## FINANCIAL SECTOR ASSESSMENT PROGRAM UPDATE ISRAEL: STRESS TEST OF THE BANKING, INSURANCE AND PENSION SECTORS — TECHNICAL NOTE (MARCH 2012)

## Source details

**Canonical URL:** [FINANCIAL SECTOR ASSESSMENT PROGRAM UPDATE ISRAEL: STRESS TEST OF THE BANKING, INSURANCE AND PENSION SECTORS — TECHNICAL NOTE (MARCH 2012)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2012/_cr1288.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2012/_cr1288.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2012/_cr1288.pdf.json)

---

### Executive summary — design, scenarios, and headline findings
- Exercise scope:
  - Comprehensive analysis of solvency and liquidity risks of key banking and insurance institutions; stress tests carried out in collaboration with Bank of Israel (BOI) and Capital Markets, Insurance and Savings Division (CMISD) of the Ministry of Finance (MOF).
- Banking tests and methods:
  - Top-down balance sheet stress test and single-factor shocks by BOI BSD; contingent claims analysis (CCA) stress test by BOI and IMF staff.
- Scenarios and horizon:
  - Baseline as of end-June 2011; stress test horizon to end-2014.
  - Three macro scenarios: Base, Adverse 1 (large domestic shock), Adverse 2 (serious international and domestic shock).
- Capital hurdle rates:
  - Total capital adequacy ratio (CAR) hurdle: 9 percent.
  - Core Tier 1 (CT1) hurdle: 5 percent.
- Headline results:
  - Balance sheet stress tests: banks have sufficient buffers and capital remains adequate under Base, Adverse 1, and Adverse 2 scenarios; one bank’s CT1 ratio falls to 6.9 percent under stress.
  - CCA results: projected bank credit spreads under Base of 75 to 210 bps (2010–2014); under Adverse 2 projected spreads increase to levels slightly higher than worst periods of 2008/09 with maximum spread over 300 bps.
  - Under Adverse 2 estimated total potential losses to bank creditors increase to about 1.3 percent of GDP (from about 0.4 percent).
  - Liquidity: all major banks maintain BOI prudential overall liquidity ratio above unity under strong stress scenarios; some banks fail to maintain excess of foreign currency short-term assets over liabilities.
  - Insurance and pension (LTS): shocks on LTS products show manageable effects; insurance business excluding saving products did not expose large vulnerabilities; two companies face challenges meeting new, higher capital requirements.

### Background on sector structure and risk drivers
- Financial structure:
  - Main institutions: banks and insurance companies; active markets in shares, corporate bonds, and government bonds; variety of pension, provident, and mutual funds.
  - Post-Bachar reform (mid-2005): banks divested most non-commercial activities and now focus on traditional banking; non-bank financial sector has grown rapidly.
  - Most institutions have relatively little overseas activity; foreign institutions play minor role.
- Household sector:
  - Households have relatively little mortgage or consumer debt; financial assets have built up strongly.
  - Large share of assets and liabilities indexed or with variable interest rates.
  - Mortgages carry recourse; relatively low LTV ratios for new loans.
  - House prices rose sharply since 2008, leveled off from mid-2011; residential construction and mortgage lending increased rapidly.
- Corporate sector:
  - Recovering from 2008–09 global crisis; profitability not back to pre-crisis levels.
  - Corporate default indicators rising but below 2008–09 levels.
  - Corporate concentration: six largest groups account for about a quarter of GDP.

### Scenarios and selected scenario parameter values (Table 1 key entries)
- Real GDP Growth (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.6; Adv. 1 3.4; Adv. 2 1.9
  - 2012: Base 3.1; Adv. 1 1.2; Adv. 2 -2.8
  - 2013: Base 3.4; Adv. 1 2.5; Adv. 2 1.2
  - 2014: Base 3.7; Adv. 1 2.8; Adv. 2 2.5
- Inflation (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.0; Adv. 1 3.1; Adv. 2 2.8
  - 2012: Base 2.3; Adv. 1 3.2; Adv. 2 0.6
  - 2013: Base 2.2; Adv. 1 2.8; Adv. 2 -1.4
  - 2014: Base 1.8; Adv. 1 3.1; Adv. 2 -4.8
- Exchange Rate Depreciation (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.0; Adv. 1 4.1; Adv. 2 3.6
  - 2012: Base 1.7; Adv. 1 6.8; Adv. 2 12.0
  - 2013: Base 1.3; Adv. 1 3.6; Adv. 2 -1.2
  - 2014: Base -0.4; Adv. 1 1.2; Adv. 2 -8.3
- Bank of Israel interest rate (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.0; Adv. 1 3.1; Adv. 2 2.9
  - 2012: Base 3.0; Adv. 1 3.6; Adv. 2 1.0
  - 2013: Base 2.8; Adv. 1 2.6; Adv. 2 0.5
  - 2014: Base 2.3; Adv. 1 2.1; Adv. 2 0.5
- Unemployment (Base / Adv. 1 / Adv. 2)
  - 2011: Base 5.6; Adv. 1 5.7; Adv. 2 5.9
  - 2012: Base 5.7; Adv. 1 6.5; Adv. 2 9.9
  - 2013: Base 5.9; Adv. 1 7.3; Adv. 2 11.1
  - 2014: Base 6.1; Adv. 1 7.3; Adv. 2 10.8
- TA100 Index changes (Base / Adv. 1 / Adv. 2)
  - 2011: Base -24.9; Adv. 1 -30.7; Adv. 2 -36.4
  - 2012: Base -1.6; Adv. 1 -8.7; Adv. 2 -57.3
  - 2013: Base -0.2; Adv. 1 -25.6; Adv. 2 33.9
  - 2014: Base 7.5; Adv. 1 7.5; Adv. 2 21.7
- Selected bond spread entries (examples preserved):
  - Long term-Short term Yield Spread 2012: Base 2.0; Adv. 1 3.3; Adv. 2 5.6
  - A rated bonds spread 2012: Base 3.4; Adv. 1 6.8; Adv. 2 12.3
  - BBB rated bonds spread 2012: Base 7.8; Adv. 1 16.0; Adv. 2 25.7
  - Real estate sector bonds spread 2012: Base 5.6; Adv. 1 9.6; Adv. 2 15.7

### Bank stress test methodologies and assumptions
- Coverage and data:
  - Focus on largest five banks comprising 95 percent of the system by asset size; tests based on end-June 2011 balance sheets; projection period to end-2014 with outputs for end-2011, end-2012, end-2013, end-2014.
- Methods:
  - Top-down balance sheet stress tests with satellite models for housing, corporate, household non-housing credit, profit components, profit retention, and haircuts for government and financial bonds.
  - Single-factor tests: market shocks (interest rate, exchange rate, stock market) and idiosyncratic credit shocks (largest borrower group, three largest corporate borrowers).
  - Contingent Claims Analysis (CCA) linking market and accounting information to default probabilities, expected losses and fair-value credit spreads.
- Key modeling choices:
  - Dividend pay-out rule: dividends distributed only if ROE > 6 percent and core Tier1 capital ratio > 8 percent (after distribution); dividend rate 35 percent for CT1 above 8 percent but below 8.5 percent; 50 percent for CT1 above 8.5 percent.
  - Profit tax rate: 40 percent.
  - Balance sheet growth assumed zero.
  - Risk-weighted assets change according to standardized approach rules; BOI estimated haircuts on Israeli government debt.

### Balance sheet stress test results (solvency)
- Aggregate outcome:
  - Banks’ capital remains broadly adequate under Base, Adverse 1, and Adverse 2 scenarios.
  - Despite sharp downturn in Adverse 2, most banks remain above hurdle ratios; some banks make losses and much of earnings need retention.
  - One bank’s CT1 capital ratio falls to 6.9 percent under Adverse 2 in 2012.
- Drivers of resilience:
  - Comfortable initial capitalization and profitability.
  - Very low housing default risk due to low LTVs and full recourse mortgages.
  - Favorable starting point for corporate credit losses reflecting recent strong corporate performance.
  - Negligible direct exposure to European sovereign risk.
  - Little change in RWA because banks use standardized approach.

### Single-factor sensitivity analysis (impact as percent of CT1, smallest to largest)
- Credit shock largest borrower group: Impact is 8.5 to 12.6 percent of CT1
- Credit shock largest three borrowers: Impact is 4.3 to 6.6 percent of CT1
- Stock market decline of 25 percent: Impact is 0.3 to 3.8 percent of CT1
- Exchange rate depreciation of 15 percent: Impact is 0.2 to 3.1 percent of CT1
- Interest rate increase: Impact is -5.0 to 5.0 percent of CT1
- Interest rate decrease: Impact is - -4.8 to 4.7 percent of CT1
- European peripheral exposures: Impact is 0.4 to 1.4 percent of CT1
- Interpretation:
  - Concentration risk (largest borrower group and largest three borrowers) produces the largest capital impacts.
  - Exposures to European sovereigns have a negligible impact.

### Contingent Claims Analysis (CCA) results and metrics
- Projected fair-value credit spreads:
  - Base scenario: 75 to 210 bps (2010–2014).
  - Adverse 1: somewhat higher than Base (remaining below 2008 levels).
  - Adverse 2: increases to levels slightly higher than worst periods of 2008/09; maximum spread over 300 bps (300 bps was highest for any bank in 2008).
- Aggregate expected losses to bank creditors (five largest banks):
  - Under Adverse 2 the total expected losses to bank creditors increases to 1.3 to 1.4 percent of GDP (from about 0.4 percent currently).
  - Interpretation: measure of potential contingent government liability to the banking system; low relative to comparable countries.
- Consistency:
  - Balance sheet and CCA tests are consistent in the ranking of bank vulnerability; market capital and regulatory capital shortfalls are different concepts.

### Liquidity stress test results (one-month horizon, BOI prudential ratio)
- Overall liquidity:
  - All major banks maintain BOI prudential ratio above unity under severe stress scenarios.
- Foreign currency liquidity:
  - Some banks would not maintain an excess of foreign currency short-term assets over liabilities under stress.
- Selected quantitative outcomes (based on end-2011 data):
  - Baseline (All currencies) 1.63; Baseline (Foreign currency) 1.59
  - Average (All currencies) 1.62; Average (Foreign currency) 1.49
  - Minimum (All currencies) 1.48; Minimum (Foreign currency) 1.04
  - A 10 percent outflow of short-term deposits — All currencies 1.28; Foreign currency 1.25
    - Average change from baseline 0.35; 0.33
    - Worst change from baseline 0.36; 0.50
  - A 20 percent outflow of non-resident deposits — All currencies 1.53; Foreign currency 1.30
    - Average change from baseline 0.09; 0.27
    - Worst change from baseline 0.12; 0.50
  - A bank's largest interbank claim becomes illiquid — All currencies 1.61; Foreign currency 1.52
    - Average change from baseline 0.02; 0.10
    - Worst change from baseline 0.04; 0.16
  - Short term securities become illiquid — All currencies 1.57; Foreign currency 1.38
    - Average change from baseline 0.05; 0.18
    - Worst change from baseline 0.10; 0.38
  - Memorandum items (percent):
    - Short-term assets/total assets 33.32 (All currencies) and 40.26 (Foreign currency)
    - Short-term foreign currency assets/total short-term assets ... and 22.72 (Foreign currency)
- Primary liquidity vulnerability:
  - Deposit outflows given low reliance on market funding and low securities holdings.

### Insurance and pension (LTS) stress tests — methodology and results
- Coverage and methodology:
  - LTS portfolios of life insurers, pension funds, and provident funds included; non-life and pure risk life tested with market and insurance shocks.
  - Immediate shock on portfolio values; one-year horizon implied by behavioral adjustments.
- Aggregate LTS losses (selected scenarios):
  - Historic 4th Quarter 2008 scenario — Weighted average 7.70 percent loss in LTS.
    - Insurers 9.45 percent; Provident funds 7.47 percent; Pension funds 5.58 percent.
  - Simulated local shock scenario — Weighted average 4.43 percent loss in LTS.
    - Insurers 3.28 percent; Provident funds 5.43 percent; Pension funds 2.55 percent.
- Portfolio composition and hedging:
  - Only about 20 percent of the average LTS portfolio comprised of equities.
  - Weighted average portfolio local/foreign split: 84.8 / 15.2 (In percent).
  - FX exposure: a 20 percent foreign investment position provided an important hedge; FX gain/loss entries show symmetric effects (e.g., FX gain 20% = 3.1 for weighted average portfolio; FX loss 20% = -3.1).
- Insurance non-saving products:
  - Stress tests did not expose unexpected vulnerabilities; four of the top companies can withstand a 10 percent deterioration in claims in main lines in a 4Q2008-type crisis.
  - None of six companies showed a negative solvency position in tested scenarios, but two companies already challenged by new capital requirements show substantial capital deficits in tested scenarios.
- Market risk and QIS-5 adaptation:
  - Market risk is dominant; SCR results show large charges for equity, interest, lapse and morbidity risks.
  - CMISD adaptations to EIOPA QIS-5 noted; parameters needing further study include lapses, tax deferrals, EPIFP treatment, and correlation matrices.
  - Recommendation: further calibration before assigning confidence to QIS-5 style results.

### Integrated systemic and policy implications
- Main vulnerability: concentration risk in corporate exposures (largest borrower group and largest three borrowers).
- Tail risk vigilance: low probability extreme events (regional geopolitical shocks or severe developed-economy crisis) could pose significant challenges.
- Model risk: tests rely on historical macro–risk parameter relationships that may not hold; EDFs may be affected by equity market liquidity.
- Recommended improvements:
  - Enhance solvency and liquidity stress testing procedures.
  - Enhance corporate credit risk stress testing using corporate EDFs and improved macro model linkages.
  - Improve stress testing of credit risk concentration.
  - Increase MOF–BOI collaboration on integrated bank and insurance stress testing to guide micro- and macroprudential policy.
  - Extend liquidity tests beyond a one month horizon and strengthen supervisory/company stress testing for insurers (including combined market and insurance shocks and reverse stress testing).
  - Consider simplifying the pension system to mitigate operational and legal risks and reduce costs.

### Key econometric and CCA technical findings (appendix highlights)
- Econometric links:
  - Panel GLS estimates using quarterly data (1997Q1–2011Q3; unbalanced panel 254 observations for banks after outlier removal).
  - GDP Growth and BOI Interest Rate coefficients significantly lower EDF (Change in EDF: GDP Growth -0.3053*** [0.0772]; BOI Interest Rate -0.3813** [0.1533]).
  - Inflation and TA100 returns also significant in assets return and EDF equations (see Appendix Table 3 coefficients and standard errors for full detail).
- CCA foundations and mappings:
  - Expected loss to creditors decomposed into risk-neutral default probability, LGD, and default-free debt value; fair-value spread s(T) relates to RNDP and LGD by s(T) = –(1/T) ln(1 – RNDP × LGD).
  - Market Price of Risk (MPR) parameter links global Sharpe Ratio and asset-market correlations to RNDP and spreads; Adverse 2 associated with increased MPR.
- CCA outcomes:
  - Weighted average EDF and fair-value spreads projected to end-2014 under each scenario with 2-standard-error prediction intervals; under Adverse 2 total expected losses to creditors 1.3 to 1.4 percent of GDP.

*International Monetary Fund — Monetary and Capital Markets Department; Financial Sector Assessment Program Update, Israel — Stress Test Technical Note (March 2012).*

### 2012. The views expressed in this document are those of the staff team and do not necessarily reflect

### FINANCIAL SECTOR ASSESSMENT PROGRAM UPDATE ISRAEL: STRESS TEST OF THE BANKING, INSURANCE AND PENSION SECTORS — TECHNICAL NOTE (MARCH 2012)

### Executive Summary: overall design and headline findings
- Exercise scope: comprehensive analysis of solvency and liquidity risks of key banking and insurance institutions; stress tests carried out in collaboration with Bank of Israel (BOI) and Capital Markets, Insurance and Savings Division (CMISD) of the Ministry of Finance (MOF).
- Banking stress tests: top-down balance sheet stress test and single-factor shocks by BOI BSD; contingent claims analysis (CCA) stress test by BOI and IMF staff.
- Scenarios and horizon:
  - Baseline as of end-June 2011; stress test horizon to end-2014.
  - Three macro scenarios: Base, Adverse 1 (large domestic shock), Adverse 2 (serious international and domestic shock).
- Capital hurdle rates used:
  - Total capital adequacy ratio (CAR) hurdle: 9 percent.
  - Core Tier 1 (CT1) hurdle: 5 percent.
- Main headline results:
  - Balance sheet stress tests: banks have sufficient buffers and capital remains adequate under Base, Adverse 1, and Adverse 2 scenarios; one bank’s CT1 ratio falls to 6.9 percent under stress.
  - CCA results consistent with balance sheet tests: projected bank credit spreads under Base of 75 to 210 bps (2010–2014); under Adverse 2 projected spreads increase to levels slightly higher than worst periods of 2008/09 with maximum spread over 300 bps (300 bps is the highest level for any bank in 2008).
  - Under Adverse 2 estimated total potential losses to bank creditors increase to about 1.3 percent of GDP (from about 0.4 percent).
  - Liquidity tests: all major banks maintain BOI prudential overall liquidity ratio above unity under strong stress scenarios; some banks fail to maintain excess of foreign currency short-term assets over liabilities.
  - Insurance and pension (long-term savings, LTS) tests: shocks on LTS products show manageable effects; insurance business excluding saving products did not expose large vulnerabilities; two companies face challenges meeting new, higher capital requirements.
- Key drivers of resilience highlighted:
  - Relatively comfortable initial capitalization and profitability.
  - Very low housing default risk due to low loan-to-value (LTV) ratios and recourse mortgages.
  - Favorable starting point for corporate credit losses reflecting recent strong corporate performance.
  - Negligible direct exposure to European sovereign risk.
  - Little change in risk-weighted assets (RWA) because banks use standardized approach to risk weights.
- Main vulnerability identified: concentration risk — single-factor shocks show largest capital impacts from credit shock to each bank’s largest borrower group and to the largest three individual borrowers for several banks.
- Caveats and forward-looking needs:
  - Tail risks from regional geopolitical events or severe external crises could pose challenges.
  - Model risk: stress tests rely on historical macro–risk parameter relationships that may not hold.
  - Recommended improvements: enhanced satellite models using corporate EDFs, improved macro–model linkages, better stress testing of credit risk and concentration, increased MOF–BOI collaboration on integrated bank and insurance stress testing to inform micro and macroprudential policy.

### Background on sector structure and risk drivers
- Financial structure and market:
  - Main institutions: banks and insurance companies; active markets in shares, corporate bonds, and government bonds; variety of pension, provident, and mutual funds.
  - Post-Bachar reform (mid-2005): banks divested most non-commercial activities and now focus on traditional banking; non-bank financial sector has grown rapidly.
  - Most institutions have relatively little overseas activity; foreign institutions play minor role.
  - Banking and insurance sectors are concentrated (figure evidence referenced).
- Household sector:
  - Households have relatively little mortgage or consumer debt; financial assets have built up strongly.
  - Large share of assets and liabilities indexed or with variable interest rates.
  - Mortgages carry recourse; borrower liable after default and foreclosure.
  - Banks maintain close customer relationships with access to borrower income and financials.
  - Relatively low LTV ratios for new loans (distribution categories preserved in source figures).
  - House prices rose sharply since 2008, leveled off from mid-2011; residential construction and mortgage lending increased rapidly.
  - By mid-2009 deterioration in mortgage standards observed; authorities responded with higher capital requirements for housing loans, supplemental reserve requirements, and variable interest rate mortgage limits.
- Corporate sector:
  - Recovering from 2008–09 global crisis; recovery quick but profitability not back to pre-crisis levels.
  - Corporate market default indicators rising but below 2008–09 levels.
  - Corporate sector leverage rose during crisis, declined in 2010, increased again in 2011.
  - Construction and consumer goods sectors have highest estimated default probabilities (Moody’s KMV indicators cited).
  - Corporate concentration: six largest groups account for about a quarter of GDP; these conglomerates increased leverage pre-crisis.
  - Policy response: Prime Minister committee recommendations include prohibiting control of both large financial and real entities and strengthening corporate governance.

### Bank stress test methodologies and assumptions
- Balance sheet stress tests:
  - Data: supervisory data used.
  - Satellite models: cover housing and corporate credit, household non-housing credit, profit components, profit retention behavior, and haircut models for government and financial institution bonds.
  - Single-factor tests: market risk shocks (interest rate, exchange rate, stock market) and idiosyncratic credit shocks from exposures to largest borrower groups and the three largest corporate borrowers.
  - Asset risk weighting: no large changes in RWA because banks under standardized approach.
- Contingent Claims Analysis (CCA):
  - Uses risk-adjusted balance sheets calibrated on market and accounting information to capture relationships between market capital changes, bank assets, and credit risk.
  - Econometric analysis links macro variables to default probabilities and asset price changes; scenarios used to project default probabilities, expected losses, and credit spreads at bank and aggregate levels.
- Liquidity stress tests:
  - Shocks to assets and liabilities applied.
  - Metrics: BOI prudential ratios for overall and foreign currency liquidity.
  - Main liquidity risk source: deposit outflows given low reliance on market funding and low securities holdings.

### Balance sheet stress test results (solvency)
- Overall outcome:
  - Banks’ capital remains broadly adequate under Base, Adverse 1, and Adverse 2 scenarios.
  - Despite sharp downturn in growth under Adverse 2, most banks remain above hurdle ratios; some banks make losses and much of earnings need retention.
  - One bank’s CT1 capital ratio falls to 6.9 percent under stress.
- Reasons for resilience:
  - Comfortable initial capitalization and profitability.
  - Very low housing default risk from low LTVs and full recourse mortgages.
  - Favorable initial corporate credit position from strong corporate performance.
  - Negligible direct exposure to European sovereign risk.
- Single-factor shocks:
  - Largest capital impacts from credit shock to each bank’s largest borrower group.
  - Credit shock to largest three individual borrowers materially impacts several banks.
  - Exposure to European sovereigns very small with negligible impact.

### CCA stress test results (market-implied)
- Projected bank credit spreads:
  - Base scenario: 75 to 210 bps (2010–2014).
  - Adverse 1: somewhat higher than Base.
  - Adverse 2: increases to levels slightly higher than worst periods of 2008/09; maximum spread over 300 bps (300 bps was highest for any bank in 2008).
- Aggregate loss indicator:
  - Estimated total potential losses to bank creditors under Adverse 2 increase to about 1.3 percent of GDP (from about 0.4 percent currently).
  - Interpretation: measure of potential contingent government liability to banking system; low relative to comparable countries.

### Liquidity stress test results
- Overall liquidity (BOI prudential ratio): all major banks maintain ratio above unity under strong stress scenarios.
- Foreign currency liquidity: some banks would not maintain excess of foreign currency short-term assets over liabilities under stress.
- Primary liquidity vulnerability: deposit outflows due to low market funding reliance and limited securities holdings.

### Insurance and pension (long-term savings, LTS) stress tests
- Methodology: differentiates liabilities depending on whether insurance companies or policy holders bear the risks.
- LTS product shocks: effects deemed manageable.
- Insurance business excluding savings: did not reveal large vulnerabilities.
- Dominant risk: market risk.
- Capital adequacy under new requirements: two companies face challenges meeting new, higher capital requirements (Solvency 2 preparedness needs).
- Recommended improvements: additional work on parameter estimation and risk calibration in preparation for Solvency 2.

### Integrated systemic and policy implications
- Tail risk vigilance: low probability extreme events (regional geopolitical shocks or severe developed-economy crisis) would pose significant challenges.
- Model risk: stress tests depend on historical macro–risk parameter relationships which may not be representative of future relationships.
- Recommended continued improvements:
  - Enhance solvency and liquidity stress testing procedures.
  - In corporate credit risk, stress test satellite models using corporate EDFs and enhanced macro model linkages.
  - Improve stress testing of credit risk concentration.
  - Increase MOF and BOI collaboration on integrated bank and insurance stress testing to better understand systemic risk and guide micro and macroprudential policies.

*International Monetary Fund — Monetary and Capital Markets Department; Financial Sector Assessment Program Update, Israel — Stress Test Technical Note (March 2012).*

### 7. The approaches to forecasting the main macroeconomic variables can be

### _cr1288 - 7. The approaches to forecasting the main macroeconomic variables can be

### Approaches to forecasting key macroeconomic variables
- GDP Growth:
  - Baseline scenario values are the output of the DSGE model and fit the baseline scenario of the Research Department Staff Forecast.
  - Values for Adverse scenario 1 and 2 were agreed with the IMF FSAP team, the RD and the BSD.
  - The first 6 quarters of Adverse scenario 2 (2011Q4 – 2013Q1) were based on the growth path observed in the 2008 crisis with the 2 quarters of GDP contraction seen in the 2008-2009 crisis, extended to 4 quarters of contraction.
- Inflation, Exchange Rate Depreciation, BOI Short Interest Rate and Unemployment:
  - These 4 variables are the output of the RD staff forecast after setting the GDP growth path and the relevant shocks (risk premium, global recession).
- TA100 Stock Index:
  - Baseline: RD used a VAR model including domestic (growth, inflation, depreciation, BOI interest rate and changes in TA100) and foreign variables (world growth, inflation, short interest rate, changes in S&P500 and the VIX).
  - Adverse scenario 1: changes observed at the 2001 recession were used (domestic recession).
  - Adverse scenario 2: changes observed during the 2008 crisis were used (aligned with GDP exercise).
- Long-Term and Short-Term Yield Spread:
  - Projections based on historical data.
  - Baseline long-term yield on government bonds set at an average rate (5 percent).
  - Adverse scenario 1: historical spreads observed at the 2001 recession used; assumed long-term yield would converge towards 5 percent in the last 6 quarters.
  - Adverse scenario 2: historical spreads observed at the 2008 crisis used as guide.
  - Note: Bond ratings are Israeli ratings, not international ratings.

### Table 1 — Israel: Stress Testing Scenario Parameters (selected scenario values)
- Real GDP Growth (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.6; Adv. 1 3.4; Adv. 2 1.9
  - 2012: Base 3.1; Adv. 1 1.2; Adv. 2 -2.8
  - 2013: Base 3.4; Adv. 1 2.5; Adv. 2 1.2
  - 2014: Base 3.7; Adv. 1 2.8; Adv. 2 2.5
- Inflation (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.0; Adv. 1 3.1; Adv. 2 2.8
  - 2012: Base 2.3; Adv. 1 3.2; Adv. 2 0.6
  - 2013: Base 2.2; Adv. 1 2.8; Adv. 2 -1.4
  - 2014: Base 1.8; Adv. 1 3.1; Adv. 2 -4.8
- Exchange Rate Depreciation (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.0; Adv. 1 4.1; Adv. 2 3.6
  - 2012: Base 1.7; Adv. 1 6.8; Adv. 2 12.0
  - 2013: Base 1.3; Adv. 1 3.6; Adv. 2 -1.2
  - 2014: Base -0.4; Adv. 1 1.2; Adv. 2 -8.3
- Bank of Israel interest rate (Base / Adv. 1 / Adv. 2)
  - 2011: Base 3.0; Adv. 1 3.1; Adv. 2 2.9
  - 2012: Base 3.0; Adv. 1 3.6; Adv. 2 1.0
  - 2013: Base 2.8; Adv. 1 2.6; Adv. 2 0.5
  - 2014: Base 2.3; Adv. 1 2.1; Adv. 2 0.5
- Unemployment (Base / Adv. 1 / Adv. 2)
  - 2011: Base 5.6; Adv. 1 5.7; Adv. 2 5.9
  - 2012: Base 5.7; Adv. 1 6.5; Adv. 2 9.9
  - 2013: Base 5.9; Adv. 1 7.3; Adv. 2 11.1
  - 2014: Base 6.1; Adv. 1 7.3; Adv. 2 10.8
- Change in TA100 Index (Base / Adv. 1 / Adv. 2)
  - 2011: Base -24.9; Adv. 1 -30.7; Adv. 2 -36.4
  - 2012: Base -1.6; Adv. 1 -8.7; Adv. 2 -57.3
  - 2013: Base -0.2; Adv. 1 -25.6; Adv. 2 33.9
  - 2014: Base 7.5; Adv. 1 7.5; Adv. 2 21.7
- Long term-Short term Yield Spread (Base / Adv. 1 / Adv. 2)
  - 2011: Base 1.9; Adv. 1 2.2; Adv. 2 2.2
  - 2012: Base 2.0; Adv. 1 3.3; Adv. 2 5.6
  - 2013: Base 2.2; Adv. 1 3.2; Adv. 2 4.7
  - 2014: Base 2.7; Adv. 1 3.0; Adv. 2 4.1
- AA rated bonds spread (Base / Adv. 1 / Adv. 2)
  - 2011: Base 1.1; Adv. 1 1.1; Adv. 2 1.3
  - 2012: Base 1.2; Adv. 1 2.2; Adv. 2 3.1
  - 2013: Base 1.2; Adv. 1 1.7; Adv. 2 2.2
  - 2014: Base 1.2; Adv. 1 1.7; Adv. 2 1.7
- A rated bonds spread (Base / Adv. 1 / Adv. 2)
  - 2011: Base 2.7; Adv. 1 2.7; Adv. 2 3.3
  - 2012: Base 3.4; Adv. 1 6.8; Adv. 2 12.3
  - 2013: Base 3.4; Adv. 1 4.8; Adv. 2 6.8
  - 2014: Base 3.4; Adv. 1 4.8; Adv. 2 4.8
- BBB rated bonds spread (Base / Adv. 1 / Adv. 2)
  - 2011: Base 6.1; Adv. 1 6.1; Adv. 2 7.7
  - 2012: Base 7.8; Adv. 1 16.0; Adv. 2 25.7
  - 2013: Base 7.8; Adv. 1 12.5; Adv. 2 16.0
  - 2014: Base 7.8; Adv. 1 12.5; Adv. 2 12.5
- Non-rated bonds spread (Base / Adv. 1 / Adv. 2)
  - 2011: Base 7.7; Adv. 1 7.7; Adv. 2 9.7
  - 2012: Base 8.9; Adv. 1 17.3; Adv. 2 22.7
  - 2013: Base 8.9; Adv. 1 16.5; Adv. 2 17.3
  - 2014: Base 8.9; Adv. 1 16.5; Adv. 2 16.5
- Real estate sector bonds spread (Base / Adv. 1 / Adv. 2)
  - 2011: Base 4.7; Adv. 1 4.7; Adv. 2 5.2
  - 2012: Base 5.6; Adv. 1 9.6; Adv. 2 15.7
  - 2013: Base 5.6; Adv. 1 9.0; Adv. 2 9.6
  - 2014: Base 5.6; Adv. 1 9.0; Adv. 2 9.0

*(Source: Bank of Israel)*

### BSD stress testing framework and assumptions
- Coverage:
  - Stress test exercise focused on the largest five banks in Israel, comprising 95 percent of the system by asset size (Bank Leumi Le-Israel, Bank Hapoalim, Israel Discount Bank, Mizrahi Tefahot Bank, First International Bank of Israel).
- Timing and data:
  - Tests based on end-June 2011 balance sheets; projection period until end-2014 with outputs for end-2011, end-2012, end-2013, and end-2014.
- Methods:
  - Top-down balance sheet stress tests and single factor tests using BSD supervisory methods and models.
  - Hurdle rates: total capital adequacy ratio (CAR) 9 percent; CT1 5 percent.
  - Profitability also used as a metric.
- Key modeling choices:
  - Satellite models cover housing and corporate credit, household non-housing credit, profit components, profit retention behavior, and haircut models for government and foreign financial institution bonds.
  - Single factor tests for market risk (interest rate, exchange rate, stock market shocks) and idiosyncratic credit shocks (largest borrower group, three largest corporate borrowers).
  - Dividend pay-out rule:
    - Dividends distributed only if ROE > 6 percent and core Tier1 capital ratio > 8 percent (after distribution).
    - Dividend rate 35 percent for core Tier1 capital ratio above 8 percent but below 8.5 percent.
    - Dividend rate 50 percent for core Tier1 capital ratio above 8.5 percent.
  - Profit tax rate: 40 percent.
  - Balance sheet growth assumed zero.
  - Risk-weighted assets change according to standardized approach rules; estimated impact of exchange rate changes on RWA incorporated.
  - BOI staff estimated haircuts on Israeli government debt; small holdings of foreign sovereign and European periphery sovereign debt are negligible.

### Main satellite models and calibration
- Profit model uses three satellite models:
  - Net interest income.
  - Operating and other income.
  - Operating and other expenses.
- Housing credit losses:
  - Loan loss provisions as a function of unemployment, BOI interest rate and lagged loan loss provisions.
- Household non-housing credit:
  - Loan loss provisions as a function of unemployment and changes in financial assets of households driven primarily by changes in the Tel Aviv 100 index.
- Corporate sector losses:
  - Based on a "quasi-probability of default" (quasi-PD): number of firms with loan-loss provision divided by number of active firms; linked to macro variables including the Composite Index, real interest rates, and changes in the TA100.
  - Losses for corporate sector in Adverse scenario 1 and 2 based on 99th percentile of the quasi-PD distribution.
  - Loss given default (LGD) assumption: 0.45.
  - A calibration factor (ratio between average provisions and average losses) is applied to get final loss rates.

### Results of top-down balance sheet stress tests
- Aggregate outcome:
  - Tests show banks have sufficient buffers and banks’ capital remains adequately capitalized under the Base, Adverse 1 and Adverse 2 scenarios.
  - Despite sharp downturn in Adverse scenario 2, banks’ capital ratios remain above hurdle rates; some banks make losses and much earnings need to be retained.
  - Under Adverse scenario 2, one bank had a CT1 of 6.9 percent in 2012.
- Reasons for resilience:
  - Relatively comfortable initial capitalization and profitability.
  - Very low housing default risk due to low LTVs and mortgages carrying recourse.
  - Favorable starting point for corporate credit losses reflecting recent strong corporate performance.
  - Negligible exposure to European sovereigns and limited bank risks.
  - No large changes in RWA because banks are under the standardized approach.
  - Small reduction in RWA due to exchange rate effects under standardized approach rules.

### Single factor test findings (sensitivity analysis results — smallest to largest impact as percent CT1)
- Credit shock largest borrower group: Impact is 8.5 to 12.6 percent of CT1
- Credit shock largest three borrowers: Impact is 4.3 to 6.6 percent of CT1
- Stock market decline of 25 percent: Impact is 0.3 to 3.8 percent of CT1
- Exchange rate depreciation of 15 percent: Impact is 0.2 to 3.1 percent of CT1
- Interest rate increase: Impact is -5.0 to 5.0 percent of CT1
- Interest rate decrease: Impact is - -4.8 to 4.7 percent of CT1
- European peripheral exposures: Impact is 0.4 to 1.4 percent of CT1
- Key interpretation:
  - Concentration risk (largest borrower group and largest three borrowers) has the largest potential impact on capital.
  - Exposures to European sovereigns have a negligible impact.

### Methodological caveats and recommended improvements
- Caveats:
  - Balance sheet stress tests rely on historical relationships of macro data to risk parameters which may not represent future relationships (model risk).
- Suggested improvements:
  - Continue improvements in stress testing procedures.
  - Enhance corporate credit risk stress testing using corporate sector EDFs and improved macro model linkages.
  - Improve stress testing of credit risk and concentration risk.

### Contingent Claims Analysis (CCA) framework overview
- Concept:
  - CCA is a risk-adjusted balance sheet concept linking bank asset values to equity value, default risk and bank funding costs.
  - Total market value of bank assets equals market value of equity plus "risky debt" over a specific time horizon.
  - Risky debt = default-free value of debt and deposits minus the "expected loss to bank creditors" from default over the time horizon.
- Expected loss to bank creditors:
  - Viewed as probability of bank default times loss given default times exposure (default-free value of total debt and deposits).
  - In risk-adjusted balance sheet, a decline in asset value leads to less than one-to-one decline in market equity; expected losses to creditors increase, possibly substantially under severe distress.
- Relation to accounting balance sheet:
  - Accounting balance sheet is a special case of the risk-adjusted balance sheet with zero uncertainty (zero asset volatility), where expected loss to creditors is zero and equity becomes book equity.
- MKMV (KMV/Moody's) CCA model:
  - MKMV EDFs calculated by iterating implied asset volatility; EDFs converted to risk-neutral default probabilities, then combined with sector LGD to produce Fair-Value Credit Default Swap spreads (FVCDS).
  - MKMV estimates EDFs and CCA parameters daily for 88 financial institutions and 220 corporates in Israel; model described as a leading indicator for default.

*Source: BOI, RD, and BSD content in the provided chapter.*

### 23. The risk-adjusted (CCA) balance sheet of the banks quantifies relationship of

### _cr1288 - 23. The risk-adjusted (CCA) balance sheet of the banks quantifies relationship of

### CCA approach — purpose and inputs
- Quantifies relationship of market capital level, default probability and bank credit spreads, as well as the impact of changes in global risk appetite.
- Lower levels of the market value of equity (market capitalization) are directly related to higher bank default probabilities.
- The impact of changes in global or regional risk appetite on:
  - bank expected losses to creditors,
  - bank funding costs, and
  - bank equity
  can be measured in the CCA approach.
- MKMV uses a version of CCA models and provides daily estimates of EDFs and CCA parameters for all major Israeli banks and insurance companies (and corporate).
- Box 1 described the MKMV data used as input to the stress test exercise.

### Linking macro variables to financial institutions' CCA outputs — methodology
- Focus: largest five banks in Israel (Bank Leumi Le-Israel, Bank Hapoalim, Israel Discount Bank, Mizrahi Tefahot Bank, First International Bank of Israel), comprising 95 percent of the system by asset size.
- Econometric model (institution-level):
  - General form: ittiiit Xyεβα ++=
  - Panel-estimated form (used because individual samples are small):
    - ittiiitiiit XIXyεδγαβ +++++=*
    - where iγ is a fixed effect for each institution and tiXI* is an interaction term between the institution and macro variables to capture different effects on each institution's it y.
- Decision on which macro variables to include in X was made by trying different specifications and using judgment.

### Data and macro variables used for estimation
- Sample: quarterly observations; quarterly average of the EDF and the assets of each institution.
- Sample size: minimum 34 observations and maximum 59 observations (1997Q1-2011Q3).
- Macro variables considered:
  - GDP growth (annual rates, difference from steady state)
  - inflation (annual rates, difference from steady state)
  - depreciation of effective foreign exchange (annual rates, difference from steady state)
  - Bank of Israel short interest rate (annual rates, difference from steady state)
  - TA100 return (annual rates, difference from steady state)
  - yield on non-indexed long government bonds (non Indexed, 8-10 years, annual rates) or the gap between this yield and the short interest rate
- Note: when including long term interest rate variables the estimation period is shorter since the long term interest rate is available only from mid-2001.

### Estimation results — key findings
- Banks Change in EDF:
  - An increase in GDP and increase in the BOI interest rate lowers the EDF. Coefficients are significant at the 1 percent level.
  - Exchange rate depreciation raises the EDF.
  - The higher EDF caused by higher long term interest rate needs further investigation; the effect is not economically significant.
  - TA100 returns affect the institutions differently.
- Banks Assets Returns:
  - All macro variables affect assets return positively and significantly (FX effect is lagged).
  - Positive and significant idiosyncratic effects are observed only in two banks.
  - Various specifications showed no different effect of GDP growth and/or TA100 returns across banks.
- Overall:
  - Many coefficients are significant at the 1 and 5 percent level.
  - Results broadly make sense in direction; additional specifications and further analysis would be useful.

### Projections and transformation to spreads
- Using the macro econometric model the EDFs and asset returns were projected for all five banks until the end of 2014, for all three stress scenarios.
- A 2-standard-error-of-the-prediction confidence interval was calculated for each institution forecast.
- The EDFs were transformed into fair value credit spreads (fair value spread = estimate of the bank CDS spread, absent liquidity premiums and distortions from government guarantees).

### CCA stress test results and metrics
- Fair value spread history and projected ranges:
  - Fair value spread at the end of 2008 ranged from 150 to 300 basis points (bps) and declined to below 110 bps in 2010.
  - Under the Base scenario, spreads increased somewhat to between 75 to 210 bps from 2010 through 2014.
  - Under the Adverse 1 scenario spreads were at a slightly higher level (remaining below 2008 levels).
  - Under the Adverse 2 scenario spreads increase to a higher level than during the worst periods of the financial crisis in 2008/09; the maximum spread is over 300 bps (300 bps is the highest level for any bank in 2008).
- Total expected losses to bank creditors (five largest banks) as percent of GDP:
  - Under the Adverse 2 scenario the total expected losses to bank creditors increases to 1.3 to 1.4 percent of GDP.
  - Interpretation: this can be viewed as a measure of the potential contingent liability of the government to the banking system.
- Consistency:
  - Market capital and regulatory capital shortfalls are different concepts and cannot be compared directly.
  - The balance sheet and CCA tests are consistent with each other in the ranking of bank vulnerability.
- Appendix IV contains more detailed outputs on past and projected EDFs and spreads.

### Caveats and recommended improvements for CCA stress tests
- CCA stress tests rely on historical relationships of macro data to EDF and asset return; such relationships may not be representative of future relationships (model risk).
- EDF estimates incorporate forward looking equity information but may be affected by less than liquid equity markets or over/under shooting.
- Going forward improvements and refinements in CCA stress testing procedures should be continued.

### Liquidity stress tests — brief summary (transitioning content)
- Liquidity risk stress tests focused on the change in short-term assets and liabilities based on the BOI’s supervisory model using a one month horizon based on end-2011 data.
- Tests carried out by BSD for four severe stress scenarios, by total currency positions and foreign currency positions separately.
- Aim: ensure a bank maintains unencumbered and high quality liquid assets above expected liquidity needs for a one month horizon (liquidity ratio exceeds unity) even under severe stress scenarios.
- Results:
  - All banks would be able to maintain sufficient liquid assets under strong stress scenarios (Table 2).
  - For foreign currency liquidity positions alone, some banks would not be able to maintain an excess of foreign currency short-term assets over liabilities.
  - Deposit outflows are potentially the main risk to maintaining liquidity.
  - Recommendation: improvements in liquidity stress testing practices, including extensions beyond a one month horizon, should be continued.
- Selected quantitative outcomes (based on end-2011 data):
  - Baseline (All currencies) 1.63; Baseline (Foreign currency) 1.59
  - Average (All currencies) 1.62; Average (Foreign currency) 1.49
  - Minimum (All currencies) 1.48; Minimum (Foreign currency) 1.04
  - A 10 percent outflow of short-term deposits — All currencies 1.28; Foreign currency 1.25
    - Average change from baseline 0.35; 0.33
    - Worst change from baseline 0.36; 0.50
  - A 20 percent outflow of non-resident deposits — All currencies 1.53; Foreign currency 1.30
    - Average change from baseline 0.09; 0.27
    - Worst change from baseline 0.12; 0.50
  - A bank's largest interbank claim becomes illiquid — All currencies 1.61; Foreign currency 1.52
    - Average change from baseline 0.02; 0.10
    - Worst change from baseline 0.04; 0.16
  - Short term securities become illiquid/ — All currencies 1.57; Foreign currency 1.38
    - Average change from baseline 0.05; 0.18
    - Worst change from baseline 0.10; 0.38
  - Memorandum items (percent):
    - Short-term assets/total assets 33.32 (All currencies) and 40.26 (Foreign currency)
    - Short-term foreign currency assets/total short-term assets ... and 22.72 (Foreign currency)
  - Note: Baseline excludes Israeli treasury bills.

*Source: BOI, and staff estimates.*

### 45. A series of dedicated stress tests were carried out in close cooperation with the

### 45. A series of dedicated stress tests were carried out in close cooperation with the authorities

### Stress testing scope and approach
- Stress tests covered:
  - LTS portfolios of life insurers, pension funds, and provident fund excluding the guaranteed return government bonds deemed risk free.
  - Non-life and pure risk life business, with shocks including market risks and insurance risks impacting claims ratios in different lines of business.
  - Two top-down stress tests by authorities and additional tests conducted by the insurance market under CMISD (similar to the QIS-5 exercise led by EIOPA) and companies’ own stress tests.
- Tests assessed market, insurance, operational, and legal risks; combined market and insurance risks were recommended for revealing hidden vulnerabilities.
- Reference to stress matrix and Appendix V for details on the severe local shock scenario. (Footnote 10 in source.)

### Long Term Savings (LTS) — Results and interpretation
- Aggregate finding:
  - Market shocks similar to the 4th quarter of 2008 and a simulated severe local shock would result in a 7.7 percent loss on the long term savings, individual portfolios, on average.
- Portfolio composition (top-left panel, Figure 7):
  - Only about 20 percent of the average LTS portfolio was comprised of equities.
  - Weighted average portfolio local/foreign split: 84.8 / 15.2 (In percent).
  - Participating policies local/foreign: 79.4 / 20.6 (In percent).
  - Provident funds local/foreign: 88.2 / 11.8 (In percent).
  - New pension funds local/foreign: 82.1 / 17.9 (In percent).
- Scenario-specific changes in value of LTS (Change in value of LTS, in percent; selected entries preserved exactly):
  - Weighted average portfolio: RF drop 20% = 1.5; RF increment 20% = -1.5; Spread +50 bps = -0.4; Spread +100 bps = -0.8; Spread +200 bps = -1.7; Equity drop 20% = -3.3; Equity drop 30% = -5.0; FX gain 20% = 3.1; FX loss 20% = -3.1; Stress test scenario Q4 2008 = -4.3; Local Shock Stress Test Scenario = -7.7.
  - Participating policies: Stress test scenario Q4 2008 = -3.3; Local Shock Stress Test Scenario = -9.4.
  - Provident funds: Stress test scenario Q4 2008 = -5.4; Local Shock Stress Test Scenario = -7.5.
  - New pension funds: Stress test scenario Q4 2008 = -2.5; Local Shock Stress Test Scenario = -5.8.
- Interpretation and policy relevance:
  - Results suggest resilient LTS portfolios; the limited equity allocation significantly shielded portfolios from equity market declines.
  - Findings appear to dismiss possible activation of the government guarantee (stop loss after a 50 percent loss in value) for pensions close to retirement under tested scenarios.
  - Differences in impact range from 10 percent to 6 percent depending on the provider of the LTS; anticipated to diminish as harmonizing regulation takes effect.
  - A 20 percent foreign investment position provided an important hedge.
  - Main risks appear shifted from market risk to operational and legal risks due to product flexibility and mobility, variety of collective agreements on management fees, and regulatory risk from changing market and economic conditions.
  - Recommendation: Simplifying the pension system should be considered as mitigation for operational and legal risk and to reduce costs.

### Insurance business (non-saving products) — Results and interpretation
- Aggregate result:
  - Stress tests on nonlife and pure risk life products did not expose unexpected vulnerabilities.
  - Four of the top companies are sufficiently capitalized to withstand with ease 10 percent deterioration in their claims in main lines of business in the midst of a 4th quarter 2008 similar crisis.
- Capital and solvency outcomes (Figure 8; Change in insurers’ capital surplus in percentage points; entries preserved exactly):
  - Company A: scenarios include values -0.9 -0.3 -0.4 -1.2 -1.6 1.5 0.6 1.2 1.2 0.9 0.3 0.0 (as listed).
  - Company B: scenarios include values -0.4 -0.5 -0.6 -0.8 -1.4 -4.0 -4.4 -4.5 -4.6 -5.1 -4.9 -5.4 (as listed).
  - Company C: scenarios include values -0.6 -0.4 -0.6 -1.1 -1.6 -10.9 -11.5 -11.3 -11.5 -11.9 -11.9 -12.5 (as listed).
  - Company D: scenarios include values -1.3 -0.8 -1.3 -2.1 -3.4 -3.5 -4.9 -4.3 -4.8 -5.6 -5.6 -6.9 (as listed).
  - Company E: scenarios include values -0.7 -0.7 -0.8 -1.4 -2.2 -2.8 -3.5 -3.4 -3.5 -4.2 -4.2 -5.0 (as listed).
  - Company F: scenarios include values -0.2 -1.5 -1.8 -1.7 -3.5 -9.6 -9.7 -11.1 -11.4 -12.9 -11.3 -13.1 (as listed).
- Observations:
  - None of the six companies showed a negative solvency position in the tested scenarios (bottom panel, Figure 8).
  - Two companies, already challenged to meet the new capital requirements, show substantial capital deficits in the tested scenarios.
  - Market risk remains the major risk; in some companies deterioration of claims can positively impact capital position due to reserve release and tax credits.

### QIS-5 adaptation, calibration needs, and supervisory practices
- QIS-5 exercise:
  - Market-wide adapted QIS-5 exercise carried out by whole market within 2 months.
  - CMISD adaptations to EIOPA QIS 5 parameters included straightforward items: definition of risk free interest discount curve, treatment of government bonds and local equities.
  - Parameters needing further study: lapses, tax deferrals, treatment of EPIFP, appropriate correlation matrices for composite companies; insurance risks require further calibration.
- Results summary:
  - Small decrease in asset value, confirming current practice of market-consistent asset valuation.
  - Resulting liabilities experienced a reduction, indicating some conservatism in current solvency regime.
  - Assumptions on liquidity premium, lapses and extrapolation of the risk free interest curve may have influenced results.
  - SCR results exhibit large charges for equity, interest, lapse and morbidity risks — suggesting need for proper calibration before assigning confidence to results.
- Supervisory/company stress testing:
  - Companies’ stress tests (required by authorities) combine historical events and single shocks; 2002 and 2008 crises are selected scenarios, complemented with market shocks on equity, spreads, yield curves, sovereign risk and combinations.
  - Reported results show single digit losses in technical provisions and in extreme cases impact on capital in the order of 25 percent.
  - Recommendation: Supervisory stress tests run by companies need to be more stringent; combine insurance and market risks in scenarios; add insurance shocks to historical scenarios (e.g., earthquake during a 2002 crisis scenario); introduce reverse stress tests.
  - Recommendation: Design a normalized industry wide stress test based on inputs from companies’ own stress testing to allow comparison of solvency across companies and detection of systemic risks.

### Integrated systemic risk stress testing — recommendations
- Coordination and model improvements:
  - Stress testing analyses for banks, insurance companies and other sectors should be more closely integrated to improve systemic risk analysis.
  - CMISD needs to work more closely with the BOI on insurance stress testing (e.g., improving satellite models to include macro-financial variables and devising consistent scenarios).
  - Stress testing of systemic risk should be developed to capture systemic risk dependence between institutions (e.g., through the systemic CCA).
- Policy guidance:
  - Stress testing activities should be designed to guide micro- and macroprudential policy to enhance financial stability.
  - Integrated stress testing should be expanded to analyze macroprudential policies including household credit risk, contingent liability risks, risk transfer between sectors (e.g., corporate, insurance companies, groups, and other sectors).
  - Stress testing can analyze potential benefits (i.e., reduced vulnerability from lower default risk and implied credit spreads) of possible new microprudential and macroprudential measures.
  - The special macroprudential unit can benefit from integrated systemic risk stress testing to analyze implications of macroprudential policies.

### Appendix highlights — Banking stress test matrix and satellite models
- Top-Down Balance Sheet ST and Top-down Contingent Claims Analysis ST covered the 5 largest institutions:
  - Market share: 93 percent of assets; 95 percent of total sector lending (both columns).
  - Data and baseline date: Supervisory data, End-June 2011; consolidated banking group (Top-Down Balance Sheet ST). Balance sheet data (public) plus market data (MKMV inputs), End-June 2011 for CCA ST.
  - Stress test horizon: End-June 2011 to end 2014 (both).
  - Scenario analysis: Base, Adverse 1 (domestic shock), Adverse 2 (serious international shock).
    - Adverse 1: 1.9 percent GDP decline from Base.
    - Adverse 2: 5.9 percent GDP decline from Base.
    - Base scenario GDP growth of 3.1 percent in 2012 (see Table 1 in source).
- Sensitivity analysis (single-factor shocks listed exactly):
  - Credit shock to largest corporate group and largest three borrowers.
  - 25 percent decline in stock market value.
  - 15 percent change in exchange rate.
  - 200 basis point change in interest rates.
  - Vulnerable European government debt shock: 30 and 10 percent write off.
- Results (select findings preserved exactly):
  - CAR for all banks is above 9 percent hurdle rate.
  - CT1 for all banks is above 5 percent threshold rate.
  - One bank has CT1 of 6.9 percent under Adv 2 scenario in 2012.
  - CCA capital ratios, expected losses and implied spreads are higher than 2008/2009 crisis under Adverse 2.
  - If a threshold of 300 bps for the implied credit spread is used, spreads for two banks go above or are very close to this threshold in 2012.
- Sensitivity analysis – Single Factor (smallest to largest impact as percent CT1; preserved entries):
  - Credit shock largest borrower group: Impact is 8.5 to 12.6 percent of CT1.
  - Credit shock largest three borrowers: Impact is 4.3 to 6.6 percent of CT1.
  - Stock market decline of 25 percent: Impact is 0.3 to 3.8 percent of CT1.
  - Exchange rate depreciation of 15 percent: Impact is 0.2 to 3.1 percent of CT1.
  - Interest rate increase: Impact is -5.0 to 5.0 percent of CT1.
  - Interest rate decrease: Impact – 4.8 to 4.7 percent of CT1.
  - European peripheral exposures: Impact is 0.4 to 1.4 percent of CT1.
- Satellite models and key model features:
  - Profits modeled via three satellite OLS models for NII (Net Interest Income), OI (Operating Income and Other Income), OE (Operating and Other Expenses). Dependent variable in these OLS models is the ratio of the profit components to total assets.
  - Models include a 2008 crisis dummy (D). Sample used: 1997Q1-2010Q4 (aggregate results of five major banks; adjustments applied to fit individual banks).
  - Housing credit losses: linear regression estimating loan-loss provision to housing credit with macro factors unemployment rate (U) and BOI interest rate (IR). D is 2006 Q1 dummy. Sample: 2002Q1-2010Q4.
  - Households other credit losses: linear regression using unemployment rate and changes in financial portfolio balance of the public (ASSETS). Sample: 2002Q1-2010Q4.
  - Corporate losses: Wilson (1997) based approach; default rate modeled with logistic function and logit transformation estimated using macro variables (Composite Index changes approximated by GDP growth, real interest rate R_IR, changes in TA100). Monte Carlo simulation used to determine loss distribution of corporate credit portfolio. Sample: 1997Q1-2010Q4.
  - Losses of the business sector in Scenario 1 and Scenario 2 were based on 99th percentile of the quasi PD distribution.

*Source: Israeli authorities, and staff estimates.*

### Appendix Figure 10 Israel: Quarterly Changes in the Composite Index and GDP

### Appendix Figure 10 Israel: Quarterly Changes in the Composite Index and GDP

### Appendix figure and observable series
- Figure plots quarterly changes in the Composite Index (GDPCI) and GDP for the period 1998–2010 on a vertical scale from -4 to 4.

### Sovereign and Foreign Financial Institution Haircuts
- Adverse Scenario 1:
  - Uses historical spreads observed at the 2001 recession.
  - Assumes convergence of the long term yield towards the above mentioned average rate in the last 6 quarters of the scenario.
- Adverse Scenario 2:
  - Uses historical spreads observed at the 2008 crisis.
  - Extends the highest long interest rate for a year (2012).
  - Since the economy doesn't recover and the recession is long, the spread declines slowly and in the last 5 quarters is set equal to the level observed at 2009Q4 (after the crisis peak and in the first stages of recovery, when BOI interest rate was still very low).
- Method to obtain long term yield:
  - Projected spreads are added to the BOI interest rates to get the long term yield.
- Haircut calculation:
  - Israeli sovereign bonds haircut = quarterly changes in these yields multiplied by the average duration (about 2 year).
  - Financial institutions haircuts:
    - Based on the FINRA/Bloomberg Investment Grade U.S. Corporate Bond Yield Index.
    - RD took the gap between the index and 5YR US bond yield.
    - For base and scenario 1 (no global crisis) no increase in yields and haircuts for financial institutions bonds was assumed.
    - For scenario 2 the path is based on a rising yield towards the level observed in 2008Q4 and 2009Q1 and then trending to the level observed in 2011Q3.
    - Financial institution haircut = quarterly changes in these yields multiplied by the average duration (2 year).

*Source of BOI credit exposure reference: BOI Report on Credit Exposure.*

### Supervisory Model for Liquidity Risk Assessment – Asset/liabilities with a maturity of up to one month ratio
- Definition:
  - Ratio = liquid assets / liquid liabilities (both referring to items with maturity up to one month).
- Assets (components and recognition weights):
  - Cash and deposits with the Bank of Israel minus the liquidity requirement
  - plus deposits with the Treasury
  - plus deposits in banks minus intergroup bank deposits
  - plus 94 percent of the total cash flow from Israeli government bonds and Bank of Israel instruments ("Makam")
  - minus the total cash flow of pledged assets in favor of the clearing house and stock exchange
  - plus 80 percent of the total cash flow of foreign sovereign bonds graded AA- and higher
  - plus a 50 percent factor for foreign sovereign bonds graded BBB- up to AA-
  - plus 50 percent of the total cash flow of other corporate bonds graded A- and higher
  - plus borrowed or purchased securities in Repo agreements
  - plus mortgage loans
  - plus 80 percent of other loans
  - plus loans to the Israeli government
  - plus 95 percent of assets from derivative instruments after applying all conversion factors
  - plus 85 percent of other assets.
- Liabilities (components and recognition weights):
  - Deposits from the Bank of Israel
  - plus deposits from banks minus intergroup bank deposits
  - plus 20 percent of the gap between deposits from the public and large deposits
  - plus 50 percent of large deposits
  - plus Israeli government deposits
  - plus securities borrowed or purchased in a Repo agreement
  - plus bonds and subordinate debenture instruments
  - plus liabilities from derivative instruments
  - plus liabilities
  - plus 10 percent of all off-balance sheet items after applying all conversion factors.
- Data source: BOI Liquidity risk system in the ZPTM application, Charts 1-4.

### Adjusted standard model
- Metric: Liquid asset / total liabilities with a maturity of up to one month.
- Adjusted Assets:
  - Cash and deposits with the Bank of Israel minus the liquidity requirement
  - plus deposits with the treasury
  - plus deposits in banks minus intergroup bank deposits
  - plus 94 percent of the total cash flow from Israeli government bonds and Bank of Israel instruments ("Makam")
  - minus the total cash flow from pledged assets in favor of the clearing house and stock exchange.
- Adjusted Liabilities:
  - Total liabilities with a maturity of up to one month.
- Rationale for adjustment:
  - BOI "Proper Conduct of Banking Business" directives No. 342 use a conservative definition; adjustment allows recognizing 94 percent of Israeli government bonds held, similar to haircut applied by Capital Markets Division of the BOI when used as collateral for a monetary loan.

### Appendix III. Contingent Claims Analysis (CCA) — core principles and model representation
- CCA foundations:
  - Generalization of option pricing theory (Black-Scholes (1973), Merton (1973)).
  - Builds risk-adjusted balance sheets from three principles:
    - (i) liabilities (equity and debt) values derived from assets;
    - (ii) liabilities have different priority (senior and junior claims);
    - (iii) assets follow a stochastic process.
- Definitions and relationships:
  - Total market value of assets A(t) = equity market value E + risky debt D.
  - Default barrier B = present value of promised payments on debt; default occurs when A < B.
  - Equity value E = implicit call option on assets with exercise price B.
  - Expected loss to creditors = implicit put option P on assets with exercise price B.
- Key analytic expressions (as presented):
  - Equity value (call option):
    - E(t) = A(t) N(d1) – Be–rT N(d2)
    - d1 = [ln(A/B) + (r + 1/2 σ^2) T] / (σ sqrt(T))
    - d2 = d1 – σ sqrt(T)
    - where r is the risk-free rate, σ is asset return volatility, N(d) is cumulative standard normal.
  - Put option expression presented in analogous form.
- Calibration and implied asset estimation:
  - Traditional Merton (1974) calibration requires observed equity value E, equity volatility σE, and distress barrier to solve for implied asset value A and implied asset volatility σA.
  - Once A and σA are known, expected losses to bank creditors (implicit put E_P) can be calculated.
- Decomposition of expected loss (implicit put) into:
  - (i) risk-neutral default probability (RNDP)
  - (ii) loss given default (LGD)
  - (iii) value of default-free debt (B).
- Spread relationship:
  - s(T) relates to RNDP and LGD by: s(T) = –(1/T) ln(1 – RNDP × LGD)
  - Yield-to-maturity relationship and credit spread definitions provided (s = y – r) and algebraic derivations shown.

### Modeling Default Risk (process and probabilities)
- Asset return stochastic process:
  - dA / A = μ dt + σ dε, where μ is asset drift, σ is asset return standard deviation, ε ~ N(0,1).
- Actual probability of default:
  - For horizon t: actual probability = N(d) with d = [ln(A0 / B) + (μ – 1/2 σ^2) t] / (σ sqrt(t)) (expressions shown in text).
- Risk-neutral probability of default:
  - Uses r (risk-free rate) as drift; risk-neutral default probability is larger than actual probability when μ > r.
- Visual explanation:
  - Appendix Figures 11a and 11b illustrate distributions of asset value at T, promised payments (default barrier), and comparison of actual vs. risk-neutral default probabilities.

### Market Price of Risk (MPR) and effect on spreads
- MPR concept:
  - Translates actual default probability to risk-neutral default probability; reflects investors' risk appetite.
  - Increase in MPR associated with systemic increase in average volatility of bank assets, higher RNDP, and higher expected losses to bank creditors.
- MKMV formulation:
  - MPR = SR × ρ_AM (i.e., ,AM SR ρ λ = in text where λ is market price of risk).
  - Using two-moment CAPM derivation:
    - β = ρ_AM (σ_A / σ_M)
    - SR is market Sharpe Ratio.
    - Thus ,AM SR λ ρ = and λ = (μ_A – r) / σ_A; alternative expressions provided.
- Empirical ranges and drivers:
  - ,AM ρ is usually around 0.5 to 0.7 (calculated bank by bank in the MKMV CreditEdge model).
  - SR around 0.55 to 1.2 during the last few years.
  - Main driver of MPR is global Sharpe ratio.
- Scenario linkage:
  - Four stress test scenarios are associated with four scenarios for the MPR.
  - Adverse Scenario 2 assumes declining GDP growth corresponds to a severe crisis (similar to 2008–2009) and is associated with an increase in MPR.
- Appendix Figure 12:
  - Shows historical Market Sharpe Ratio and projected path for Adverse Scenario 2 (scale 0 to 1.2).

### Appendix IV. Econometric model linking macro factors to CCA outputs
- Objective:
  - Explain changes in bank EDFs and asset returns using macroeconomic environment.
- Specification (institution level):
  - y_it = α_i + β X_t + ε_it (initial representation)
- Panel specification (to address small sample per institution):
  - y_it = α_i + γ_i + β X_t + δ (I_i × X_t) + ε_it (interaction term included to capture different effects by institution).
- Data and variables:
  - EDFs and asset returns for Israel's five largest banks.
  - Quarterly macro variables: GDP growth (annual rates, difference from steady state), inflation (annual rates, difference from steady state), depreciation of effective foreign exchange (annual rates, difference from steady state), Bank of Israel short interest rate (annual rates, difference from steady state), TA100 return (annual rates, difference from steady state), yield on non-indexed long government bonds (non Indexed, 8–10 years, annual rates) or gap between this yield and the short interest rate.
  - Quarterly average of EDF and assets per institution produced samples between 34 and 59 observations (1997Q1–2011Q3).
  - Converted to an unbalanced panel giving 254 observations for banks (after extracting outliers).
- Estimation choices:
  - GLS estimation allowing for heteroscedasticity (homoscedasticity rejected).
  - Autocorrelation not modeled since dependent variables are not autocorrelated.
  - Various specifications tested; insignificant variables omitted.
  - Fixed effects and interactions used to capture bank-specific responses to macro variables.
- Outputs and usage:
  - RD provided forecasts for all institutions until the end of 2014 for all three scenarios and for both assets return and change in EDF.
  - A 2-standard-error-of-the-prediction confidence interval was calculated for each institution forecast.
- Note on limitations:
  - Small sample sizes per institution motivated panel approach; interactions partially mitigate the equal-parameter constraint across institutions.
  - Further research and additional specifications are noted as desirable.

*Italic source: Appendix Figure 10 and accompanying text from the provided IMF content unit.*

### conclusion is that the CI for the insurance sector is wider than the one for the banking sector.

### _cr1288 - conclusion is that the CI for the insurance sector is wider than the one for the banking sector.

### Banks — Regression Results (Appendix Table 3)
- Sample: Number of id 5; Observations 254 (Assets Return), 207 (Change in EDF).
- Dependent variables: Banks - Assets Return; Banks - Change in EDF.
- Coefficients and standard errors (in brackets):
  - GDP Growth: 0.0664*** [0.0164] (Assets Return); -0.3053*** [0.0772] (Change in EDF)
  - Inflation: 0.0563*** [0.0186] (Assets Return); -0.2849*** [0.0857] (Change in EDF)
  - Depreciation: 0.1034*** [0.0266] (Assets Return)
  - Depreciation (t-1): 0.0279*** [0.0067] (Assets Return)
  - BOI Interest Rate: 0.1955*** [0.0242] (Assets Return); -0.3813** [0.1533] (Change in EDF)
  - Long Interest Rate: 0.1133** [0.0519] (Change in EDF)
  - TA100 Returns: 0.0060*** [0.0015] (Assets Return)
  - Dummy for inst. 2: 0.0026 [0.0022] (Assets Return)
  - Dummy for inst. 3: 0.0085*** [0.0029] (Assets Return)
  - Dummy for inst. 4: 0.0089*** [0.0033] (Assets Return)
  - Dummy for inst. 5: 0.002 [0.0021] (Assets Return)
  - Dummy for inst. 1 X TA100 Returns: -0.0652*** [0.0178] (Change in EDF)
  - Dummy for inst. 2 X TA100 Returns: -0.0449*** [0.0074] (Change in EDF)
  - Dummy for inst. 3 X TA100 Returns: -0.0255** [0.0104] (Change in EDF)
  - Dummy for inst. 4 X TA100 Returns: -0.1446*** [0.0368] (Change in EDF)
  - Dummy for inst. 5 X TA100 Returns: -0.0560*** [0.0203] (Change in EDF)
- Constant: 0.0074*** [0.0015] (Assets Return).
- Notes:
  - Standard errors are in brackets.
  - Significance notation: *** - 1 percent significance, ** - 5 percent significance, * - 10 percent significance.
  - The smaller samples in rows 3 and 4 are a result of using the long interest rate, whose data is available only since mid 2001.

### CCA Stress Test Results — Banks (Appendix Figure 13)
- Metrics shown:
  - CCA Stress Test Results – 1 year EDF (percent) and Fair Value Credit Spreads (bps).
  - Weighted Average EDF 5 banks past and Base scenario; Weighted Average 5 yr FV spread (bps) banks past and Base scenario.
  - Weighted Average EDF 5 banks past and Adv 1 scenario; Weighted Average 5 yr FV spread (bps) banks past and Adv 1 scenario.
  - Weighted Average EDF 5 banks past and Adv 2 scenario; Weighted Average 5 yr FV spread (bps) banks past and Adv 2 scenario.
- Chart elements (as labeled in source):
  - EDF -Base; EDF; EDF + 2SE; EDF -2SE.
  - Weighted 5YR FVCDS -Base; Weigthed 5YR EICDS; Weigthed 5YR EICDS + 1SE; Weigthed 5YR EICDS -1SE; Weigthed 5YR EICDS + 2SE; Weigthed 5YR EICDS -2SE.
  - EDF -Scenario 1; EDF -Scenario 2 with similar EDF and FVCDS series and SE bands.
- Axis scales appearing in figures:
  - EDF axis: 0.0000, 0.2000, 0.4000, 0.6000, 0.8000, 1.0000, 1.2000.
  - FV spread axis examples: 0.0, 50.0, 100.0, 150.0, 200.0, 250.0, (300.0 for some scenarios).

### Insurance — Top-Down Stress Test Matrix for Long Term Savings Providers (Appendix V)
- Institutions covered:
  - Life insurers, pension funds and provident funds.
  - The whole market of long term savings is included, differentiated by type of provider.
- Market share / balance sheet aggregates:
  - 100 percent of the market.
  - NIS 140 billion insurance providers.
  - NIS 300 billion, provident funds.
  - NIS 95 billion, new pension funds.
- Data Source:
  - Regulatory data as of June 2011.
- Methodology / supervisory models:
  - Immediate shock on the value of the portfolio.
  - Stress test horizon: Immediate shock on the value of the portfolio (one-year horizon implied by behavioral adjustments section).
- Shocks — Scenario analysis (single factor shocks):
  - Historic based scenario: 4th Quarter 2008:
    - Foreign investments -15 percent
    - Loans -4.4 percent
    - Corporate bonds -9.4 percent
    - Government bonds and deposits 4.6 percent
    - Equity -29.8 percent
  - Simulated local shock scenario:
    - FX depreciation -20 percent
    - Corporate spread 200 bp
    - Risk free interest + 20 percent
    - Equity -30 percent
- Risk factors:
  - Equity prices, yield curve, production, lapses, etc. (see above shock listings).
- Behavioral adjustments:
  - Managerial and policyholders’ reactions: None for one-year horizon.
- Regulatory and accounting:
  - Definition of solvency: No capital impact since focus on depreciation of portfolio value (risk-sensitive regime is necessary).
  - Accounting requirements: Mark to market valuation is preferable.
- Results — Asset losses:
  - Historic based scenario: 4th Quarter 2008
    - 9.45 percent loss in LTS provided by insurers
    - 7.47 percent loss in LTS provided by provident funds
    - 5.58 percent loss in LTS provided by pension funds
    - Weighted average 7.70 percent loss
  - Simulated local shock scenario:
    - 3.28 percent loss in LTS provided by insurers
    - 5.43 percent loss in LTS provided by provident funds
    - 2.55 percent loss in LTS provided by pension funds
    - Weighted average 4.43 percent loss

*IMF staff report: _cr1288 - conclusion is that the CI for the insurance sector is wider than the one for the banking sector.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2012/_cr1288.pdf_