Stress testing for competitive advantage beyond regulatory - - PowerPoint PPT Presentation

Stress testing for competitive advantage beyond regulatory compliance

Led by:

The Center for Financial Professionals & Dennis Bennett, MRMIA

Presenters:

Tally Ferguson, BOK Financial Venkat Iyer, Santander Chris Smigielski, TIAA Bank

Tally Ferguson Director of Enterprise- Wide Risk BOK Financial

MEET THE PRESENTERS

Dennis Bennett CEO & Founder MRMIA Venkat Iyer Director of PPNR Forecasting Santander Chris Smigielski Vice President, Director of Model Risk Management TIAA Bank

Components of Performance

• Banks are Financial Intermediaries
• Money is the product
• Interest Rate is the Price
• Asset/Liability Management
• Interest Rate Risk (Rates/Yield Curve)
• Balance Sheet Assumptions
• Earnings at Risk (NII simulation)
• Earnings Trajectory and Likelihood
• Interest Income
• Loans
• Investment Securities
• Trading Securities
• Interest Expense
• Deposits
• Wholesale Funding
• Brokered Deposits
• Provision for Loan Loss
• Net Interest Income (NII)

1

Components of Performance

• Competitive Landscape
• Product Offerings
• FinTech Companies
• Volume and Price Impact
• Stress Testing
• Scenario Planning
• Comprehensive Assessments
• Measure ‘Realistic’ Responses
• Non Interest Income
• Fees
• Transactions
• Asset Management
• Brokerage and Trading
• Other Services
• Non Interest Expense
• People
• Infrastructure
• Fraud

2

Insight into Strategy Decisions

1

Good

• Navigate

businesses to financial success in the face of uncertainty Bad

• embarking on a

venture/business line that we cannot deliver or that leads to ruin

Insight into Strategy Decisions

• Sources of Ruin

2

Uncertainty

• Range of economic

and environmental factors considered too narrow

• product/service

knowledge base too limited Unexpectedly Poor Performance

• impact of

environmental factors not well modeled

• Controls not put in

place, or not effective Panic Exit

• Survived the tail

event without knowing it

• “Anchoring” error

from behavioral finance

Images from historicalwallpapers.blogspot.com

Stress Testing as a counter to sources of ruin

• Uncertainty
• CCAR/DFAST imposed discipline on forecasting PPNR from

economic factors

• Idiosyncratic stress testing can give us insight into environmental

factors

• Litmus tests
• What economic factors will lead to deviations from

expectations?

• What environmental factors will influence our results
• For which of these factors can we measure impact on

product/service results

• How confident are we assessing risk from factors whose

impact we cannot measure

Insight into Strategy Decision

Stress Testing as a counter to sources of ruin

• Unexpectedly Poor Performance
• Apply backtesting discipline we developed for CCAR/DFAST to

idiosyncratic stress tests

• Use Expected Shortfall or Conditional VaR approach for

products/services where worst case is not known

• Define controls to keep results in the fairway and test them

Insight into Strategy Decision

Stress Testing as a counter to sources of ruin

• Panic Exit
• If expected shortfall analysis leads to untenable results, don’t

enter the business, or change controls.

• Emphasize that low probability events are NOT zero probability

events and should not be dismissed.

• Challenge board and executive level risk tolerance decisions – are

they really as big as they say they are?

• Re-assess risk/return trade-off before exiting.

Insight into Strategy Decision

1

Aligning stress testing with business processes

Ø Planning and budgeting

• Use stress testing tools to support planning projections for strategic (medium to longer term) and

budget (typically one year) decisions

• Use of macro-sensitive and driver based models in addition to traditional bottoms-up FP&A

methods will increase transparency and flexibility of the budget process Ø Capital allocation

• Cascade top of the house stress absorption capital to individual LOB/products
• Use of granular PPNR and loss models will assist downward allocation of risk based capital.

Variance between base and stress scenarios provide good insights into LOB capital requirements and future scenario design

• Granular models enable product level risk and return trade-offs

Ø Portfolio optimization

• New products and initiatives can be assessed for relative risk-return attractiveness
• Constrained optimization frameworks can be built with risk, return and capital targets and/of

constraints

• Mix of segments within or across products can be assessed within the optimization objectives

2

Planning: Revenue components and stress drivers

§ PPNR components: Interest income, Interest Expense, Non-interest income, Non-interest Expense § Interest income is driven by loan balances and yields § Generally declines under stress due to balance contraction and lower yields on new and variable rate loans. § Other interest income moves due to changes in securities run-off to optimize the balance sheet § Interest expense driven by deposits for most commercial banks § Generally decreases under stress due to decline in balances and lower yields § Can be off-set by other borrowing costs based on funding requirements § Idiosyncratic events can be used to stress funding constraints § Non-Interest income driven mostly by fee based revenue § Loan, deposit and investment fees decline under stress due to asset and balance declines § Operating lease assets flow through non-interest income – impact similar to loan balances § Assets such as mortgage servicing rights can see significant volatility based on interest rates § Non-Interest expense driven by personnel and operating expenses § Personnel expenses generally decline under stress but trade-offs exist between categories § Operating lease depreciation, operational loss and FDIC insurance expenses can drive NIE higher

PPNR = Pre-Provision Net Revenue

3

Planning: Methodologies for revenue components

Methodology summary

§ Total Balance and component approaches practiced § Total balance used when assets/liabilities are less structured or high volatility in component level drivers

§ Non maturing deposits § Revolving lines of credit

§ Component approaches preferred for structured assets such as amortizing term loans § Combinations of statistical and expert judgement models § NII approaches have been enhanced through:

§ Rigorous model risk management through SR11-7 § Quantitative teams collaborating closely with

business experts: Developers and validators

§ Judicious use of properly governed expert

judgement/qualitative approaches

§ Better understanding of business drivers:

Relationship between pricing, originations and risk drivers

§ Use of component level build-up where feasible § Improvements in aggregation process: integrating

PPNR and loss models implementation in a seamless manner

Total Balance approach Balance/Line growth models Utilization Book Yield Component approach Existing book New Book Contractual amortizations/ yields Prepayment models Originations Yields Prepayment Amortization – Loan characteristics

4

Planning – Benchmarking with models

Model with Internal data Model with industry data

• Internal data is typically more granular but can have limited history – may not capture multiple

economic cycles

• External data may not have desired granularity but can provide multiple economic cycles
• Benchmark models with external data provides useful comparisons with internal trends
• Can be used to assess:
• Consistency of scenario projections
• Difference in sensitivity between internal and market trends
• Similarity/difference in macro economic drivers
• Influence of multiple economic cycles on macro sensitivity

2,000 4,000 6,000 8,000 10,000 12,000 14,000

5

Portfolio optimization illustration

5.35 4.65 3.13 2.93 2.44 2.26 2.18 2.03 1.80 1.50 1.20 1.19 0.93 0.93 1.97 1.89 1 3 5

• 2.0%

0.0% 2.0% 4.0% 6.0% 8.0% 10.0%

Risk Adjusted Yield by Scenario Return-to-Risk Ratio: Base Risk Adjusted Yield/(Base - Stress)

Products/Business Lines

Legend: (Risk Adj. Yield)

Base Adverse Severely Adv

• Risk can be defined as variance from base to stress … combine PPNR and provisions
• Portfolios/products can be ranked on the basis of risk or return-risk
• Provides insights into
• What portfolios provide excess return for the amount of risk taken?
• Are scenarios adequately stressing the material portfolios?
• Is a new product/initiative accretive to the portfolio in terms of risk-return?
• How to optimize the mix of products in a portfolio

Impact of new product

SWOT Analysis – Risks and Opportunities

1

• Strengths
• Weaknesses
• Stress Testing
• Scenario Planning / Systems and People
• Comprehensive Assessments
• Measure ‘Realistic’ Responses
• Risk Appetite/Governance
• Data / Modelling
• Competitive Landscape
• Product Delivery Preferences
• Branch Complements Electronic Delivery

SWOT Analysis – Risks and Opportunities

• Non-Bank Competition (Amazon)
• Rising Rates & Flatter Yield Curve
• Strategies for a Dynamic Marketplace
• Artificial Intelligence and Machine Learning
• Enhanced Analytical Capabilities
• Enhanced Service Capabilities
• Automation to Reduce Operating Costs
• C-suite more Risk-attuned than 10 years ago

2

• Threats
• Opportunities

Capital Allocation

• Reasons for allocating Capital

Incentive

• Scarce capital
• Limited return
• Reward for

efficient use of capital Informed decisions

• Does

product/service generate sufficient return

• n capital
• Does

product/service crowd out another that better uses capital? Maximize Shareholder Value

• Efficient use of

capital builds up return on equity faster

• Higher volatility
• f returns should

demand higher returns

1

Capital Allocation

• Some examples

2

[-2.74, -2.18] (-2.18, -1.62] (-1.62, -1.06] (-1.06, -0.50] (-0.50, 0.06] (0.06, 0.62] (0.62, 1.18] (1.18, 1.74] (1.74, 2.30] (2.30, 2.86] (2.86, 3.42] 10 20 30 40 50 60 70 80

Interest Income Loans

• No results 3 standard deviations below 0
• Clear average around 0 standard deviations

[-3.82, -3.26] (-3.26, -2.70] (-2.70, -2.14] (-2.14, -1.58] (-1.58, -1.02] (-1.02, -0.46] (-0.46, 0.10] (0.10, 0.66] (0.66, 1.22] (1.22, 1.78] (1.78, 2.34] (2.34, 2.90] (2.90, 3.46] 10 20 30 40 50 60 70

Interest Income Trading

• Unclear if average is at 0 standard deviations
• Clear evidence of events outside of 3 standard deviations

Capital Allocation

• Some examples

3

• Clear average around 0 standard deviations
• Noticeable frequency of over 3 standard deviation moves
• Good candidate for Expected Shortfall/Conditional VaR study

[-3.82, -3.26] (-3.26, -2.70] (-2.70, -2.14] (-2.14, -1.58] (-1.58, -1.02] (-1.02, -0.46] (-0.46, 0.10] (0.10, 0.66] (0.66, 1.22] (1.22, 1.78] (1.78, 2.34] (2.34, 2.90] (2.90, 3.46] 10 20 30 40 50 60 70

Interest Income Trading

• Unclear if average is at 0 standard deviations
• Clear evidence of events outside of 3 standard deviations

[-3.23, -2.67] (-2.67, -2.11] (-2.11, -1.55] (-1.55, -0.99] (-0.99, -0.43] (-0.43, 0.13] (0.13, 0.69] (0.69, 1.25] (1.25, 1.81] (1.81, 2.37] (2.37, 2.93] (2.93, 3.49] 10 20 30 40 50 60 70

Trading Fees

Capital Allocation

• What Capital do we use?

Risk Based

• Credit based

exposure

• Market Risk

for large trading activity banks

• Benefits

government, OECD bank,

• ff-balance

sheet exposure Leverage

• Balance

sheet-based exposure

• Captures

Matched book positions

• Benefits off-

balance sheet, fee income businesses Economic

• Unexpected

Loss

• Model-

specific

• Hits the “3

standard deviation” businesses

4

Thank you for attending

You may also be interested in…

The future of stress testing: Regulatory changes, leveraging technology and moving towards automation

View full speaker line-up and agenda at: www.stress-testing-usa.com