THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
September, 2018
The BoR Approach to MacroprudentalStress Testing
Financial Stability Department
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING2
Assessment of losses and absorbing capacity of financial system in the stress scenario taking into account contagion
risks and feedback effects within the system (banks, non-bank financial organizations, non-financial sector)
Detection of most significant risks for financial system and economy
Close monitoring of identified vulnerabilities and development of policy toolkit
(precautionary and anti-crisis measures)
Macroprudentialstress test
is conducted by the authority responsible for ensuring financial
stability (the Bank of Russia)
PURPOSE OF MACROPRUDENTIAL STRESS TESTING
*The Bank of Russia published the Concept of macroprudential stress testing for public consultations in November 2017
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING3USE OF MACROPRUDENTIAL STRESS TEST FOR FINANCIAL STABILITY PURPOSES
Planning anti-crisis
measures
Implementing
macroprudential
policy
•Assessing amount of funds that may be
required for recapitalization of banks
•Planning anti-crisis refinancing policy in
ruble and FX
•Timing of regulatory forbearance
measures
Macroprudential stress testing is supposed
to provide analytical support for
macroprudential policy framework
(countercyclical capital buffer and
macroprudential add-ons to risk weights)
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING4KEY FEATURES OF MACROPRUDENTIAL STRESS TEST OF THE BANK OF RUSSIA
Group basis (the largest banks and financialgroups, covering 80% of the financial system)
The most granular data(trade repository database,transaction based reports,surveys of financial sector)to assess financial risks
Relevant stress scenario (extreme butplausible commodity shock), all significantfinancial risks (market, liquidity, credit), fivetime horizons
Mutual financialexposures and contagionrisks in the financial sector(network analysis of moneymarket, derivative market)
Macroprudential stress testof the Bank of Russia
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING5MACROPRUDENTIAL STRESS TEST PERIMETER
Stress test perimeter (2017) – 39 largest banking and financial groups
• Banks (73 banks which account for about 85% of the banking sector assets)
• Non-governmental pension funds with an amount of assets under management
in each fund exceeding 20 billion rubles (14 NPFs which accumulate almost 90%
AUM)
• Systemically important insurance companies (14 insurance companies, which
account for more than 60% of assets)
Stress test perimeter extension (2018)
• Broker-dealers
• Development institutions (Vnesheconombank, DOM RF)
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
On demand side (world economic
growth decline)
Global risks
On supply side (extensive growth
of oil production in US and other
countries)
KEY GLOBAL RISKS WHICH MAY HAVE NEGATIVE IMPACT ON RUSSIA
Decline in oil prices
Potential shock• Major central banks policy normalization
• Escalation of trade conflicts
• Political risks
Spillover effects
Market risks (volatility, higher
rates, capital outflow, FX liquidity
shortage)
Negative impact on
financial sector
Macroeconomic effects
(GDP growth decline, budget
deficit)Second round
effects
6
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING7
HISTORICALLY OIL PRICE HAS BEEN A KEY FACTOR DETERMINING PRIVATE CAPITAL FLOWS IN RUSSIA
Source: Bloomberg, Thomson Reuters, Bank of Russia.
0
20
40
60
80
100
120
140
-100
-80
-60
-40
-20
0
20
40
31.0
3.2
00
1
31.1
2.2
00
1
30.0
9.2
00
2
30.0
6.2
00
3
31.0
3.2
00
4
31.1
2.2
00
4
30.0
9.2
00
5
30.0
6.2
00
6
31.0
3.2
00
7
31.1
2.2
00
7
30.0
9.2
00
8
30.0
6.2
00
9
31.0
3.2
01
0
31.1
2.2
01
0
30.0
9.2
01
1
30.0
6.2
01
2
31.0
3.2
01
3
31.1
2.2
01
3
30.0
9.2
01
4
30.0
6.2
01
5
31.0
3.2
01
6
31.1
2.2
01
6
30.0
9.2
01
7
From June 2004 to June 2006 Fed funds rate (upper bound) wasincreased from 1% to 5.25%, till August 2007 the rate remained at5.25%Since December 2015 Fed funds rate (upper bound) was increased from0.25% to 1.75%
Net private capital inflow in/outflow from Russian banking sector,quarterly balance of payment data (bln USD)
Urals (USD per barrel), right axis
Oil price and capital flows
1.
2.
1
.
2
.
20%
25%
30%
35%
40%
45%
50%
55%
60%
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
20
14
20
15
20
16
20
17
Share of oil revenues in total federal budget revenues (%)
Share of revenues from exports of crude oil and oil products inthe total export earnings (%)
Source: Ministry of Finance, Bloomberg.
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
•The base price of oil is supposed to redistribute additional oil and gas revenuesbetween phases of the oil cycles, smoothing fluctuations in the budget expenditures
8BUDGET RULE HELPS TO LIMIT VOLATILITY CAUSED BY OIL PRICE FLUCTUATIONS
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Stress testing
horizonType of risk Perimeter
2 working days Market risk, liquidity risk Banks, National Clearing Centre (CCP)
1 month Market risk, liquidity risk Banks
1 quarter Market risk, interest rate riskBanks, insurance companies, non-
governmental pension funds
1 yearMarket risk, interest rate risk,
credit risk, insurance risk
Banks, insurance companies, non-
governmental pension funds
2 yearsMarket risk, interest rate risk,
credit risk
Banks, insurance companies, non-
governmental pension funds
STRESS TESTING HORIZON AND TYPES OF ASSESSED RISKS 9
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Materialization of a stress scenario (approach to each market instrument is defined separately on historical data - 10 years or crisis periods, CVaR 1%)
The need for additional liquidity
• To satisfy margin calls by adding cash or securities to CCP
• To maintain positions on the OTC market
Correction of positions in the money market
• Participants with liquidity shortage decrease amount of lending
Participants having insufficient amount of liquid assets fail to fulfill obligations to CCP
If the lines of defense are insufficient CCP implements loss-allocation procedures
STAGES OF A 2-DAY STRESS TEST IMPLEMENTATION 10
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING11RESULTS ON A 2-DAY HORIZON
The ratio of high liquid assets to required amount of liquidity
-50
0
50
100
150
200
250
300
Bank 1
Bank 2
Bank 3
Bank 4
Bank 5
Bank 6
Bank 7
Bank 8
Bank 9
Bank 1
0
Bank 1
1
Bank 1
2
Bank 1
3
Bank 1
4
Bank 1
5
Bank 1
6
Bank 1
7
Bank 1
8
Bank 1
9
Bank 2
0
Bank 2
1
Bank 2
2
Bank 2
3
Bank 2
4
Bank 2
5
Bank 2
6
Bank 2
7
Bank 2
8
Bank 2
9
Bank 3
0
Bank 3
1
Bank 3
2
Bank 3
3
Bank 3
4
Bank 3
5
Bank 3
6
Bank 3
7
Bank 3
8
Bank 3
9
Bank 4
0
Bank 4
1
Bank 4
2
Bank 4
3
Stress scenario Baseline scenario
Liquidity
shortage
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Macro scenario
Interest rate
risk
Exchange rate
Equities RiskCredit risk for
Russia
Non-ruble interest
rates for Russian
eurobonds
Credit risk
spread
Ruble
interest
rate
Shares and
depositary
receipts
Russian
sovereign
bonds
Russian
corporate
bonds
Russian
sovereign
eurobonds
Russian
corporate
eurobonds
Foreign
bonds
Russian equities
price drop
Foreign
exchange risk
Modified duration
MARKET RISK 12
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING13
Corporate portfolio stress testing
Macroeconomic
parameters of the stress
scenario (Monetary
Policy Department)
Financial indicators
(Federal State Statistics
Service)
Projected financial
indicators
Data on loan
indebtedness (credit
registry)
Projected provision
Estimate of loan loss
provision taking into
account collateral
Neuralnetwork
Panelregression
CREDIT RISK
• Gross profitability of the borrower• Current liquidity ratio• Borrower’s autonomy ratio• Interest rate coverage ratio• Relation to tradable sector • Current provision
• Number of days the loan is overdue
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Baseline
scenario
Stress
scenario
Number of banks with liquidity shortfall
Banking groups (of which systemically important credit institutions)
Liquidity shortfall – difference between deposit outflow and amount of liquid assets
(banks with liquidity shortfall)
Liquidity deficit coverage
Amount of intragroup liquidity support
BoR emergency liquidity support measures
Liquidity deficit after support measures
Banking groups with liquidity shortage after support measures (of which systemically
important credit institutions)
14LIQUIDITY SHORTAGE ON A 1-MONTH HORIZON (IN RUBLES AND FX)
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Baseline
scenario
Stress
scenario
Number of banks with capital shortfall
Banking groups (of which systemically important credit institutions)
Total losses
Market risk*
FX revaluation
Capital needed to comply with regulatory required level of capital adequacy ratio (at 8%)
Groups of systemically important credit institutions
Other groups
15CAPITAL SHORTAGE ON A 1-MONTH HORIZON
* Market risk can be partly mitigated by temporary relax measures - the right not to revaluate bonds (transferthem to the held to maturity bond portfolio), as in December 2014. These measures can help “to win time” andlimit contagion effects.
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Cumulative dataBaseline
scenario
Stress
scenario
Total losses, of which
Credit risk
Market risk
Interest rate risk
Credit risk of insurance companies and non-governmental pension funds
Results of insurance companies activities (including insurance risk)
Positive FX revaluation
Capital deficit in the financial sector (capital needed to comply with regulatory required level of
capital adequacy ratio at 8%)
Systemically important credit institutions
Other groups
Number of groups with capital shortfall
Systemically important credit institutions
Other groups
16STRESS TEST ON 1-YEAR AND 2 YEAR HORIZONS
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
• Contagion effect reflects the amount of losses suffered by the stress test participants due todefault of their counterparties in the previous period
• Prevention of the contagion effect at an early stage helps to reduce the capital deficit in thelate horizons
1 Month 1 Quarter 1 Year 2 Years
Str
ess
scenario
Number of groups with capital/liquidity shortfall
Total losses
Losses from network effect
Share of network effect, % 36 59 23 20
17CONTAGION EFFECT
Group-defaulters with the ratio of contagion effect to capital deficit > 1 can be attributed to
“hidden” systemically important organizations. Elimination of the contagion effect coming from
these groups may reduce the capital deficit of other stress test participants
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
Bank 1
Bank 2
Bank 3
Bank 4
Bank 5
Bank 6
Bank 7
Bank 8
Bank 9
Bank 1
0
Bank 1
1
Bank 1
2
Bank 1
3
Bank 1
4
Bank 1
5
Bank 1
6
-8
-4
0
4
8
12Before ST
Additional Capital
Capital Conservation Buffer
SIB Buffer
Bank 1
Bank 2
Bank 3
Bank 4
Bank 5
Bank 6
Bank 7
Bank 8
Bank 9
Bank 1
0
Bank 1
1
Bank 1
2
Bank 1
3
Bank 1
4
Bank 1
5
Bank 1
6
After ST
Additional Capital
Capital Conservation Buffer
SIB Buffer
Deficit Low buffer
MACROPRUDENTIAL MEASURES
• Capital deficit (identified as a result of macropru stress test) can be covered withcountercyclical capital buffer (CCyB)
• Banks with substantial capital buffer can be insensitive to increasing CCyB, but may bevulnerable to particular shocks => potential to use additional Pillar 2 requirements or sectoralmacropru add-ons
18
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING19
Comparison and cross-checking the results of the top-down stress test
conducted by the Financial Stability Department with the results of the bottom-up
stress test
Analysis of the network effect:
• Interconnectedness of the stress test participants
• Results of the macropru stress test taking into account influence of support
measures and contagion effects within and between the groups
Attribute analysis of total losses of macropru stress test participants broken down
by participants, risks, instruments
Extrapolation of losses (by type of risks) to other organizations beyond the
macropru stress test perimeter
RECOMMENDATIONS FOR SUPERVISORS
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING20APRIL 2018: IMF MISSION ON MACROPRUDENTIAL STRESS TESTING
Using a set of scenarios
• Development of an annual plan on stress testingfor 2019 (including macropru ST, supervisory ST)
• Creation of a Working Group on stress testing atthe BoR
Current work at the BoRIMF key recommendations
Clearer organization process and internalcooperation between departments
Implementation of an expected loss basedapproach for credit risk, better integrationof models
Additional data collection for risk analysis
Enhancing the use of contagion models
• Improvement of reporting forms on interest raterisk, liquidity risk
• Plan to conduct a survey of expected losses onbank portfolios
Development of several scenarios for top-down ST(cyclical, structural, alternative, etc.)
Work on PD model to assess credit risk
Work in progress
THE BOR APPROACH TO
MACROPRUDENTIAL STRESS TESTING
THANK YOU FOR YOUR ATTENTION!
Top Related