more stable value should give more protection in case of an economic recession

(*) Depends on the use of Standard vs. Advanced approach and the characteristics of the rating models.
© Oliver Wyman LON-FSP22401-197

34

4 Concluding remarks

Summary and conclusions

• Asset Quality Review and Stress test are increasingly becoming the standard methodological approaches used by various regulators in Western Economies to supervise Bank’s safety and soundness • These approaches are widely appreciated by the market professionals, who see the outcome of their application as an implicit “stamp” guaranteeing the quality of Banks’ balance sheet and financial accounts. Still, their application requires huge efforts and commitment from a wide range of actors involved
• Should a trend towards greater usage of these methodology emerge also in MENA region, we see six main implications – two of them for Regulators, the remaining four applicable to Banks
– Regulatory Authorities should 1. adapt their approach in AQR/stress test application to recent best practices and 2. introduce rigorous Recovery and Resolution Planning requirements – Banks, in turn, should 3. ensure availability of relevant data, 4. introduce advanced methodologies and tools to allow appropriate stress simulations, 5. preserve capital adequacy under stress scenarios and 6. steer business decisions with the objective to maintain required robustness under severe assumptions

© Oliver Wyman LON-FSP22401-197

36