## 8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of Bermuda ## Source details **Canonical URL:** [8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of Bermuda](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14133.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14133.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14133.pdf.json) --- ### I. Introduction: role and objectives of stress testing - Stress testing defined as forward-looking technique measuring sensitivity to low-probability, high-impact events. - Objectives in financial stability context: - Forecast impacts to identify vulnerabilities from rapid deterioration in operational and market environments. - Inform policy discussion in surveillance (FSAP, Article IV, crisis programs) and validate internal economic capital models. - FSAP vs supervisory stress tests: - FSAP: surveillance, medium-term focus, very severe but plausible scenarios, no immediate supervisory implications. - Supervisory: identify capital shortfalls, may require management capital plans/recovery actions (restrict dividends, variable remuneration). ### II. Overview and framework for insurance stress testing - Purpose of paper: - Review state of system-wide solvency stress testing for insurance. - Provide guidelines for consistent implementation for macroprudential surveillance. - Augment banking-focused asset risk stress testing with underwriting/liability risk guidance. - Paper components: - Characteristics of a stress testing framework and IMF application to insurance markets. - Cross-country Stress Testing Matrix (STeM) comparison across FSAPs. - Guidance for macroprudential stress testing frameworks for insurance and country authorities preparing for FSAPs. ### A. Macroprudential stress testing for insurance (MPS) - Objective: identify, measure, and monitor vulnerabilities and risks of multiple firms to mitigate systemic risk. - IAIS three-step MPS process: - i. Determine key indicators of macro-financial vulnerabilities across business models and distinguish traditional vs non-traditional/non-insurance activities. - ii. Define risk factors and transmission channels to identify common exposures, concentrations, interdependencies, spillovers and contagion risks. - iii. Integrate key risk drivers and implement meaningful shocks to determine supervisory action, operational changes, and policy measures. - Status and practice: - Development still in its infancy; IAIS survey finds most supervisors perform macroprudential surveillance mainly via trend monitoring. - IAIS published initial guidelines (IAIS, 2013a). - Stress testing use limited and more often microprudential; ICP 24 expects supervisors to monitor vulnerabilities and analyze plausible unfavorable future scenarios. ### B. Key differences between banks and insurers and implications for stress testing - Funding and cash-flow: - Insurers funded largely by upfront premiums (stable cash flows); not subject to bank-style runs or immediate collateral calls in same way. - Liability structure and payout timing: - Insurance liabilities are technical provisions; payouts managed over time; long-term for life insurers. - Risk types and cyclicality: - Insurance-specific risks (mortality, morbidity, casualty, liability) largely idiosyncratic and diversifiable; banks’ credit/liquidity risks highly cyclical. - Integration and interconnectedness: - Insurers have relatively low interconnectedness in payment/clearing systems but resolvability for large complex insurers can be harder. - Funding structure and liquidity: - Absence of maturity transformation reduces some funding fragility; negative duration gap common (“short-long mismatch”). - Lapse risk: economic conditions or higher rates can trigger lapse/surrender requiring asset sales. - Valuation challenges: - Emphasis on best/current estimates of liabilities; actuarial models may not capture stochastic properties under stress (mortality, lapse rates, catastrophe, long-tail perils). - Implication: - Stress tests must address both asset and liability sides with emphasis on underwriting/liability risks and model/valuation uncertainty. ### Data and practice observations from FSAPs (selected findings and coverage) - Since 1999, "22 stress tests of the insurance sector" have been published out of a total of "256 FSAPs (as of February 2014)". - Insurance stress testing secondary to banking in FSAPs; bank stress testing integral to every mission. - BU approaches relied on national supervisory frameworks and prudential data; TD approaches less common due to data constraints. - Average market coverage in FSAP insurance stress tests: about "70 percent" on average; two thirds of exercises range between "55 and 85 percent". - Comprehensive system-wide coverage examples: 30 firms in Switzerland and the United States; 78 in France. - Drivers of increased focus: NTNI activities and designation of G-SIIs (IAIS, July 2013). ### Methodology, valuation, and scenario design - Scope and object of analysis: - Assess insurance business lines, investment portfolio composition, intra-group connectedness, interlinkages with other financial institutions, solvency standards, tax regime, and policyholder protection funds. - BU vs TD: - BU predominant in FSAPs; TD useful for quick results but hindered by granularity/data limitations. - Robust TD modeling requires global valuation/solvency regime and harmonized reporting. - Valuation approaches: - Three major approaches preserved exactly as in source: - Accounting basis (e.g., Solvency I): historical prices/cost accounting. - Risk-based approaches (e.g., RBC, Solvency II): risk-based with possible alleviations (volatility/matching adjustments). - Market-consistent valuation (e.g., SST): discount cash flows with appropriate risk-free rates; assets market-valued. - FSAP valuation practice: fully market-consistent valuation applied to a few countries (Belgium (2013), Canada, Portugal, Spain, Singapore (2013), Switzerland); Belgium used statutory, “near market-consistent” and fully market-consistent valuations in different exercises. - Scenario design principles: - Scenarios must be severe yet plausible, driven by identified macro-financial vulnerabilities and transmission channels. - Combine historical and simulated/hypothetical outcomes with supervisory judgment to maintain plausibility. - Relevant risk categories: market and credit risks, interest rate risk, foreign exchange risk, liquidity risks, underwriting risks, concentration and interconnectedness. - Use multiple scenarios/combinations to capture business model diversity. ### Typical calibrated scenarios and parameter magnitudes (examples preserved verbatim) - Asset shocks (examples and frequencies): - Equity risk: about "-30% (but up to -50%)". - FX risk: around "+/-20% (but up to +/-50%)". - Real estate risk: about "-20% (but up to -50%)". - Interest rate risk: about "+/-200bps parallel shift (H frequency)"; national approaches often about "+/-100bps parallel shift (H frequency)". - Credit spreads: increase of credit spreads by up to "50%", downgrade of counterparties by "2-4 notches". - Liability and underwriting shocks: - Life underwriting—mortality/morbidity/longevity: mortality/morbidity/longevity of annuitants "about +25% each". - Lapse/surrender rates: mass lapse "about 25% (but up to 50%)". - Non-life natural catastrophe: set to maximum historical claims experience, such as "1-in-50 years probable maximum loss (PML)"; national approaches sometimes use "1-in-200 years". - Other non-life shocks: "+10% in the cost of claims"; "+15% higher frequency of claims greater than EUR 30,000". - Credit risk and bond downgrade scenarios (selected historic supervisory calibrations): - "1.5 percent loan loss in Japan (2003)" - "4.4 (9.4) percent loss on loans (corporate bonds) in the case of Israel" - FX typical variation between "15 and 35 percent"; extreme examples: Netherlands "45 percent depreciation of the euro"; Denmark "+/- 40 percent"; South Africa "+/- 50 percent". - Life underwriting shocks in exercises: - Mortality shocks "between 15 and 30 percent". - Pandemic/morbidity shocks "mostly in a range of 15 to 25 percent above the baseline assumptions". - Lapse increases historically specified as "50 percent" (Spain and Portugal) and "30 percent" (Guernsey); Belgium (2013) used "(+30 percent)" only where surrender value exceeded technical provision. ### Aggregation, horizons, and output metrics - Aggregation methods observed across FSAPs: - No aggregation of sensitivities; simple summation; summation of combinations; aggregation via correlations (sometimes setting correlation coefficients to 1); sole use of a correlation matrix (Swiss FSAP). - Time horizons: - Single-period stresses predominant; multi-year horizons used by Canada (up to "five years"), Singapore ("three-year"), Japan ("two-year" in one exercise). - Output measures: - Main output: change in capital adequacy (solvency ratio), often over a "one-year risk horizon". - Solvency ratio: excess of equity capital over prescribed capital requirement (PCR) or national capital standard. - Many jurisdictions have a minimum capital requirement (MCR) or minimum solvency margin (MSM); breach triggers strong supervisory actions. - Capital resources and eligibility: - Careful consideration of inclusion of subordinated debt, intangible assets, deferred tax assets; ICP 17.11.34 categorization into highest/medium/lowest quality capital. ### Liquidity and non-financial risks - Liquidity risk overview: - Long-term funding profile typically less susceptible than banks, but liquidity/maturity mismatches and negative duration gap still common. - OTC derivatives, securities lending, repo transactions can generate short-term collateral/margin needs distinct from insurance liability cash flows. - Relevance to non-life and reinsurance: - Particularly relevant for reinsurers required to settle large claims after catastrophes. - Supervisors monitor liquid assets vs potential payment amounts; no well-established industry standard for liquidity stress testing yet. - Suggested liquidity stress inputs: - Large catastrophe claims; lower future premiums; collateral needs from OTC derivative transactions. ### Box 7 — Case Study: Belgium (representative FSAP practice) - Design and scope: - BU exercise covering six largest insurers (> "70 percent" of sector), using mid-2012 prudential data (one insurer used end-September 2012). - Scenario design: - NBB calibrated four market risk factors for mild and severe scenarios: interest rates, equity prices, corporate spreads, sovereign spreads. - Additional shocks: mass lapse in life business and PML for a catastrophic tail event. - Methodology: - Insurers aggregated impacts using a correlation approach similar to Solvency II standard formula; compared own funds to SCR and MCR (treated as fixed reference points). - Validation and communication: - Emphasized due diligence and validation checks: firm-specific validation, peer group analysis, time-series comparison. - Communicate results in a non-technical manner; present sensitivity to assumptions and use standardized templates for comparability. - Representative parameter magnitudes summarized: - Equity: about "-30% (but up to -50%)". - FX: around "+/-20% (but up to +/-50%)". - Real estate: about "-20% (but up to -50%)". - Interest rates: about "+/-100bps" to "+/-200bps". - Credit spreads: increases up to "50%"; downgrades "2-4 notches". - Life underwriting: mortality/morbidity/longevity "about +25% each"; mass lapse "about 25% (but up to 50%)". - Non-life catastrophe: "1-in-50 years PML" typical; national approaches sometimes "1-in-200 years". ### Box 8 — Case Study: Bermuda (detailed) - Market context: - Bermuda is the third largest (re)insurance market globally; focus on property and casualty risks. - BMA requires all large firms to perform an annual stress test and submit results with Capital and Solvency Return. - Framework and reporting: - Firm-level stress testing via internal models or prescriptive shocks; submission must include key assumptions, post-stress statutory assets and liabilities, with and without reinsurance, and occurrence and relative return period results. - Prescribed financial scenarios: - Severe decline in equity prices of "40 percent" without diversification across markets. - Widening of credit spreads; valuation haircuts; general upward shift in yield curve of "50 basis points". - Credit spread widenings calibrated between "163 basis points" for "AAA" to "3,188 basis points" for "BB or lower". - FX shocks based on four times difference of maximum implied annualized volatility between 1 Jan. 2008 and end-2011 and long-term average since 1 Jan. 2005. - Sovereign risk shock calibrated from 99th percentile historical density of CDS forward contracts. - Underwriting scenarios: - Prescribed catastrophe scenarios (Lloyd’s Handbook on Realistic Disaster Scenarios): U.S. windstorm, U.S. earthquake, Non-U.S. windstorm, Non-U.S. earthquake, aerospace/aviation, marine. - Non-peak perils: U.S. oil spill, U.S. tornadoes, Australian flooding, Australian wildfires. - Additional risks: pandemic, terrorism, worst-case annual aggregate catastrophe loss scenarios combining economic and underwriting losses. - All lines of business included net of protection (reinsurance, retrocession, insurance-linked securities). - IMF 2007 BU system-wide solvency stress test of ten large commercial (re)insurance and long-term insurance companies (case study findings): - Catastrophic events significantly negative for aggregate capital; scenarios combining catastrophes with recession had greatest impact. - No firm failed to meet regulatory capital requirements under scenarios. - Shortcomings identified: incomplete coverage of financial market effects and use of accounting data rather than economic valuation. - Developments and enhancements: - BMA introduced BSCR in 2010 addressing prior shortcomings; introduced worst-case annual aggregate catastrophe loss scenario. - Future enhancements: BMA capacity to execute system-wide industry-level stress testing regularly and expanded granular treatment of catastrophe risk. ### IV. Discussion and conclusion — key issues and considerations - System-wide stress tests for insurance have risen in importance post-crisis. - Key differences imply bank-focused tests are not directly transferable; need to address liability-side risks and differing sensitivities (e.g., interest rate shock may raise insurer solvency and reduce bank solvency). - Trend toward market-consistent valuation and convergence of regulatory standards (IAIS BCR for G-SIIs, Solvency II) supports greater precision and comparability. - Ten key design/calibration/interpretation issues preserved verbatim as themes: - i. Risk-factor evolution over time. - ii. Valuation methodology robustness. - iii. Trade-off between accuracy and timeliness. - iv. Total balance sheet approach. - v. Aggregation and diversification assumptions. - vi. Flexibility through combinations and magnitudes. - vii. Beyond solvency—profitability and liquidity measures. - viii. Multi-period scenarios. - ix. Intra-group transactions and consolidated reporting limitations. - x. Secondary impacts and avoiding distortive supervisory incentives. - Role of qualitative analysis and expert judgment: - Qualitative elements (business strategy, underwriting practices, innovation in risk transfer) remain essential given limits of data and modeling for tail risks. *Source: _wp14133 - References .............................................................................................................* ### References ............................................................................................................. ### _wp14133 - References ### References and Appendices - References ................................................................................................................................59 - Appendix I—Tables .................................................................................................................71 - Appendix II―Additional Background ....................................................................................81 ### Tables - 1. Overview of Possible Validation Checks ............................................................................50 - 2. Summary of Key Assumptions in IMF Stress Testing of Insurance Sectors .......................52 ### Figures - 1. Overview of IMF FSAPs and Completion of Insurance Stress Tests ....................................7 - 2. Number of Completed Insurance Stress Tests in FSAPs Before and After the Global Financial Crisis ..................................................................................................................8 - 3. Stylized Insurance Balance Sheet and Solvency Control Levels .........................................15 - 4. Stress Testing Process ..........................................................................................................17 - 5a. Overview of Solvency Regimes—Risk Measurement .......................................................26 - 5b. Overview of Solvency Regimes—Valuation Standards ....................................................26 - 6. Elements of Risk Assessment and Scope of FSAP Stress Testing ......................................29 - 7a. Presentation Templates of Outputs (hypothetical singe-period test) .................................46 - 7b. Presentation Templates of Outputs (hypothetical multiple-period test) ............................47 ### Boxes - 1. General Macro-Financial and Systemic Risk Implications for Insurance ...........................16 - 2. The Taxonomy of Stress Testing Approaches .....................................................................18 - 3. Recessionary Scenarios in the Insurance Sector ..................................................................31 - 4. Assessing the Impact of Low Interest Rates on Insurance Activities ..................................33 - 5. Liquidity Risk in Insurance ..................................................................................................37 - 6. Examples of Supervisory Approaches of Insurance Stress Testing .....................................39 - 7. Case Study: Belgium Insurance Stress Test for the FSAP ..................................................49 *Source: _wp14133 - References .............................................................................................................* ### 8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of ### 8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of Bermuda ### I. Introduction: role and objectives of stress testing - Stress testing is a forward-looking technique measuring sensitivity of a portfolio, an institution, or an entire financial system to low-probability, high-impact events. - Well-formulated stress tests use sensitivity and/or scenario analyses to assess capacity to absorb shocks from key macro-financial risks. - In financial sector stability analysis, stress tests aim to forecast impacts to identify vulnerabilities from rapid deterioration in operational and market environments, from system level down to firm and portfolio level. - Stress tests can be sector-limited or cross-sectoral to capture interconnectedness of banks, insurers, and other market participants. - Over the last decade, stress testing has become central to the Fund’s financial sector surveillance; it is a key component of the Financial Sector Assessment Program (FSAP) and is used in Article IV and crisis program work. - FSAP stress tests differ from supervisory stress tests: - FSAP: designed for surveillance, medium-term focus, typically involve very severe but plausible scenarios to assess system resilience; results inform policy discussion and have no immediate supervisory implications. - Supervisory: aim at identifying potential capital shortfalls and may require management capital plans/recovery actions (including restricting dividends and variable remuneration). - FSAP stress testing has greater severity due to a longer forecast horizon and comprehensive scenario-based treatment affecting assets and liabilities simultaneously. - Stress testing also helps validate internal economic capital models to demonstrate resilience to extreme shocks. ### II. Overview and framework for insurance stress testing - Purpose of paper: review current state of system-wide solvency stress testing for insurance; provide guidelines for consistent implementation for macroprudential surveillance; augment banking-focused asset risk stress testing with underwriting risk discussion affecting insurance liabilities. - Paper components: - Articulate main characteristics of a stress testing framework and demonstrate application in IMF surveillance of insurance markets. - Compare implementation of various stress tests across FSAPs using detailed cross-country Stress Testing Matrix (STeM). - Discuss properties and guidance for constructing macroprudential stress testing frameworks for insurance and for country authorities preparing for FSAPs. ### A. Macroprudential stress testing for insurance (MPS) - MPS objective: identify, measure, and monitor vulnerabilities and risks of multiple firms within a country and/or across boundaries to mitigate systemic risk. - Macroprudential stress testing for insurance builds on transmission channels affecting investment and underwriting performance within the broader financial system. - IAIS three-step MPS process for insurance: i. Determine key indicators of macro-financial vulnerabilities across insurance business models and distinguish traditional vs non-traditional/non-insurance activities. ii. Design conceptual approach defining risk factors and transmission channels to identify common exposures, concentrations, and interdependencies that generate spillovers and contagion risks. iii. Develop macroprudential framework integrating key risk drivers and implement meaningful shocks to determine supervisory action, operational changes, and policy measures to mitigate severity and duration of sector distress with adverse effects on the real economy. - Status of MPS in insurance: - Development is still in its infancy. - IAIS survey found most supervisory authorities carry out macroprudential surveillance activities, predominantly monitoring trends and analyzing system-wide impact of macroeconomic variables on domestic insurance markets; international data analysis has received less attention. - IAIS published initial guidelines of MPS for insurance (IAIS, 2013a). - Use of stress testing for MPS: - Limited; supervisors increasingly use stress tests for microprudential purposes. - Insurance Core Principles (ICP), revised in 2011, introduced enterprise risk management for solvency, including stress testing/scenario analysis—largely focused on viability of individual institutions rather than system-wide joint risk factor impacts. - Supervisors are expected to monitor vulnerabilities and carry out analysis of “plausible unfavorable future scenarios with the objective and capacity to take action at an early stage, if required,” aimed at identifying and mitigating systemic risk (ICP 24). - Solvency stress tests for insurance typically assess capital impact of shocks to risk factors on the total balance sheet: - Common macro-financial risk factors: interest rates/credit spreads, asset risks, foreign exchange rates. - Underwriting risks require specific stress approaches: deterioration of technical provisions (including reserves and best estimates), demographic risks, and catastrophic risks. - Differentiation by business model matters (e.g., life insurance with/without minimum guarantees, non-life short-tail, non-life long-tail, non-proportional reinsurance). ### B. Key differences between banks and insurance companies and implications for stress testing - Fundamental contrasts imply limited usefulness of directly transplanting bank-focused stress testing approaches to insurers. - Principal differences and implications: - Funding and cash flow: - Insurers funded largely by upfront premium payments, producing stable cash flows (an “inverted production cycle”); banks fund illiquid longer-term assets with short-term liabilities. - Insurers typically do not face immediate collateral calls or customer runs in the same way banks do. - Liability structure: - Insurance liabilities are technical provisions (often long-term for life insurers) backed by diversified investment portfolios composed mostly of high-quality assets. - Payouts from claims can be managed over time, reducing speed of cash outflows in stress. - Risk types and cyclicality: - Insurers face credit, operational, and market risks, but insurance-specific risks (mortality, morbidity, casualty, liability) are largely idiosyncratic and generally independent of the economic cycle, allowing diversification gains. - Banks’ credit and liquidity risks are highly correlated with the economic cycle. - Integration in financial infrastructure: - Insurers are not organizers of payment or clearing systems and have limited direct intra-system claims and liabilities, exhibiting relatively low levels of interconnectedness domestically and internationally. - Lower interconnectedness reduces negative externalities from failure but may complicate resolvability for large complex insurers. - Regulatory constraints on risk generation: - Insurance regulation limits underwriting to insurable interest; insurers generally control and mitigate existing risks rather than generate additional risks. - Reinsurance transfers are partial; most risk remains on ceding insurer’s balance sheet. - Funding structure and liquidity: - Absence of maturity transformation and transaction clearing services means insurers’ liquidity positions are less influenced by external funding conditions; strong operating cash flows via upfront premiums and longer-term retail funding provide resilience. - Insurers can be insolvent yet remain liquid due to long-term nature of business, but liquidity risk can arise from asset-liability mismatches and cash flow management: - Asset-liability matching is critical; insurers often have a negative duration gap (“short-long mismatch”). - Cash flow management: liabilities are normally not redeemable on demand; claims are paid via sale of liquid assets when needed. - Lapse risk: adverse economic conditions or higher interest rates can trigger higher lapse rates, requiring asset sales to meet surrender payments. - Valuation challenges: - Long-term funding places emphasis on valuation methods for best/current estimates of liabilities; actuarial models may not fully capture stochastic properties of risk factors under stress. - Calibration errors and sparse empirical observations can cause model deficiencies for mortality, lapse rates, catastrophe risk, pandemics, terrorism, or long-tail perils (e.g., asbestos). - Overall implication for stress testing: - Insurance stress tests must explicitly address both asset and liability sides, with particular attention to underwriting/liability risks and valuation/model uncertainty under stress. - Macroprudential stress testing should identify common exposures, concentration risks, and interdependencies unique to insurance business models rather than relying solely on bank-oriented frameworks. ### Data and practice observations from FSAPs (selected findings) - Since 1999, 22 stress tests of the insurance sector have been published out of a total of 256 FSAPs (as of February 2014). - Insurance stress testing has played a secondary role relative to banking in FSAPs; bank stress testing is integral to every mission. - Reliance on national supervisory frameworks for bottom-up (BU) approaches is greater for insurance due to balance sheet differences and lack of global solvency/valuation standards; BU approaches are more resource-intensive. - Increased focus on systemic risk in insurance has been driven by assessments identifying vulnerabilities from non-traditional/non-insurance (NTNI) activities and the designation of G-SIIs (IAIS, July 2013), which introduced insurance-relevant indicators and proposed enhanced supervision, recovery and resolution, and capital requirements. _Italic: Source — IMF staff compilation from “8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of Bermuda” (excerpt)._ ### Box 1. General Macro-financial and Systemic Risk Implications for Insurance ### Box 1. General Macro-financial and Systemic Risk Implications for Insurance ### Overview - Economic cycles impact investment income and underwriting performance of insurance companies over time. - Macro-financial linkages vary by business line and by technical factors influencing pricing and reserving of insurance products. - Footnote references in source: 1, 2, 3, 4, 5. ### Life insurance: sensitivity to economic volatility - Certain life insurance activities exhibit a high correlation with economic volatility because they rely on stable investment returns to match expected claims over the long run. - Life companies typically have higher asset leverage than non-life insurers and longer duration investments, increasing susceptibility to secular changes in credit spreads and interest rates (unless sufficiently hedged). - Lower interest rates: - Heighten re-investment risk for new funds generated from premiums. - Increase the present value of future claims, potentially creating critical asset-liability mismatches despite temporary asset valuation gains. - Monetary easing during economic slowdowns (possibly combined with higher asset impairments) lowers investment income and could jeopardize returns of life insurers. - Consequences may include some firms being forced to lower guaranteed premium rates or returns in capital-intensive investment products. - Footnote 2 notes that during the financial crisis several mitigating factors allowed life insurers to mitigate investment risks and that realization of adverse effects can be reduced by regulatory forbearance, product designs, and/or personal tax regimes. ### Non-life insurance: underwriting cycle and catastrophes - Underwriting performance broadly tracks economic growth, which influences available capacity and future pricing as insurers adjust to changing demand and cost of capital. - Large catastrophe losses tend to be followed by premium hardening due to lower insurance capacity. - Cost of replenishing capacity is accentuated if the insurance cycle coincides with economic downturns when rising risk aversion of investors and depressed asset prices raise the cost of capital. - Excess capacity pushes pricing lower on renewals; acceleration of price declines is more likely if coinciding with economic booms and lower cost of capital. - Price dynamics are influenced by the extent to which renewal rates trail expected underwriting losses. - Selective price increases become more likely if the long-term loss trend outpaces historical price increases at the margin. - Higher rates of inflation during periods of economic recovery can adversely affect provisioning and reserve adequacy, especially if changes in claims activity negatively impact performance in real terms. - Footnote 4: An inflationary effect beyond expectations (which implies higher nominal insurance cover due to price appreciation) could cause insurers being under-reserved for future claims. - Erratic occurrence of natural catastrophes and man-made disasters drive significant changes in underwriting performance; system-wide impact depends on firm-specific and/or cross-sectional concentration of exposures. - Footnote 3: Growing popularity of insurance-linked securities on natural catastrophes increases linkage of some insurers to capital markets, while outsourcing of insurance risk via alternative risk transfer mechanisms has arguably muted insurance-cycle impact on some business lines. ### Liquidity, non-traditional life products, and funding arrangements - Some non-traditional forms of life insurance are inherently more susceptible to cyclical effects than mainstream individual life insurance. - Funding arrangements via capital markets (repurchase agreements, security lending, and OTC derivatives) may require more liquidity over shorter time periods than insurance claims. - Potential high quality collateral calls from OTC derivatives transactions or margin calls from cash collateral reinvestment from securities lending differ markedly from long-term cash flow projections associated with insurance liabilities. - Large transactions of liquidity swaps could worsen insurers' liquidity positions by significantly reducing available cash and liquid assets. - Cash flow models for security financing transactions are generally derived from mark-to-market valuations and can give rise to margin calls if funding liquidity deteriorates. - Insurance-backed contracts (e.g., institutional investment and third-party asset management products such as guaranteed investment contracts (GICs)) implied some liquidity risk to the extent policyholders could surrender contracts at short notice with limited penalties. - Such surrender behavior could create a cash flow scenario comparable to a bank run if contracts are surrendered on short notice. - Historical example (footnote 5): In August 1999, holders of GICs issued by General American Life Insurance Co. exercised put options requiring the life insurer to rapidly repay principal and interest, causing its parent firm, General American Group, to go into administration. *Box 1. General Macro-financial and Systemic Risk Implications for Insurance* ### Box 2. The Taxonomy of Stress Testing Approaches ### Box 2. The Taxonomy of Stress Testing Approaches ### Stress testing purposes and approaches - Purposes include macroprudential surveillance, microprudential supervision, crisis management, and risk management. - Performed by bottom-up (BU), top-down (TD), or combined approaches: - BU tests: companies perform tests based on prescribed assumptions/scenarios and provide results to supervisory authority. - TD tests: supervisory authority runs tests using input data from companies (regular reporting or public disclosure). - Macroprudential stress tests: determine system-wide resilience; can be single-sector or cross-sectoral (banks, insurers, other market participants). System-wide tests in IMF FSAPs generally fit this description. - Microprudential stress tests: determine firm-specific vulnerabilities; commonly used by insurance supervisors, especially where risk-based solvency systems are absent. - Crisis management stress tests: assess actual or potential capital needs of distressed companies (examples cited: SCAP, CCAR in early 2009; EBA recapitalization exercise in 2011). - Risk management stress tests: run by individual financial institutions for internal portfolio and planning purposes. - Reverse stress tests: - Set a target threshold for an output measure (e.g., a solvency ratio of 100 percent) and identify shock magnitude to breach threshold. - Condition can be relaxed by providing shocks to other risk factors. - Guidance note: reverse stress tests should be used only for material and relevant risks to avoid meaningless results; example—low equity exposures can render reverse tests implausible unless extremely large equity decline assumed; conversely, relevant for large parallel shifts in the interest rate term structure. ### A. Object of Analysis - Conduct a thorough insurance market analysis of external factors and business conditions, acknowledging differences in business models, role in domestic financial sector, and international linkages. - Issues to elaborate for insurance stress testing: - Insurance business: - Prevalence of specific lines of business; share of traditional vs. investment-linked life insurance. - Existence of country-specific products (e.g., state-sponsored catastrophe insurance or retirement schemes). - Profitability breakdown, duration of liabilities, average guaranteed interest rates, penalties for early termination, use of reinsurance, and risk-mitigating features like profit sharing. - Investment portfolio: - Asset composition (equity, bonds, loans, real estate, alternative investments such as private equity, hedge funds, and commodities). - Breakdown by countries, sectors, duration and liquidity (e.g., Level 1, 2, and 3); hedging transactions, especially for interest rate risks. - Connectedness within groups/conglomerates: - Corporate structure including foreign group entities and special purpose vehicles. - Intra-group transactions (committed funding arrangements, securities holdings/lending with parents or other group companies) and intra-group reinsurance activities. - Possible supervisory interventions (e.g., ring-fencing). - Note: intra-group transactions often involve higher liquidity risk under stress; funding needs can increase reliance on intra-group transactions. - Interlinkages to other financial institutions: - Exposures via asset exposures (equity, senior bonds, subordinated debt, commercial paper), deposits, derivatives, securities lending/repos. - NTNI-related exposures (derivatives trading, securities lending/repo) usually negligible except for large firms and monoline insurers. - Insurers with large investment portfolios could provide liquidity to banks via liquidity swaps and other funding commitments; such arrangements are prone to contagion in stress. - Solvency standards: - Existence of a risk-based system, coverage of risks, level of confidence, time horizon, calibration of risk factors. - Tax regime: - Shock absorption effect of deferred tax assets/liabilities; relative tax advantages of saving products offered by insurers vs. bank products. - Policyholder protection funds: - Existence and coverage, relative to bank deposit protection. ### B. Determination of Scope - Capital assessment under stress should capture all material risks and provide total view of capital adequacy—aggregated or individual basis—of legal entities and/or across groups. - Non-financial group activities can be excluded if non-material; non-insurance financial activities need consideration. - Group participation usually involves inclusion of foreign businesses; less granular group information makes TD tests more difficult and validates BU results more complicated. - Home supervisor running group-wide tests should communicate results to host supervisors. - Representative sample selection can reduce burden; coverage decisions depend on market structure and relevance of NTNI activities. - Market coverage considerations: - Calculate life and non-life sector coverage separately. - If market highly concentrated, including largest companies by premiums or assets is usually sufficient. - Include medium/small insurers if they: conduct significant NTNI activities, are highly vulnerable to certain shocks, are highly relevant to real economy/financial sector (e.g., credit/mortgage insurance), or are very interconnected (e.g., reinsurer or part of conglomerate). - Historical FSAP practice: - Most FSAP insurance stress tests have been for samples comprising the largest firms only. - Market coverage reached about 70 percent on average, with two thirds of exercises ranging between 55 and 85 percent. - Comprehensive system-wide coverage sometimes involved large numbers of firms (examples given: 30 each in Switzerland and the United States; 78 in France). - Average market coverage of insurance companies is smaller than that of banks in FSAP stress tests. ### C. Methodological Framework and Data Quality - Model/technique choice should be proportionate to nature, scale, and complexity of insurance sector; more sophisticated models increase estimation uncertainty. - Insurance stress tests traditionally BU; IAIS (2003) proposed standardization; IAIS (2013a) and IAA (2013) provided additional guidance. - National progress: EIOPA and ECB advanced stress testing in Europe; EIOPA conducts regular BU stress tests with customized scenarios. PRA (UK) issued statements on approach to insurance supervision including stress testing references. - TD insurance sector stress tests less developed due to data constraints: - TD advantage: quick results; hindered by lack of granular data (duration, credit quality of bonds, hedging instruments, reinsurance program details, liability duration). - In absence of critical data, BU stress test is necessary first step to gather parameter ranges for TD tests. - Robust TD modeling requires global valuation/solvency regime and harmonized reporting/disclosure standards. - IAIS survey indicates increasing reliance on a mixture of TD and BU approaches (IAIS, 2011c). - IMF FSAP practice: - Predominantly applied BU stress tests using prudential data, with three exceptions: Israel (TD only); Portugal and South Africa (TD complemented BU). - Prudential data use does not preclude public data for BU preparation and cross-validation. - Only one exercise (Japan, 2003) relied solely on public data; lack of granularity (e.g., cannot distinguish Yen vs. foreign currency bonds or fixed- vs. floating-rate bonds) limited modeling of interest rate shock effects. - Data granularity considerations: - Data should be granular enough to account for intra-group transactions and banking-insurance transactions. - Most supervisory stress tests require firm reporting on legal entity basis; many stress tests completed only on consolidated basis (European Union), which may miss intra-group transaction impacts. - If intra-group and banking-insurance transactions are salient, more granular prudential information on legal entity basis is needed. - Some FSAP stress tests completed on both solo and consolidated basis. ### D. Valuation and Capital Resources - Capital adequacy under stress influenced by interaction of economic impact of risk factors and characteristics of solvency regime. - Stress test projects economic change in balance sheet for adverse scenarios based on risk sensitivity and valuation of exposures. - Valuation standards typically prescribe fair valuation for investment assets and best estimate valuation for insurance liabilities. - Cash flow projections should reflect realistic future demographic, legal, medical, technological, social, and economic developments; discount rates must be consistent with observable market prices matching timing, currency, and liquidity of liabilities. - Note: exclude insurer’s non-performance risk to avoid uneconomic volatility in net assets. - Distortions can arise from conservatism and degree of risk-sensitivity in valuation standard: - Alleviations based on conservative assumptions (asset prices and best estimate liabilities including discount rates) vary across countries and trade off robustness (reduced procyclicality) against risk of understating liabilities. - Choice of risk sensitivity and valuation standard significantly impacts estimated changes in liability-matched asset values and technical provisions. - Comprehensive analysis of applicable valuation standards should occur prior to developing stress testing. - Three major valuation approaches in solvency regimes: - Accounting basis (e.g., Solvency I in the EU): historical prices/cost accounting without consideration of actual risk; less suitable for quantifying economic impact of asset price and interest rate changes. - Risk-based approaches (e.g., Risk-based Capital in the United States, Solvency II in the EU): Solvency II includes alleviations (volatility adjustment, matching adjustment, convergence period) decreasing sensitivity to interest rate and spread changes, resulting in lower technical provisions and higher own funds. - Market-consistent valuation basis (e.g., Swiss Solvency Test): discount cash flows for technical provisions with appropriate risk-free rate based on asset swap rates (after controlling for credit risk) or replicated with sovereign bonds only; assets valued on available market prices. - Market-consistent valuation generates fair value representation and best estimates of liabilities; robust validation necessary to minimize model risk and valuation uncertainty. - Examples of risk-measure and valuation regime features (as presented): - Solvency II: consolidated; standardized/internal model at 99.5% VaR; diversification benefit. - BSCR: consolidated; factor-based/internal model at 99.0% TVaR; diversification benefit. - Japan SMR: solo, branch, consolidated; factor-based at 95.0% VaR (assets) and >99% VaR (liabilities). - RBC: solo basis; factor-based at 98.0% VaR. - Solvency I: consolidated; balance sheet/leverage ratio without consideration of actual risk. - SST: solo*; internal model at 99.0% TVaR. - Practical observations: - Most national supervisory stress tests allow/require historical cost accounting for insurance liabilities while valuing assets market-consistently. - Cost-based valuation assumes hold-to-maturity behavior by insurers, which may be inconsistent with actual asset turnover and can understate economic impact in stress. - Simplifying/conservative assumptions can overstate capital resources and understate risk measurements, leading to misleading solvency assessments. - Removing mitigating factors and adjustments tends to lower capital resources (recognizing economic loss) and increase capital requirements versus cost-based valuation. - Inflating discount rates (adding credit spread to risk-free rate) for estimating technical provisions under conservative approaches could overstate solvency ratios and increase liquidity risks ex ante. - FSAP valuation practice: - Valuation standards used in FSAP insurance stress tests heavily influenced by national solvency regimes and vary significantly. - Fully market-consistent valuation applied only to a few countries (Belgium (2013), Canada, Portugal, Spain, Singapore (2013), and Switzerland), with some adjustments for unrealized gains/losses in other cases (Japan (2013)). - Belgium (2013) used statutory (cost) accounting, a “near market-consistent” valuation per EIOPA’s QIS5 on Solvency II, and a fully market-consistent valuation of both assets and liabilities. ### E. Scenario Design and Other Assumptions - Stress tests require suitably severe yet plausible scenarios defined by macro-financial shocks to assess rapid deterioration of investment and underwriting performance. - Scenarios should quantitatively assess adverse changes in risk factors affecting earnings/capital/reserves and complement routine supervisory reporting insights. - Scenarios ideally supplement solvency frameworks by addressing risks not considered or by augmenting severity of existing relevant risk factors. - Scenario design driven by relevance of identified macro-financial vulnerabilities and transmission channels to insurance business. - Scenario construction requires projections of possible future risk materialization and transmission to insurers’ assets, liabilities, and revenues. - Scenarios should be plausible relative to firms’ capacity to control and mitigate vulnerabilities and should inform capital planning and strategic responses (“use test”). - Scenario generation approach: - Combine historical and simulated/hypothetical outcomes with supervisory judgment/validation/expert opinion to avoid hindsight bias and to incorporate potential future aberrations. - Worst-case generation with qualitative overlay must still satisfy plausibility, defined probabilistically and consistent with historical correlations of risk factors. - Risk measures within plausible domain should satisfy axioms of coherence (sub-additivity, monotonicity, positive homogeneity, translation invariance). - Relevant risk categories to cover: - Market and credit risks (corporate and sovereign), interest rate risk from asset-liability mismatches, foreign currency risk, liquidity risks, underwriting risks, concentration risks, and interconnectedness with other financial institutions. - Consider main market features: share of traditional vs. unit-linked business, average guaranteed rates, modified duration of assets and liabilities, profitability, investment portfolio composition, and degree of interconnectedness. - Diversity of insurance market: - Business models and supervisory frameworks differ across jurisdictions; relevance and impact of macro-financial shocks differ accordingly. - Use multiple scenarios and combinations to capture variation of outcomes. - Typical scenarios: - Recessionary scenario — decline in equity and property prices, increase in credit spreads, higher lapse rates, higher defaults of mortgage borrowers, offset by lower interest rates. - Banking/financial/sovereign crisis — higher credit spreads for financials, default of a large bank counterparty, stress of all asset classes including sovereign bonds. - Inflation scenario — claims inflation, rising interest rates, offset by rising equity and property prices. - Non-life underwriting shock — large natural or man-made catastrophe claim, possibly combined with reinsurer default or decline in equity prices in affected country. - Life underwriting shock — high lapse rates and/or a pandemic. - Combinations of the above. *Source: IMF staff (Box 2 text).* ### Box 3. Recessionary Scenarios in the Insurance Sector ### Box 3. Recessionary Scenarios in the Insurance Sector ### Macro-financial linkages between recessions and insurance liabilities - A decline in household wealth and disposable income could lower premium income of insurers; however the effect is likely to vary across different lines of business. Life insurance but also motor insurance tends to be more sensitive to a downturn of the economy (EIOPA, 2013b). - Lower household income and wealth can cause higher lapse rates in life insurance. The size of this shock would depend on the incentives of policyholders to surrender their policy; if the surrender value is low (and/or the interest rate levels drop below the implied return from guaranteed term insurance), the lapse rate in a recession tends to decline. - Higher corporate default rates could increase claims from financial guarantees and credit insurance. These lines of business are heterogeneous (trade and export financing; mortgage insurance; credit enhancements of structured finance). Each sub-category requires specific assumptions on probability of default and loss given default. - Higher claims could also be attributable to higher operational risk from insurance fraud; Dionne and Wang (2013) identify a business-cycle pattern in fraudulent claims for the Taiwanese automobile theft insurance market. - While macro-financial linkages of underwriting activity are generally limited, there can be tail dependence between market and insurance risks during times of stress (e.g., large catastrophes affecting asset prices and/or funding conditions). - The cost of replenishing capital after large insurance losses would be accentuated if the insurance cycle coincides with rising risk aversion of investors and higher cost of capital during an economic downturn. - Longer-term uncertainty about the ultimate consequences of a catastrophe (or a pandemic like the SARS outbreak in 2002/03) tends to be negative for the stock markets. ### Catastrophes, market reactions, and inflationary effects - Amending a catastrophe scenario by subsequent market stresses (especially a stock price decline and potentially some currency fluctuations) could be considered. - Higher rates of inflation during periods of economic recovery can adversely affect provisioning and reserve adequacy in non-life underwriting, especially if changes in claims activity negatively impact investment performance in real terms. - Some non-traditional insurance activities (funding arrangements via capital markets such as repo, security lending and OTC derivatives) differ markedly from long-term cash flow projections associated with insurance liabilities and are inherently more susceptible to cyclical effects than mainstream insurance business (IAIS, 2013c). - Evidence note: Recent studies show that even highly disastrous events like Hurricanes Katrina and Sandy in the United States or the 2011 Tōhoku earthquake and tsunami in Japan had only very short-lived effects on the respective domestic stock markets (Wang and Kutan, 2013). More severe declines occurred for specific sectors like (re)insurance, discretionary consumer goods and tourism, while broad market indices did not decline significantly. ### Scenario specification and calibration - The scenario specification should start from a baseline that reflects a likely future development of a consistent combination of changes in risk factors. - Forecasts for parameters like interest rates, credit spreads or real estate prices could be directly derived from a macroeconomic model; other factors may be estimated exogenously or based on expert judgment. - Adverse scenarios should be defined as deviations from the baseline scenario. - Use a sufficiently high but realistic confidence level for the calibration of the magnitude of shocks. - For sensitivities in single risk categories, one might use confidence levels and determine the actual shock based on time series analysis (for instance, by means of bootstrapping procedures). ### Aggregation approaches and time horizons in FSAP stress tests - Traditionally, most FSAP stress test exercises apply single-factor shocks. - Aggregation approaches used across exercises include: - No aggregation of sensitivity results (examples: Belgium (2006), Spain, Denmark (2007)). - Simple summation (examples: Guernsey, South Africa, Isle of Man, United States, Japan (2012), Singapore (2013)). - Summation of certain combinations of single-factor shocks (examples: France (2005), Mexico). - Aggregation via correlations (sometimes complemented by simple summation, i.e., setting correlation coefficients to 1) (examples: The Netherlands, Luxembourg, Belgium (2013)). - Sole use of a correlation matrix (example: Swiss FSAP). - Single-period stresses remain the predominant modeling framework used by national supervisory authorities. - Authorities in Canada, Singapore and the United States apply multi-year period scenarios with projection horizons of up to five years; others use single-period or instantaneous shocks. - Exceptions to single-period FSAP approaches include Japan (2012) with two-year, Singapore (2013) with three-year, and Canada with five-year projection horizons. Extending to a multi-year horizon requires additional assumptions when designing scenarios. ### Advantages and caveats of multi-year scenarios - Advantages: - Identify medium- and long-term vulnerabilities from gradual erosion of solvency. - Inform remedial actions, recovery plans, and capital planning decisions. - Capture intertemporal effects of shocks (e.g., impact of lower solvency/rating downgrades on cost of funding/underwriting capacity) and mitigating factors (deferred tax assets, dividend policy, managerial actions). - Better reflect the long-term nature of most underwriting activities (except “short-tailed” non-life insurance). - Caveats: - Multi-year horizons diminish the accuracy of solvency forecasts under stress. - Effectiveness of management actions (changes in hedging, product design, dividend payouts) are difficult to model and compare across firms, risking inconsistent implementation. - FSAP exercises tend to abstract from quantitative assessment of mitigating factors but recognize scope for managers to allocate losses among current and future benefits and equity. ### Interest-rate environment, low rates, and insurer vulnerability (summary points from Box 4) - Insurers are large investors in fixed income instruments, equity and real estate; they are vulnerable to abrupt falls in asset prices from reassessment of risk premia (which implies an increase in nominal interest rates). - Low yields increase insurers’ long-term liabilities in today’s terms. In most cases (with the exception of most non-life insurance business lines), the duration of liabilities exceeds that of available investment assets. - On the asset side, low interest rates reduce investment returns and increase reinvestment risk; this is pronounced for firms matching long-term low-risk investments to guaranteed rates of return. - For legacy life insurance books, future premiums cannot be changed to reflect lower investment returns; higher value of interest-dependent assets usually cannot compensate for the higher present value of liabilities due to the “short-long duration mismatch.”1 - Low interest rates would require insurers to either increase premiums for the same expected future claims payments or lower guarantees to policyholders. - Adverse effects vary by balance sheet structure and business type. Low interest rates may be unlikely to cause a serious solvency impact on non-life business (particularly protection-oriented lines) in absence of negative demand effects and lower expense due to low inflation expectations. - Protection-oriented life products (mortality, disability, long-term care) allow insurers to compensate lower investment returns with higher risk charges, but demand for these products is susceptible to economic conditions and likely to decline during recessions. - Insurance supervisors identify the current environment of low interest rates as a major risk for the life insurance industry (EIOPA, 2013a; Antolin and others, 2011; Swiss Re, 2012). - Quantification of capital impact from low interest rates is not straightforward; instantaneous interest rate shocks without market-consistent valuation cannot capture long-term solvency effects. - Examples of multi-year quantitative approaches: - French and others (2011): project cash flows based on existing investment portfolio and duration of insurance liabilities, assuming maturing bonds re-invested at a lower market rate and unchanged asset allocation. - Deutsche Bundesbank (2013): scenario analysis over a 10-year horizon for bonus and rebate provisions to draw conclusions about solvency ratios of 85 German life insurers. The analysis, based on Kablau and Wedow (2012), finds that “a stress scenario with a prolonged period of low interest rates, more than one-third of German life insurers would no longer be able to fulfill the regulatory own funds requirements under the current solvency regime (Solvency I) by 2023. [...] This result is attributable primarily to high guaranteed interest rates (Deutsche Bundesbank, 2013, p. 69).” ### Key risk factors to include and their modeling considerations - General point: relevance of particular risk factors depends on jurisdictional business models, products, and typical investment portfolios. Forward-looking indicators of monetary conditions (interest rates and inflation) and asset valuations in capital markets (equity and debt prices) are generally most significant. - i. Interest rate risk - One of the most important risk factors, especially for life insurers offering long-term annuities with guarantees since asset duration is usually shorter than liability duration. - Modeling approaches vary: simple parallel shifts of the term structure to advanced modeling aligned with macroeconomic projections. - In a recessionary scenario, interest rates would likely decline or remain at a low level given (expected) accommodative monetary policy; inflationary pressure could lead to upward-moving interest rates. - Generally, short-term interest rates tend to be more volatile than long-term rates (Box 4).44 - ii. Equity - Typical component in insurance stress tests; relevance of equity exposures has decreased in many countries. - Main challenge: determine shocks for diverse equity exposures (listed stocks, private equity, hedge funds, alternative asset classes) and strategic participations. - iii. Real estate - Property price shocks can be designed similarly to equity shocks; high heterogeneity across property types (residential vs. commercial, forestry, project development). - Shocks could apply to investment assets and self-used property. - iv. Foreign exchange - Exchange rate risks are usually seen as less relevant because many solvency regimes include strict matching rules for business written and investments held in foreign currencies. - Instead of multiple bilateral shocks, a general appreciation or depreciation of the local currency may be adequate. - For realistic macro scenarios, note some currencies tend to appreciate in a crisis due to a “flight to safety.” - v. Credit risk - Can be highly relevant due to investment holdings of fixed-income instruments, derivative transactions, contractual relations with reinsurers and direct lending where allowed.45 - Modeling approaches: (a) model counterparty default risk with stressed probabilities of default and losses given default (historic evidence scarce, approximations needed); (b) stress market prices of bonds by assuming higher credit spreads, with different shocks for corporate vs. sovereign bonds and consideration of potential “flight to safety.” - vi. Concentration risk - Prevalent on asset and liability sides. Stress combined exposure to a single counterparty (e.g., a reinsurer to whom business is ceded while the undertaking also holds asset exposures to the same entity). - Stress concentrated banking exposures by assuming default of largest bank counterparty and model effects across seniority of instruments; if the bank acted as a distribution channel, lower premiums could feed into the stress scenario. - vii. Liquidity/funding risk - Insurers invest premium income from long-dated gross claims and gross life assurance provisions in high-quality assets to support predictable short-term payment obligations. - An abrupt rise in frequency/severity of claims (exceptional string of large natural catastrophes) could drain liquidity and overwhelm non-life insurers’ liquidity management capacity.46 - Liquidity risks could materialize in life insurance if structural changes in claims activity and/or negative cash flows from exceptional surrender behavior (“lapse risk”) increase payment obligations above actuarial expectations. - Unexpected surrender payments due to higher lapse rates would require use of cash reserves or asset sales to meet obligations. - viii. Contagion risk - Arrangements between banking and insurance activities can be prone to contagion during severe stress (committed funding arrangements, contingent intra-group transactions). - Loss of confidence in capital markets could make banking sides of conglomerates vulnerable to large deposit withdrawals and/or run-off of liabilities; both banks and insurers sustaining sharp investment value decreases could increase reliance on intra-group transactions or contingent funding arrangements. *Source: Box 3. Recessionary Scenarios in the Insurance Sector, _wp14133* ### Box 5. Liquidity Risk in Insurance ### Box 5. Liquidity Risk in Insurance ### Liquidity risk overview - Rising liquidity risk tends to amplify the deterioration of a firm’s capital position under adverse scenarios and should be considered an essential element of insurance stress test. - In general, the long-term funding profile of insurers is less susceptible to funding shocks than banks (although such risks cannot be excluded). - Insurance companies may still have liquidity and maturity mismatches, and the duration gap tends to be negative (especially for life insurers). - Some financial transactions, such as the use of OTC derivatives for hedging and securities lending, could create short-term cash flow needs (such as high quality collateral) that are markedly different from long-term cash flow projections associated with insurance liabilities and are inherently more susceptible to the financial market effects. - In many countries, insurance regulations are imposed to limit liquidity risks, such as investment limits for loans or real estate, prohibition of certain derivatives and securities lending transactions to protect the interest of policyholders. ### Relevance for non-life insurers and reinsurers - Stress testing liquidity risk in insurance is most relevant for non-life insurance and reinsurance. - Liquidity stress tests can shed a light on specific vulnerabilities faced by reinsurers that would have to settle large claims after a major natural catastrophe. - In many countries, insurance supervisors are monitoring liquidity positions of reinsurers and non-life insurers by comparing their liquid assets with potential payment amount of large claims. - There is no well-established market practice of liquidity stress within the industry yet. - One possible approach is to make use of cash flow projections with certain stress scenarios (such as large claims from catastrophe events, lower future premiums from commercial lines in response to greater competition, and collateral needs from OTC derivative transactions). - Note: This is also relevant for life insurers experiencing a significant increase in surrender rates. ### Stress testing approaches and scenario design - Supervisory monitoring: supervisors in many jurisdictions compare liquid assets with potential payment amounts of large claims; however, a standardized industry practice for liquidity stress testing is not yet established. - Suggested stress inputs include: - large claims from catastrophe events; - lower future premiums from commercial lines in response to greater competition; - collateral needs from OTC derivative transactions. - Stress tests should consider the mismatch between short-term cash needs generated by certain transactions and long-term liability cash flows. ### Insurance-specific risks (IAIS, 2003) - i. Underwriting risk - Commercial considerations regarding the pricing and coverage of insurable interest are influenced by rapid changes in the volume of the underwriting portfolio, uncertainty of the claims experience (e.g., the volume and timing of claims), and tolerance for variations in expenses. - Dependence on intermediaries (such as brokers and securities underwriters), the possibility of higher reinsurance rates, and the effects of high pricing uncertainty in new or emerging markets (possibly complicated by insufficiently understood insurance risk and reserving requirements) represent considerable challenges to the risk management of insurers. - ii. Deterioration of technical provisions - Includes the adequacy of the technical claims and other underwriting provisions, the uncertainty of the claims experience (in terms of the frequency and size of claims), the length of the claims development (including possible outcomes relating to any disputed claims, particularly where the outcome is subject to legal proceedings), the impact of inflation, the effects of increasing longevity on pension products, the guarantees and options in policy terms, the risks of early policy surrenders which can be linked to variations in interest rates, and other social, economic, legislative and technological changes. - iii. Demographic risks - Changes in long term trends of mortality can have a significant permanent impact on the life insurance industry. While shocks do not change the underlying trend, they heavily affect both the level and volatility of mortality rates and long-term pay-outs. - iv. Catastrophe risks - Reflects the ability of insurers to withstand catastrophic events, increases in unexpected exposures, latent claims or aggregation of claims, or the possible exhaustion of reinsurance (or alternative risk transfer) arrangements, and the appropriateness of the underlying assumptions and calibrations underpinning catastrophe models. - Insurers use commercially available models to simulate and estimate the possible cost of claims arising from natural catastrophes and man-made disasters; models are based on historic claims and are constantly updated, but there is a fundamental model risk, especially for low-probability events. *Source: Box 5. Liquidity Risk in Insurance, _wp14133 - Box 5. Liquidity Risk in Insurance_* ### 1.5 percent loan loss in Japan (2003) or a 4.4 (9.4) percent loss on loans (corporate ### _wp14133 - 1.5 percent loan loss in Japan (2003) or a 4.4 (9.4) percent loss on loans (corporate ### Credit risk and bond downgrade scenarios - Historical and supervisory scenarios included: - "1.5 percent loan loss in Japan (2003)" - "4.4 (9.4) percent loss on loans (corporate bonds) in the case of Israel" - Downgrade scenarios of bond holdings (e.g., "two to four notches") were frequent during the early phase of the financial crisis (Guernsey and the Isle of Man). - Absolute shocks (e.g., "an increase by 50 bps") or relative shocks (e.g., "multiplying current spreads with a factor of 1.5") are now common for traded securities. - Differentiation of spread increases by rating class used in the Netherlands, Belgium (2013), and Singapore (2013). - Most credit risk scenarios applied only to corporate bond exposures; sovereign stress tests have been added only recently (Luxembourg and Belgium (2013)). ### Foreign exchange risk - FX shocks inclusion: - Foreign exchange risk was included in every other exercise; only half of all stress tests included an explicit shock to FX rates. - Typical assumed variation of the external value of the domestic currency: between "15 and 35 percent". - Notable severe shocks: - Netherlands: "45 percent depreciation of the euro" - Denmark: "+/- 40 percent" - South Africa: "+/- 50 percent" ### Life underwriting risk - Inclusion and calibration: - Life underwriting risk was included in fewer than half of the exercises. - Mortality shocks: mortality rates exceeding baseline assumptions by "between 15 and 30 percent". - Testing of lower-than-expected mortality rates in Spain and Portugal. - Combined tests of higher mortality and increased longevity for annuitants (Guernsey, South Africa, and Isle of Man). - Pandemic or higher morbidity tests: shocks "mostly in a range of 15 to 25 percent above the baseline assumptions" (Spain, Guernsey, South Africa, Isle of Man, and the United States). - Lapse and surrender rates during recessions included in five exercises: - Earlier specifications: general increase in lapse rates "50 percent" (Spain and Portugal) and "30 percent" (Guernsey). - Belgium (2013): higher lapse rates "(+30 percent)" assumed only where surrender value exceeded the technical provision. ### Non-life underwriting risk - Main approaches and historic scenarios: - Non-life underwriting risks were incorporated mainly via natural catastrophe scenarios. - Historic scenarios tested: - Portugal: Lisbon earthquake of 1755 - France (2005): severe windstorm in 1999, where claims were assumed to be "twice the amount of the historical claims" - Belgian FSAP (2013) included the PML expected over a "40-year risk horizon". - Other non-life stresses: - Increase of "10 percent in the cost of claims" - "15 percent higher frequency of claims greater than EUR 30,000" (Spain) - Worsening technical result and higher operating costs (Netherlands and Belgium (2006)) ### Other risk factors - Examples and augmentations: - Netherlands FSAP added a commodity shock and an increase in implied volatilities (also used in the South African FSAP). ### Aggregation of risk factors and approaches to total impact - Common practice and methodological considerations: - Risk factors are aggregated to determine joint impact of shocks. - Most supervisory stress tests combine single-factor shocks that are individually determined rather than jointly calibrated using historical sensitivities. - Aggregation approaches include use of one or more correlation matrices (embedded in several solvency regimes, most notably Solvency II), which implies random changes with a given correlation — a potential inconsistency with the shock concept. - An alternative is a linear combination of separate risk factor impacts, preserving individual risk factor impacts at high statistical confidence. - Some supervisory stress tests and FSAP exercises adopt a dual approach, assessing capital adequacy under stress with and without aggregation/diversification effects. ### Output measures and solvency metrics - Primary output and definitions: - Main output variable: change in capital adequacy due to a pre-defined shock and/or scenario over a single- or multi-period forecast horizon. - Total balance sheet approach: insurance liabilities need to be covered by assets at all times subject to risk factor impacts. - Solvency frequently defined as capacity to maintain a positive net asset value with a high level of statistical confidence, usually over a "one-year risk horizon". - Solvency ratio: excess of equity capital (assets minus liabilities, usually with some restrictions) over the prescribed capital requirement (PCR) or another national capital standard. - Many jurisdictions also have a minimum capital requirement (MCR) or a balance sheet-based minimum solvency margin (MSM) as a minimum threshold; breach triggers strongest supervisory actions (e.g., business suspension and revocation of licenses). ### Capital resources, eligibility, and interpretation under stress - Considerations on capital quality and availability: - Assessment of capital adequacy is heavily influenced by the definition of capital resources and their availability under stress. - In jurisdictions with stringent statutory requirements for current (best) estimates of technical provisions and margin requirements, it might be appropriate to include some reserves in regulatory eligible capital resources. - Including less reliable capital instruments and assets (e.g., subordinated debt, intangible assets, deferred tax assets, deferred acquisition costs) requires careful consideration of their loss absorption capacity. - When insurers rely on lower-quality capital instruments, adjustment before or after the stress test and careful interpretation of results are necessary. - Current (best) estimate defined as "the expected present value of all relevant future cash flows that arise in fulfilling insurance obligations, using unbiased, current assumptions." - ICP 17.11.34 categorizes capital into: (i) highest quality capital; (ii) medium quality capital; and (iii) lowest quality capital. ### Interpretation, granularity, and presentation of results - Reporting, granularity, and complementary metrics: - BU (bottom-up) approaches allow detailed analysis due to direct involvement of firms, enabling disaggregated contribution analysis of individual shocks and consideration of business/external factors, operational and structural considerations (e.g., policyholder participation or deferred tax assets/liabilities). - If shocks affect both available and required capital, a disaggregated view of stress test results is desirable. - Mitigating/aggravating influences (business strategy, market competition, management behavior) are often assessed qualitatively; management actions and hedging are recognized but authorities often require reporting results without recognizing those actions. - Aggregated reporting levels (due to confidentiality) increase importance of presentation format; segmentation into life, non-life, and reinsurance is common. - Results should be reported for each year of the forecast horizon with measures of dispersion, such as the inter-quartile range (between the "25th and the 75th percentile"). - Contribution of different risk drivers, risk mitigation effects, and recognized diversification effects should be shown. - Multi-year projections require thoughtful presentation of timing and duration of stresses and contributions of individual shocks. - Few exercises presented non-solvency figures (examples: France included post-shock policy yields; Canada analyzed impact on net income). ### Alternative valuation metrics and publication practice - Valuation alignment and disclosure: - Accounting measures (e.g., net income, return on equity, return on assets) may be inconsistent with valuation approach; use of an alternative valuation metric such as market-consistent embedded value (MCEV) is suggested (difference between market value of assets less market value of liabilities). - Publication of stress test results by insurance supervisors has generally been limited compared with banks. - EIOPA refrained from publishing firm-by-firm results of the 2011 EU stress test, noting the test "is based on a future regulatory regime and not necessarily indicative of any current solvency problems (EIOPA, 2011a)." - FSAP stress test results are published as part of FSSA reports and Technical Notes after approval by the IMF’s Executive Board in consultation with national authorities; publications present aggregated results (e.g., solvency ratios before and after stress, dispersion measures such as inter-quartile range, minimum and maximum) and are linked to sector structure and predominant business models. *Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14133.pdf* ### Box 7. Case Study: Belgium Insurance Stress Test for the FSAP ### Box 7. Case Study: Belgium Insurance Stress Test for the FSAP ### Design and scope of the exercise - Undertaken as a BU exercise to determine the capacity of the sector to absorb a combination of single-factor shocks affecting each capital component. - Covered the six largest insurers, comprising more than 70 percent of the insurance sector. - Conducted by insurers themselves in collaboration with the FSAP team and National Bank of Belgium (NBB) staff based on mid-2012 prudential data. - One insurer used end-September 2012 data due to corporate restructuring; other tests used data as of end-June 2012. - Consideration of transition from Solvency I to Solvency II and examination of regulatory reform impacts on general conditions affecting risk factors. ### Scenario design and calibration - NBB calibrated four market risk factors for a mild and a severe adverse scenario: - interest rates - equity prices - corporate spreads - sovereign spreads - Additional shocks included: - a mass lapse event in the life business - realization of the largest probable maximum losses (PML) on a single catastrophic tail event (man-made or natural) - The non-life catastrophe and the life insurance mass lapse were identical for both mild and severe scenarios. - Macro-financial linkages emphasized: interest rates and inflation, asset valuations in capital markets (equity and debt prices), and general risk aversion (credit spreads). ### Methodology and capital assessment - Insurers calculated overall capital impact by aggregating individual impacts of each risk factor using a correlation approach similar to the Solvency II standard formula. - Own funds available under each scenario were compared with the Solvency Capital Requirement (SCR) and the Minimum Capital Requirement (MCR), subject to eligibility conditions. - Exercise treated SCR and MCR as fixed reference points despite acknowledging they change during stress; main scenario effect was on own funds. - Approach combined actuarial liability revaluations and market-consistent considerations depending on insurer inputs. ### Validation of results - Emphasized due diligence for BU approaches: actuarial assessment of liabilities requires detailed underwriting portfolio knowledge and stochastic properties, difficult in TD exercises. - Benefits and risks of BU collaboration: - Benefit: greater accuracy through industry participation. - Risk: inconsistent system-wide outcomes if firms apply different assumptions without sufficient adjustments. - Recommended validation techniques: - Firm-specific validation against financial soundness indicators and prudential measures from supervisory reporting, public disclosure, surveys, and audits. - Peer group analysis to identify outliers across firms and business lines. - Time-series comparison including historic trends and previous stress test results. - Table 1 (overview of validation checks) highlights point-in-time and time-series checks, disclosed sensitivities, impact of stress, exposure, deviations in assumptions of future baseline (premiums, claims, lapses), and deviations from historic averages for RoE/RoA and dividend payout ratios. ### Communicating outcomes and disclosure considerations - Effective communication and disclosure of identified vulnerabilities are as important as design and execution. - Stress test results should be presented in a non-technical manner to attract senior decision-making attention. - Present sensitivity of outcomes to assumptions and key risk drivers as robustness checks; use standardized reporting templates when tests are regular to aid comparability. - Frequency of stress tests should reflect: - risk-sensitivity of supervisory framework - dynamics of the insurance sector - influence of external changes on validity of results - During periods of stress or greater market uncertainty, run frequent sensitivity analyses and/or additional ad hoc stress tests based on updated data. - Periodically update stress scenarios to reflect changing risk factors and insurance market trends to avoid firms optimizing for a single predictable scenario. - Balance publication of results against potential negative market reactions; complement disclosure with credible backstops and recovery/resolution arrangements, including a readily available fund to protect policyholders of companies that show capital shortfall under stress. - Empirical precedent: - SCAP 2009 (United States) was well-received where credible recapitalization commitment existed. - CEBS 2009 exercise failed to allay investor concerns where relevant risks were excluded and backstops were absent (Ong and Pazarbasioglu, 2013). ### Representative assumptions and parameter magnitudes (from FSAPs and national approaches) - Coverage and completion: - BU completion is common; TD used occasionally to complement BU. - Typical firm coverage in small countries: 4-6 largest firms; in larger countries more firms (e.g., 30 firms in the United States). - Valuation basis: - Mostly statutory accounting; sometimes market-consistent valuation applied. - Risk horizon: - Single-period (one year) stress horizon common; more recently multiple periods (2-3 years) in some exercises. - Asset shocks (examples and frequencies): - Equity risk: About -30% (but up to -50%) (H frequency). - FX risk: Around +/-20% (but up to +/-50%) (H frequency). - Real estate risk: About -20% (but up to -50%) (H frequency). - Interest rate risk: About +/-200bps parallel shift (H frequency); national approaches often about +/-100bps parallel shift (H frequency). - Credit spreads: Increase of credit spreads by up to 50%, downgrade of counterparties by 2-4 notches (H frequency). - Liability and underwriting shocks: - Life underwriting—mortality/morbidity/longevity: mortality/morbidity/longevity of annuitants (about +25% each); occasional testing of pandemic (M frequency). - Lapse/surrender rates: mass lapse of about 25% (but up to 50%) (L frequency). - Non-life natural catastrophe: usually set to maximum historical claims experience, such as 1-in-50 years probable maximum loss (PML) (L frequency for FSAPs); national approaches often use peak risk such as 1-in-200 years (M frequency). - Other non-life underwriting shocks: large variation; usually around +10% average cost of claims and +15% higher frequency of claims (L frequency). - Risk mitigation and aggregation: - Reinsurance and hedging treated often on a firm-specific basis; risk aggregation/diversification effects applied via correlation assumptions or simple summation of single-factor shocks (M frequency). *Source: IMF (Box 7 text as provided).* ### 7. Ri sk fa c to r s 1 / ### 7. Ri sk fa c to r s 1 / ### IV. DISCUSSION AND CONCLUSION - System-wide stress tests for insurance have increased in importance since the end of the last financial crisis. - Systemic relevance of insurance companies is generally different (and in most jurisdictions smaller) than that of banks, but interlinkages with banks and other financial institutions may increase through products, markets and organizational arrangements. - Enhancements are warranted in supervisory processes, risk management, and flexible approaches to resolvability to minimize adverse externalities. - Most stress testing remains focused on the viability of individual institutions to instantaneous shocks rather than system-wide robustness to joint risk-factor impacts, despite: - (i) growing complexity of interconnections among insurance companies and with other financial institutions; and - (ii) the extent to which interlinkages cause potential spillover and contagion effects. - Movement toward market-consistent valuation within solvency regimes means stress test results can inform thematic reviews of vulnerabilities and help integrate stress testing into the supervisory framework. - A more integrated approach would ideally be based on a common framework for banking and insurance stress testing, or at least consistent assumptions. - Recent FSAPs (Belgium, Canada, United States) show closer coordination between banking and insurance stress testing. - Developing common scenarios requires considerable effort because of complementary balance sheet differences and transmission channels, notably: - (i) different time horizons used in specifying stresses (instantaneous/single-factor shocks in insurance vs. scenario-based/multi-period sensitivity to multiple risk drivers in banks); - (ii) different (and in the extreme, opposite) sensitivities to the same shocks; - (iii) risk factors with adverse scenarios for insurance include many liability-side risks in addition to asset-side risks affecting both insurers and banks. - There is a case for alignment with pension fund exercises (example: Israel (IMF, 2012a)). - Illustrative contrast: For instance, a positive shock to interest rates tends to generate higher levels of solvency among insurers, especially long-term underwriters, whereas the opposite holds true for banks. ### Key issues affecting the design, calibration, and interpretation of insurance stress tests - i. Risk-factor evolution over time: - Calibration premised on comprehensive assessment of industry trends, interconnections (including non-traditional and non-insurance activities in insurance groups), and capital market conditions. - Understanding differences in business models and behavioral characteristics under stress is fundamental to assessing transmission channels. - ii. Valuation methodology robustness: - Impact of shocks depends on valuation methodologies, whose robustness may be undermined by the rare, non-recurring events that generate systemic risks. - Quantitative measures and actuarial valuation models based on robust statistics may be limited when prices and parameters do not converge to long-term expectations. - iii. Trade-off between accuracy and timeliness: - Historical sensitivity of sample firms to macro-financial shocks is essential for assessing combined impacts over a forecast horizon. - Reliance on past experience can increase predictability but may induce hindsight bias; early warnings gain accuracy only as realization becomes more probable, limiting recalibration flexibility. - Trend toward more precise and consistent assessment: - Convergence of regulatory standards and supervisory practices enhances model precision. - IAIS work on a global solvency regime and the development/field-testing of a basic capital requirement (BCR) for G-SIIs by the end of 2014 will influence future risk-based global insurance capital standards (IAIS, 2013e). - Introduction of Solvency II in the EU influences stress test design toward more shocks, higher granularity, and convergence in taxonomy and methodology. ### Considerations for evolving macroprudential stress testing of insurance - i. Total balance sheet approach: - Most stresses impact both assets and liabilities; assess using a total balance sheet approach. - Interest rate shocks modeling common in market-consistent liability valuation, but other channels (e.g., inflation linked to claim patterns, credit insurance claims during recessions) must be considered. - Calibration of liability shocks often requires weighing prescribed parameters against firms’ own modeling assumptions. - ii. Aggregation and diversification assumptions: - Aggregate risk impact should not include diversification benefits among risk factors except where economically plausible. - Accounting for dependence structure is reasonable, but combining multiple risk factors with diversification effects complicates reliable capital assessment. - Simple aggregation of risk-factor impacts preserves stochastic assumptions; frequent use of correlation may underestimate potential losses. - Use of non-linear dependence can provide more reliable joint tail risk insights (Jobst, 2013b). - iii. Flexibility through combinations and magnitudes: - Data on single risk-factor impacts can measure impacts of different combinations on system-wide capital adequacy with varying statistical confidence. - Non-linearity in price changes of certain products (financial derivatives, embedded options, non-proportional re-insurance contracts) must be considered. - iv. Beyond solvency—profitability and liquidity measures: - Accounting measures (e.g., net income, profitability) inform dynamics of solvency buffers and decisions on dividends/bonuses. - Liquidity measures provide insights when investment assets become illiquid or when severe claims shocks occur (e.g., mass lapse events, catastrophes). - v. Multi-period scenarios: - Extending single-period shocks to multi-period scenarios helps identify medium- to long-term vulnerabilities from gradual solvency erosion and informs remedial actions and recovery plans. - Market-consistent valuation may not fully capture uncertainty of future cash outflows for long-tail liabilities (e.g., asbestos-related claims recognized decades later). - vi. Parallel banking and insurance stress tests: - May include different risk factors but should be based on the same target variables defined by general economic deterioration. - A common metric allows integrated system-wide analysis and conglomerate-level analysis; supplementary sensitivity analyses help for the less-affected sector. - vii. Intra-group transactions: - Impact of scenarios on intra-group transactions should be assessed by comparing stress-testing results; consolidated reporting alone can miss material intra-group effects. - viii. Secondary impacts: - Deteriorating financial positions can lead to higher cost of capital, constrained capital mobility, limited underwriting, rating downgrades—link secondary impact analysis to contingency, recovery and resolution planning. - ix. Avoiding distortive supervisory incentives: - Stress-testing frameworks should avoid inducing uneconomic insurer behavior (e.g., reducing asset duration, increasing maturity mismatches, altering product design to influence stress outcomes). - Variation of stress test scenarios over time reduces incentive to "game" tests. - x. Role of qualitative analysis and expert judgment: - Despite advances in risk analytics, qualitative elements (business strategy dynamics, underwriting practice evolution, innovation in risk transfer) require periodic reassessment. - Granular data will never be sufficient for reliably modeling tail risk; expert judgment remains crucial. - Qualitative analysis should include reputational risk, competitive environment, and existing risk controls that influence gross impact of risks. *Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14133.pdf* ### Appendix I―Tables ### Appendix I―Tables (_wp14133 - Appendix I―Tables) ### Overview of Insurance Stress Tests in FSAPs (Table A1) - FSAP missions and FSSA/TN publications (selected entries): - Japan: June 2002, October 2002, March 2003 — FSSA: September 2003 (IMF, 2003) - Singapore: November 2002, July/August 2003, September 2003 — FSSA: April 2004 (IMF, 2004c) - Netherlands: October/November 2003, March 2004 — FSSA: September 2004 (IMF, 2004a) - France: February 2004, May 2004 — FSSA: November 2004 (IMF, 2004b); TN: June 2005 (IMF, 2005) - Belgium: December 2004, March 2005 — FSSA: February 2006 (IMF, 2006a) - Spain: June/July 2005, October/November 2005 — FSSA: June 2006 (IMF, 2006b); TN: June 2006 (IMF, 2006c) - Denmark: November 2005, May 2006 — FSSA: October 2006 (IMF, 2006d); TN: March 2007 (IMF, 2007b) - Mexico: February/March 2006 — FSSA: October 2006 (IMF, 2006e); TN: May 2007 (IMF, 2007f) - Portugal: December 2005, May 2006 — FSSA: October 2006 (IMF, 2006f); TN: January 2007 (IMF, 2007a) - Switzerland: November 2006 — FSSA: June 2007 (IMF, 2007d); TN: June 2007 (IMF, 2007e) - Bermuda: June 2007 — FSSA: October 2008 (IMF, 2008b) - South Africa: May 2008 — FSSA: October 2008 (IMF, 2008a) - Isle of Man: September 2008 — FSSA: September 2009 (IMF, 2009a); TN: September 2009 (IMF, 2009b) - United States: October/November 2009, February/March 2010 — FSSA: July 2010 (IMF, 2010a); TN: July 2010 (IMF, 2010b) - Guernsey: March 2010 — FSSA: January 2011 (IMF, 2011a); TN: January 2011 (IMF, 2011b) - Luxembourg: November 2010 — FSSA: June 2011 (IMF, 2011e) - Israel: November 2011 — FSSA: April 2012 (IMF, 2012a); TN: April 2012 (IMF, 2012c) - Japan: November/December 2011, March 2012 — FSSA: August 2012 (IMF, 2012d) - France: January 2012, June 2012 — FSSA: December 2012 (IMF, 2012e) - Belgium: November 2012, January 2013 — FSSA: May 2013 (IMF 2013a); TN: May 2013 (IMF, 2013b) - Singapore: May 2013, July/August 2013 — FSSA: November 2013 (IMF, 2013d) - Canada: June 2013, September 2013 — FSSA: February 2014 (IMF, 2014a); TN: March 2014 (IMF, 2014b) - Notes: - FSSA = Financial System Stability Assessment - TN = Technical Note on Stress Testing ### Overview of National Supervisory Stress Testing Approaches (Table A2) - Reference documents and national approaches (selected entries): - Austria: Financial Sector Assessment Program Update, Technical Note—Factual Update and Analysis of the IAIS Insurance Core Principles (IMF, 2008c) - Bermuda: Stress/Scenario Analysis (Class 4, Class 3B and Insurance Groups) and Stress/Scenario Analysis (Class 3A) (BMA, 2013a and 2013b; Appendix Box 8) - Canada: Stress Testing Guideline (OSFI, 2009) - Czech Republic: Models for Stress Testing in the Insurance Sector (Komárková and Gronychová, 2012) - Denmark: Financial Sector Assessment Program - Detailed Assessment of Observance of the Insurance Core Principles (IMF, 2007c) - European Union (EIOPA): Specifications for the 2011 EU-wide Stress Test in the Insurance Sector (EIOPA, 2011b) - Germany: Conducting of Stress Test (BaFin, 2004) - Guernsey: Stress Testing of the Guernsey Insurance Sector (Guernsey Financial Services Commission, 2011) - Japan: Supervisory Guidance for Insurers (JFSA, 2013) - Singapore: ERM Notice (MAS, 2011), Stress Testing on Financial Condition of Life Direct Insurer (MAS, 2013) - Switzerland: White Paper of the Swiss Solvency Test (Swiss Federal Office of Private Insurance, 2004) - United Kingdom: Stress and Scenario Testing (FSA, 2008) - United States: Own Risk and Solvency Assessment (ORSA) (NAIC, 2013) ### IMF FSAPs: Specification of Insurance Stress Testing (Table A3 — Key dimensions and selected country specifications) - General notes: - Source: IMF. - TD = top-down, BU = bottom-up. - Footnotes used in table: 1/ based on gross premium income; 2/ based on insurance liabilities; 3/ only policies where lapses result in loss; 4/ PML = probable maximum loss. - Common structure in FSAP insurance stress testing: - 1. Scope - Approaches used: TD and BU (examples show both BU and TD or combined BU/TD). - Coverage reported as numbers of life, non-life, and composite insurers; relevance often reported as a percent (e.g., "86% (life, based on assets)" for Japan; "77% (life), 45% non-life /154% (based on assets)" for Singapore). - Reporting basis: solo and/or consolidated. - Data sources: prudential, public, or a combination; some entries list "n.a." where not available. - 2. Valuation Basis - Assets and liabilities valuation: often "n.a." or specified as "market-consistent" (e.g., Spain and Switzerland use market-consistent). - 3. Scenario Design - Macro-financial linkage/transmission: typically "Combination of single (instantaneous) shocks" or "scenario analysis + single factor shocks". - Risk horizon: typically "single period"; several FSAPs use multi-year horizons (e.g., Japan: 2 years; Israel: single period; Canada: 5 years; Singapore: 3 years; France: single period). - 4. Risk factors (selected calibrated shocks from sample countries) - Assets — Credit risk: - Japan: 1.5% credit loss on loan book - Singapore: -2% and -7% in corporate bond prices (Singapore and other SE Asian markets); -2% and -3% rest of the world - Netherlands: credit spreads +30% and +50% for investment grade; +40% and +60% for speculative grade; +25% implied volatility - France: credit spreads +50 bps - Spain: realization of implied PD of Basel II risk-weights; alternatively credit spread shock - Switzerland: credit spread (+50 bps) - Assets — Equity risk: - Japan: -20% - Singapore: -10% and -20% (Singapore and other SE Asian markets); -5% rest of the world - Netherlands: -25% and -40% for developed countries; -30% and -50% for developing countries and private equity; +25% implied volatility - France: -30% - Switzerland: +/-35% - Canada: between -21% and -49% in first year (for Canada entry in later table) - Assets — FX risk: - Singapore: between -3% and +3% against various currencies - Netherlands: 30% and 45% depreciation of the EUR; +25% implied volatility - Spain: +/- 40% - Mexico: +/- 15% - Switzerland: +20% (CHF against EUR, GBP and USD) - Assets — Real estate risk: - Singapore: -10% and -20% (Singapore commercial); -5% rest of the world - Netherlands: -10% and -20% - France (2013 entry): between -34% and -54% over three years - Guernsey: -20% to -15% - Assets — Interest rate risk: - Japan: +100 bps parallel shift - Singapore: -60 bps in short-term rates (with unchanged long-term rates); +150 bps in short-term rates (long-term rates +50 bps) - Netherlands: +/-100 bps parallel shift; +/-200 bps parallel shift; +25% implied volatility - France: +/-100 bps parallel shift; +300 bps parallel shift; steepening of yield curve; flattening and upward shift of yield curve - Spain: +/- 200 bps parallel shift; steepening; flattening - Mexico: +250 bps parallel shift; -100 bps parallel shift - Switzerland: -190 bps in domestic interest rates; -114 bps in foreign interest rates - United States (2010): +/- 200-300 bps parallel shift of entire yield curve (later entries include country-specific variations) - Liabilities — Life underwriting shocks: - Mortality/morbidity/longevity: varied use; some entries include mortality (+/- 15%) (Spain); mortality (+25%), morbidity (+25%) and longevity (+25%) in some jurisdictions (Isle of Man, Guernsey, Luxembourg entries). - Lapse/surrender rates: examples include +50% (Spain), +/-50% (Mexico), +30% (US Guernsey entry), +20% in first year (Canada 2014 entry). - Liabilities — Non-life underwriting shocks: - Natural catastrophe: - Belgium: doubling the claims of a storm event in 1999 - Spain: earthquake with a probability of 1-in-250 years - Bermuda, Isle of Man, Guernsey: explicit inclusion or "yes" for catastrophe in various entries - Other non-life underwriting shocks: increases in claim costs and frequencies (Spain: +10% average cost of claims; +15% higher frequency of claims > 30,000 EUR). - Other risks and operational/technical items: - Singapore (2004): receivables: -10% and -20%; loans and other receivables -5% and -10% - Netherlands: 50% increase in worst technical result in last 5 years; 50% increase in maximum cost in last 5 years; -30% and -45% in commodity prices; +25% implied volatility in commodities - Mexico: premium shock: zero nominal premium growth; loss rate increases (life 10%, accidents and health 13%, P&C 20%, auto 5% and catastrophe 10%) - Risk aggregation / diversification: - Methods vary: simple summation; aggregation with correlation (e.g., correlation of 0.5 or 0 between shocks); use of correlation matrices similar to QIS2/QIS5 or Solvency II; Swiss Solvency Test correlation matrix used in Switzerland. - 5. Regulatory capital standards (examples) - Solvency I (Netherlands, France earlier entries) - Swiss Solvency Test (SST) (Switzerland) - Solvency II SCR (Portugal and some reporting) - RMM (Bermuda) - RBC (United States) - Singapore Risk Based Capital; MCCSR (Canada) - Statutory accounting and market-consistent frameworks used depending on country and exercise - 6. Presentation of results (selected presentations) - Dispersion measures: distribution of losses as percentage of shareholder equity; min/max impact on solvency ratio; distributions of solvency ratios before and after shock; boxplots with company data points; anonymized company-by-company data. - Contribution of individual shocks: impact of individual shocks on shareholder equity, solvency ratio, available capital, required capital, and risk-bearing capital. - Other results: number of companies below regulatory thresholds (e.g., number of companies with solvency ratios after stress below 100% and recapitalization needs as percent of liabilities), capital shortfall as percent of market solvency requirement, number of companies below 300 percent RBC, number of companies with negative net asset value. ### Selected national specifications and methodological notes (Table A4 highlights and notes) - Sources cited in table: BMA, CNB, EIOPA, FINMA, and IMF. - Notes and methodological points: - TD = top-down, BU = bottom-up. - Switzerland participates in EIOPA stress test based on the same specification and scenarios but applies the Swiss Solvency Test (SST) for capital assessment. - Credit spread scenario calibration note: The credit spread scenario is calibrated annually to historical price changes of rating-specific baskets of credit default swaps (CDS) with maturity terms of three years at a statistical confidence of 99th percentile over an estimation period starting on 1 January 2006. - CTE = conditional tail expectation (a generic term for "Tail VaR" or "Expected Shortfall"). - Sovereign risk shock: considered from a creditor perspective by examining potential valuation changes and impairment charges of mark-to-market and hold-to-maturity assets; haircuts calculated from expected valuation changes of liquid government (benchmark) bonds, assuming an increase of sovereign distress but not a general shift in the yield curve. - Economic view to interest rate risk: in addition to other approaches, the economic view to the interest rate risk of assets and liabilities is applied in some exercises. *Italic source: Appendix I―Tables, _wp14133 - Appendix I―Tables* ### 1. Scope ### _wp14133 - 1. Scope ### 1. Scope (coverage, relevance, reporting, frequency) - Approaches: BU; BU; BU; BU; BU; BU (table of multiple national supervisory approaches). - Coverage examples listed: - entire sector - all large commercial (re)insurers [Classes 4 and 3B] - entire sector (but small insurers as well as term life and unit-linked insurers can be exempted) - large and middle-sized insurers - entire sector - major European (re)insurance entities (200 firms) - Relevance of the coverage examples: - 100% - 100% - n.a. - 90% of gross premium written - 100% - 50% of gross premium written - Reporting Basis examples: - solo - solo/consolidated - n.a. - solo - n.a. - consolidated (excl. banking activities) - Frequency examples: - semi-annual for life and health insurers, annual for non-life insurers - annual - annual - annual - annual - annual - Additional national examples (Germany, Guernsey, Japan, Singapore, Switzerland, United Kingdom (PRA), United States (NAIC)): - Coverage: most insurance firms (but small insurers may be exempted); 6 life with liabilities > GBP 50 mln and 22 non life firms with gross premium earned > GBP 15 mln; cell companies are included.; All insurers, re-insurers and branches; all insurers; all insurers under supervision; major life insurers; all life and health (re)insurers. - Relevance examples: 88% 5/; unspecified; 100%; 100%; 100%; n.a.; 100%. - Reporting Basis examples: solo; n.a.; n.a.; solo/consolidated; solo/consolidated/granular; solo/consolidated; solo (legal entity). - Frequency examples: annual (TD), quarterly (BU) 6/; ad hoc; n.a.; annual; annual/semiannual; annual; annual. ### 2. Valuation Basis (assets and liabilities) - Assets valuation examples: - statutory accounting - statutory accounting - statutory accounting - statutory accounting 4/ - statutory accounting - statutory accounting - Liabilities valuation examples: - statutory accounting - statutory accounting - statutory accounting - statutory accounting 4/ (rough estimation of the change in the deficiency provision in life insurance) - statutory accounting - market-based approach of best estimate of liability projected into financial statements - Additional national examples: - Assets: statutory accounting; statutory accounting; Not specified; statutory accounting; market-consistent; statutory accounting; statutory accounting. - Liabilities: statutory accounting; statutory accounting; Not specified; statutory accounting; market-consistent; statutory accounting; statutory accounting. ### 3. Scenario Design (source, specification, risk horizon, confidence level) - Source of scenarios: - provided by supervisor; prescriptive shocks (multiple entries) - mostly principles-based with some prescriptive shocks - Specification of shocks, macro-financial linkage/transmission: - combination of single (instantaneous) factor shocks; no specification of general macroeconomic conditions - combination of single (instantaneous) factor shocks; contains macro-financial linkages of insurance and capital market shocks (two adverse scenarios (depression and loss of confidence)) - combination of single (instantaneous) factor shocks; baseline and adverse scenarios with 0% inflation change and a single inflationary scenario - combination of single (instantaneous) factor shocks; contains macro-financial linkages for some jurisdictions - some entries: mostly single (instantaneous) factor shocks; no specification of general macroeconomic conditions - Risk Horizon examples: - single period (stress is assumed to occur at the end of a one-year horizon) - single period - multiple periods (5 years (life), 3 years (non-life)) - single period - Confidence level examples: - n.a. - 99.0% Tail-VaR (99.0% CET) - n.a. (multiple entries) - Additional national examples: - Sources: provided by supervisor; prescriptive shocks; general guidelines but principle-based approach; historical and hypothetical shocks/scenarios; provided by supervisor; prescriptive (standardized) shocks; deterministic scenarios prescribed by regulator; stochastic scenarios generated by prescribed scenario generator. - Macro-financial linkage/transmission channel(s): combination of single (instantaneous) factor shocks; combination of (instantaneous) multiple factor shocks; some jurisdictions include single period and multiple period (3 years); one jurisdiction uses over lifetime of liabilities. - Risk Horizon national examples: single period; single period; not specified; single period and multiple period (3 years); single period; single period; over lifetime of liabilities. - Confidence level national examples: n.a.; n.a.; n.a.; n.a.; Probability in between 0.1% and 1% for prescriptive scenarios; n.a.; 70% CET for reserves and 90% CET for capital. ### 4. Regulatory capital standards - Listed standards: - Solvency I - Bermuda Solvency Standard (BSCR) - Minimum Continuing Capital and Surplus Requirements (MCCSR) - Solvency I - Solvency I - Solvency II/Swiss Solvency Test (SST) - Additional national examples: - German Solvency I - Minimum Capital Requirement of licensed insurers - Solvency Margin Ratio - Singapore Risk Based Capital - Swiss Solvency Test (SST) - Individual Capital Adequacy Standards (ICAS) - Risk-based Capital (RBC) ### 5. General comment (methodological and supervisory notes) - General observations from table entries: - contains no macro-financial specifications and amounts to a sensitivity analysis - high comprehensiveness on technical (underwriting) risks; sensitivity analysis exercises similar to Solvency II tests - approach relies on dynamic financial analysis (DFA) completed by firms, which use employ DFA techniques to model the uncertainty of insurance operations (including scenarios and subsequent responses) - additional features can be incorporated in one-year risk horizon, such as the profit/loss produced during the year, the repeated occurrence of natural disasters, and planned dividend payments - "traffic-light" system with a yellow and a red scenario; missing the thresholds of either scenario is directly linked to heightened supervisory scrutiny; yellow scenario suspended since Q3 2008 - Contains no macro-financial specifications and amounts to a sensitivity analysis; Switzerland conducted stress tests together with EIOPA based on the same scenarios (but the Swiss Solvency Test (SST) for capital assessment) - Additional national comments: - Contains no macro-financial specifications (other than market risk shocks) and amounts to a sensitivity analysis. - Simple model with selected single factor shocks. - Publicly available information limited to stress testing; approach is heavily reliant on firm-generated stress scenarios. - Publicly available information limited to stress testing for life insurers; approach is heavily reliant on internal models and firm-generated stress scenarios. - High comprehensiveness on technical (underwriting) risks.* - Aimed to evaluate the resilience of major life insurance groups to market stresses of progressive severity in order to understand the nature of the market risks which the groups are exposed to. - Stress testing is done as test on the adequacy of statutory formula reserves; at start of projection, it is assumed that assets equal liabilities (e.g., surplus is zero); metric is the present value of surplus at end of the projection period. ### 6. Output (metrics reported post-stress) - Reported outputs examples: - post-stress effect on solvency ratio (full impact, full impact net of hidden reserves, full impact net of hidden reserves and the equalization reserve) - post-stress effect on statutory assets and liabilities - statutory ratio post-stress either positive or above minimum depending on scenario - post-stress effect on Solvency I ratio and the ability to cover technical provisions with a sufficient volume of assets; economic view to the interest rate sensitivity of assets and liabilities - solvency ratio post-stress - reduction of own funds and comparison to MCR - Additional national outputs: - Asset coverage ratio over liabilities - Margin of solvency has to be higher than the minimum margin of solvency - n.a. - Risk based capital ratio post-stress - Aggregation of several scenarios to the capital requirement (target capital) - Post-stress impacts: If reserves are adequate surplus is positive at end of projection. If surplus is negative additional reserve has to be established. ### 7. Risk factors (assets, liabilities, non-life underwriting, mitigation, aggregation) - Macro scenario: - many entries: no specification of general macroeconomic conditions - Two scenarios, baseline and adverse for the downward movement of interest rates and rate of inflation affecting non-life claims volumes. - included, but no consistent specification of general macroeconomic conditions across firms - scenarios combining substantial equity, property, credit and yield shifts in some jurisdictions - Asset risk factors (examples and shocks where specified): - Credit risk: yes; (-5% for A- to BBB-; -20% for non-IG bonds); yes (rating class-specific increase of credit spreads); yes - Equity risk: yes (-20% and -35%); yes (-40%); yes (-35%, +15%, -10% and level thereafter); yes (yellow: -30%, red: -12%); yes (up to -15%); yes (stochastic) - FX risk: yes (up to -20% relative to major currencies) in some entries; other entries: — or n.a. - Real estate risk: yes (-20%); entries with (-10%); yes (yellow: -12%, red: -8%); yes (up to 11.6% for residential, up to 25% for commercial) - Interest rate risk: yes (bonds -5% and -10%); 3-month T-bill (-10 bps), long-term (-170 bps); entries with deterministic & stochastic approaches; yellow: +/- 100 bps, red: +/- 70 bps - Sovereign risk: yes in some entries; sovereign shock considered from a creditor perspective (haircuts calculated from expected valuation changes of liquid government (benchmark) bonds, assuming an increase of sovereign distress but not a general shift in the yield curve) 3/; government bond spread DNK-DEU in one entry. - Other assets: hedge funds -40% in one entry; other entries: — or yes for specific asset classes. - Liabilities / Life underwriting risk factors: - Mortality/morbidity/longevity: yes in several jurisdictions; maximum exposure to mortality and longevity shocks in one entry - Pandemic: yes in some entries; — in others - Lapse/surrender rates: included in some jurisdictions; — in others - Reinsurance: n.a. in some entries; yes in others; treatment varies by jurisdiction - Other risk factors: expense persistency, cash flow mismatch, new business (renewal), off-balance sheet items; rating downgrade (up to two notches); regulatory and political risk; second-order effects (managerial and regulatory reaction); management actions - Example: health insurance (being similar to life insurance): increase in claims by 7.5% - Non-life underwriting risk factors: - Natural catastrophe: yes in many jurisdictions; Lloyd’s RDS scenarios for P&C/own worst case scenarios; maximum exposure to two specific 1/200 year scenarios (with reinsurance allowed to be included discounted by 70%); inflation shock to claims reserves. - Other risk factors: higher frequency of claims in various lines of business; terrorism; frequency, severity, reinsurance, premium volume, misestimating liabilities, pricing; premium risk for casco/motor third party liability insurance; shock to net written premiums following a macroeconomic model; claims inflation. - Risk mitigation and aggregation: - Risk mitigation: reinsurance and hedging considered variably; some jurisdictions permit completion with and without hedging assumption; other entries treat reinsurance implicitly in internal models. - Risk aggregation/diversification effects: various aggregations of stresses, generally no diversification effects; limited approximation of peak exposure based on combined impact from the three largest underwriting risks; implicit in internal models underpinning DFA approach; aggregation by simple summation, no diversification effects applied; other entries: limited (only inflation impact on P&C scenarios). *Sources: BaFin, FINMA, IMF, NAIC, Bank of England (PRA), BMA, CNB, EIOPA, FINMA, and IMF (table compilation).* * _wp14133 - 1. Scope_* ### Box 8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of ### Box 8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of Bermuda ### Overview - Bermuda hosts the third largest (re)insurance market in the world, with globally active commercial underwriters focused on property and casualty risks. - The Bermuda Monetary Authority (BMA) requires all large firms to perform an annual stress test and submit results with their Capital and Solvency Return. - Objective: assess capital adequacy of legal entities and groups by evaluating impact of risk drivers conditional on plausible firm- and system-wide scenarios, providing understanding of loss absorbing capacity relative to shocks to asset prices, interest rates, and projected underwriting losses on the insurer’s/group’s statutory balance sheet (statutory admitted assets, admitted liabilities, and capital and surplus). ### Stress testing framework and reporting requirements - Stress testing is conducted at the firm level, either via firms’ internal models or through prescriptive shocks to risk factors in accordance with uniform guidelines and assumptions. - Exercises are based on either internal model-derived or pre-defined scenarios affecting both single entity and group-wide annual solvency return. - Firms must submit: - A description of all key assumptions and calculations used to arrive at final results. - Post-stress/scenario positions on aggregate statutory assets and liabilities immediately upon occurrence of the event, both with and without the effect of reinsurance and/or other loss mitigation instruments. - Results that comprise both the occurrence return period (e.g., 1-in-50 year event) and the relative return period using the underlying loss distribution of the aggregate net probable maximum loss. ### Economic (financial risk) scenarios (prescribed shocks) - Asset risk single-factor shocks triggered by an adverse global macroeconomic scenario: - A severe decline in equity prices of 40 percent without allowance for diversification across the markets (assume all markets are correlated and only long (asset) positions are affected). - A widening of credit spreads. - Negative shocks to asset and net open foreign currency positions (assuming a depreciation of the U.S. dollar vs. major reserve currencies). - Valuation haircuts on fixed income holdings (and long derivative positions) of sovereign debt and financial bonds (debt securities and loans), including a general upward shift in the yield curve of 50 basis points. - Calibration details referenced in the exercise for the current reporting year: - Credit spreads widen between 163 basis points for “AAA”-rated securities to 3,188 basis points for securities rated “BB” or lower. - The adverse scenario (“through-the-cycle”) is calibrated to historical price changes of rating-specific baskets of benchmark CDS with three-year maturity at a statistical confidence of 99th percentile. - The magnitude of negative foreign-exchange shock is determined based on four times the difference between the maximum implied annualized volatility of each currency (Euro, Japanese yen, Pound sterling, Swiss franc, and Australian dollar with the U.S. dollar as reference currency) between 1 Jan. 2008 and end-2011 and the long-term average since 1 Jan. 2005. - Sovereign risk shock comprises valuation changes and impairment charges applied to all net exposures and is calculated from expected valuation changes based on the 99th percentile of the historical density of one-/three-/five-/seven- and ten-year forward contracts on CDS with maturity terms between one and ten years. ### Underwriting (insurance risk) scenarios - Multiple underwriting risks affecting aggregates in-force at the beginning of the reporting period: - Prescribed property and casualty events in different scenario groupings (U.S. windstorm, U.S. earthquake, Non-U.S. windstorm, Non-U.S. earthquake, aerospace/aviation, and marine) as specified in Lloyd’s Handbook on Realistic Disaster Scenarios. - Non-peak perils not currently in vendor models (U.S. oil spill, U.S. tornadoes, Australian flooding, and Australian wildfires). - Additional insurance risks (pandemic, terrorism). - Other underwriting scenarios if prescribed perils do not apply or produce de minimis loss projections. - Projections from the worst-case annual aggregate catastrophe loss scenario, which combines economic and underwriting loss scenarios generating the largest losses and a series of loss simulations relating to extreme tail events. - A qualitative assessment of a rating downgrade by two notches (or falling below a “A-” rating, whichever is more severe) on income and liquidity positions, including relative impact/severity of collateral requirements, loss payment triggers on in-force policy contracts, claw-backs, and/or other adverse financial and liquidity implications of the downgrade. - All lines of business and exposures are included in final loss estimates net of protection such as reinsurance, retrocessional agreements, or insurance-linked securities. - Firms may substitute own worst-case scenarios for prescriptive scenarios and the BMA encourages internal stress testing tailored to firms’ risk appetite and profile. ### IMF 2007 system-wide solvency stress test (case study results) - In 2007, the IMF completed a BU system-wide solvency stress test of ten large commercial (re)insurance and long-term insurance companies as part of its Offshore Financial Center Assessment Program. - Companies used a combination of internal and vendor models to calculate capital impacts of underwriting scenarios: three natural catastrophe events, two pandemic events, and worst-case scenarios of aggregate net probable maximum loss as specified by each insurer. - Impact assessed against statutory reporting requirements at the time (change in capital and surplus, minimum regulatory premium ratio, and minimum regulatory loss reserve ratio). - Findings: - Catastrophic events would have had a significantly negative impact on aggregate capital, with the most severe impact resulting from the worst-case scenarios (only two of which included economic events in addition to natural catastrophes). - Scenarios combining catastrophic events and an economic recession had the greatest impact on solvency positions on average. - No firm failed to meet applicable regulatory capital requirements under any of the scenarios. - The exercise covered only a subset of risk factors necessary for a comprehensive assessment; incomplete coverage of financial market effects and use of accounting data rather than economic valuation were identified as shortcomings. ### Developments, shortcomings addressed, and future enhancements - Shortcomings identified in the 2007 exercise were remedied in the first stress testing guidance under the risk-based solvency regime (BSCR) introduced by the BMA in 2010. - In particular, sensitivity of some firms to combinations of extreme financial and underwriting events motivated the introduction of the worst-case annual aggregate catastrophe loss scenario. - After extending the stress testing framework to other classes of insurers (smaller commercial (re)insurers (Class 3A) and long-term business), future enhancements include: - Capacity of the BMA to execute system-wide industry-level stress testing on a regular basis. - Expanded treatment of catastrophe risk scenarios based on more granular data. *Box 8. National and IMF Stress Testing for Non-life (Re) insurance—A Case Study of Bermuda* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14133.pdf_