IFoA SA4 cheat sheet

Lloyd's operates as a subscription market: risks are syndicated across multiple syndicates, each taking a share, with a lead syndicate setting terms; the chain of security runs through members, syndicates and central assets.

The underwriting cycle links premium rates to prior loss experience and capital availability: soft markets follow surplus capital and produce rate erosion; hard markets follow capital depletion (large catastrophe or investment losses) and produce rate hardening.

Combined ratio = loss ratio + expense ratio; a ratio below 100% indicates an underwriting profit before investment income. Calendar-year and accident-year ratios differ from ultimate ratios because of reserve development.

Capital intensity differs by line: catastrophe-exposed property and liability lines consume more capital per unit of premium than short-tail, high-frequency personal lines, so returns must be judged on return on allocated capital, not loss ratio alone.

Distribution economics matter: broker, agency, MGA and direct channels carry different acquisition cost ratios and different levels of underwriting control; delegated authority requires explicit underwriting guidelines and audit rights.

Reinsurance and retrocession capacity, plus Insurance-Linked Securities and catastrophe bonds, expand peak-risk capacity; pricing of these instruments feeds back into primary rates for catastrophe-exposed business.

Demand-side drivers include exposure growth (sum insured inflation, urban concentration), legal and social inflation, and coverage penetration; supply-side drivers include capital, reinsurance pricing and regulatory capital rules.

Market cycles are not synchronised across lines or territories; specialty lines can harden while personal motor remains soft, so portfolio mix drives the group result more than any single market view.

Solvency II structure: own funds must exceed the Solvency Capital Requirement (SCR), which itself must exceed the Minimum Capital Requirement (MCR); breach of the SCR triggers supervisory intervention, breach of the MCR triggers withdrawal of authorisation.

SCR is calibrated to a 1-in-200 one-year value-at-risk of basic own funds; it may be computed by the standard formula or an internal model subject to supervisory approval.

Two capital tiers apply: Tier 1 (ordinary share capital, retained earnings, surplus funds) is highest quality; Tier 3 (deferred tax assets, net deferred tax) is limited in its eligible amount to cover SCR and MCR.

Risk margin under Solvency II is the cost of holding non-hedgeable SCR (and MCR) capital for the run-off of obligations; cost-of-capital rate is 6% per annum in the standard calibration.

Risk margin under Solvency II is the cost of holding non-hedgeable SCR (and MCR) capital for the run-off of obligations; cost-of-capital rate is 6% per annum in the standard calibration.

Prudential conduct splits: prudential regulation covers capital, governance and reporting; conduct regulation covers fair treatment of customers, product oversight and governance, and claims handling standards.

Regulatory reporting includes quantitative reporting templates (QRTs), the Solvency and Financial Condition Report (SFCR) and the Regular Supervisory Report (RSR); the ORSA is the insurer's own forward-looking assessment of capital adequacy.

Cross-border and group supervision require equivalence assessment of third-country regimes; group solvency is calculated on a consolidated or aggregation-and-deduction basis.

Regulatory change risk (capital rules, IFRS 17, climate disclosure, operational resilience rules) is a strategic risk requiring scenario planning, not merely a compliance cost.

Technical premium = expected loss cost + expected expense + cost of capital + risk margin, adjusted for reinsurance cost and investment income; the pure premium equals frequency times average severity.

Burning cost = historical losses (net of excess) divided by historical exposure, adjusted for trend and development; it is the standard starting point for excess-of-loss and catastrophe layers.

Experience rating credibility: credibility factor Z is a function of claim count and between-claim variance; the credibility-weighted estimate is Z times the insured's own experience plus (1 - Z) times the class rate.

Deductibles reduce frequency cost more than proportionally to the deductible amount because small claims are removed; co-insurance and aggregate limits cap the tail and must be reflected in the loss cost.

Policy terms that change risk: basis of valuation (reinstatement vs indemnity), claims-made vs occurrence trigger, retroactive dates, exclusions and warranties; each alters expected loss and must be priced.

Underwriting authority must be delegated with explicit limits (line size, aggregate, classes) and monitored through bordereaux review; pricing adequacy is tested by rate monitoring and technical pricing versus written rate.

Exposure measures differ by line: payroll for employers' liability, turnover for business interruption, vehicle-years for motor, total insurable value for property; wrong exposure bases distort loss ratios.

Adverse selection arises when insureds know more about their risk than the insurer; risk classification, mandatory coverage and anti-selection loadings mitigate but do not eliminate it.

Chain-ladder: ultimate = latest cumulative paid or incurred divided by the cumulative development factor; assumes past development patterns persist and applies best to stable, mature portfolios.

Bornhuetter-Ferguson: ultimate = reported to date + expected ultimate loss ratio times unearned exposure times (1 - proportion reported); it blends expectation with experience and is more stable for immature years.

Mack chain-ladder gives a distribution-free estimate of prediction error; bootstrap and Bayesian methods (e.g. Bayesian stochastic reserving) give full predictive distributions needed for capital.

Reserve risk drivers: case reserve adequacy, IBNR versus IBNER split, claim handling speed, legal environment, inflation (including social inflation) and changes in mix or limits.

Case reserves are set claim by claim; IBNR covers claims incurred but not reported and development on known claims; total reserves equal case plus IBNR plus ULAE.

Capital modelling aggregates risk distributions using copulas or correlation matrices; the sum of standalone capital overstates diversified capital, and the diversification benefit must be evidenced.

Catastrophe models produce an exceedance probability curve from hazard, vulnerability and financial modules; model uncertainty and non-modelled perils require explicit loading.

Back-testing and validation: compare previous model outputs with realised experience, test sensitivity to assumptions, and document limitations for the model use test.

Quota share: cedant cedes a fixed percentage of premium and losses; proportional reinsurance reduces both volatility and expected profit, and the ceding commission offsets acquisition cost.

Surplus treaty: cession is proportional to the amount by which the risk exceeds the cedant's retention, subject to line capacity; it is exposure-based rather than a flat percentage.

Excess of loss: reinsurer pays losses above the cedant's retention up to the limit; the cedant's retained loss is capped per occurrence and the layer price depends on expected loss and exposure to large events.

Aggregate stop loss and cat covers protect the annual result rather than a single risk; reinstatements (free or paid) restore limit and must be priced into the layer cost.

Reinsurance credit risk: recoverables are an asset subject to counterparty default; collateral, trust funds and rating requirements mitigate, and Solvency II applies a counterparty default risk charge.

Risk transfer adequacy: finite or financial reinsurance that transfers insufficient risk may be treated as financing, not insurance, for accounting and regulatory capital purposes.

Capital management levers: reinsurance purchase, dividend policy, capital raising (equity or debt), retrocession and risk transformation via ILS; each trades expected return against capital relief.

Optimal retention balances the cost of reinsurance against the reduction in capital and volatility; retention should be set where marginal capital relief per unit of premium equals the marginal cost of cover.

IFRS 17 measurement: the liability for incurred claims (LIC) and liability for remaining coverage (LRC) equal fulfilment cash flows plus the contractual service margin (CSM); the CSM is unearned profit released over coverage.

Under IFRS 17, the general measurement model uses a risk adjustment for non-financial risk and a discount rate; the premium allocation approach is a simplified option for short-duration contracts.

Insurance revenue under IFRS 17 excludes investment components and is recognised as coverage is provided; insurance service result = insurance revenue less incurred claims and expenses.

Best-estimate liabilities use discounted expected cash flows including claim handling costs; the risk margin or risk adjustment is added separately, and the discount rate must reflect the timing of claims.

Profit emergence differs by basis: accident-year underwriting result, calendar-year result including reserve development, and IFRS 17 insurance service result; comparisons across bases are invalid without reconciliation.

Key performance metrics: combined ratio, return on capital, loss ratio, expense ratio, reserve development as a percentage of opening reserves, and investment yield versus discount rate.

Discounting matters for long-tail liability: a rise in discount rates reduces the liability and creates a one-off gain, while a fall increases it; the effect is amplified by duration.

Segment reporting and embedded value analysis require allocation of capital, expenses and investment income to lines of business to reveal true profitability.

Risk taxonomy for a general insurer: underwriting, reserving, catastrophe, market (interest rate, credit spread, equity, property), credit, liquidity, operational, cyber, climate and strategic risk.

Three lines of defence: business units own risk (first line), risk and compliance functions oversee (second line), internal audit provides independent assurance (third line).

Risk appetite is expressed as quantitative tolerances (e.g. probability of breaching SCR, earnings-at-risk, maximum cat loss) plus qualitative statements; it must be cascaded to underwriting and investment limits.

ORSA requires a forward-looking assessment of capital adequacy over the business plan horizon, including stress and reverse stress tests, and a statement of overall solvency needs.

Board responsibilities: approve risk appetite, strategy and capital plan; ensure the Actuarial Function opines on technical provisions, reinsurance adequacy and underwriting policy.

Operational risk controls include segregation of duties, delegated authority limits, business continuity and outsourcing oversight; cyber risk needs scenario analysis and aggregation with other risks.

Climate risk splits into physical risk (acute and chronic perils) and transition risk (policy, technology, litigation); both require scenario analysis and disclosure.

Model risk management: independent validation, documentation, governance of model changes, and a model inventory with owners and review cycles.

Frame the problem: identify the decision, the relevant time horizon, the materiality threshold and the stakeholders before selecting methods; method choice follows the decision, not the reverse.

Triangulate: compare multiple methods (chain-ladder, Bornhuetter-Ferguson, frequency-severity, expected loss ratio) and explain divergences rather than averaging blindly.

Distinguish data limitations from model limitations: sparse data calls for credibility weighting and expert judgement; structural change (new scheme, new law) calls for scenario analysis.

Test sensitivity to key assumptions (development factors, loss ratios, discount rates, cat model version) and report the range, not a single point estimate.

Document expert judgement explicitly, with rationale, impact and review date, so that it can be challenged and audited.

Beware of spurious precision: present results to a number of significant figures appropriate to the uncertainty; a reserve with 20% uncertainty does not need four decimal places.

Consider second-order effects: reinsurance recoveries, tax, capital and dividend capacity all change when the reserve or pricing estimate changes.

Structure for the audience: conclusion first, then key assumptions, then evidence; non-technical readers need the decision and the range of outcomes, not the derivation.

State uncertainty explicitly: give a central estimate, a range and the drivers of the range, and separate model uncertainty from parameter uncertainty.

Use plain language for technical terms on first use, and avoid jargon such as 'IBNR' without a one-line explanation for non-actuarial readers.

Charts should show the distribution or range, not just a point; label axes, state units and avoid misleading truncated scales.

Peer review and sign-off: material results should be reviewed by another actuary, and the review scope and conclusions documented.

For regulatory and board papers, link results to capital, risk appetite and the ORSA, and flag any breach or near-breach of tolerances.

Answer the question asked: if asked for a reserve, give the reserve and its uncertainty; if asked for advice, state the recommended action and its consequences.

The Actuaries' Code is built on the principles of integrity, competence and care, impartiality, and compliance; it applies to all actuarial work, including work outside a formal actuarial role.

Actuaries must act with integrity, be honest about their competence and limitations, and not mislead; they must decline work outside their competence or obtain expert input.

Conflicts of interest must be identified, disclosed and managed; independence of judgement is required, and undue pressure from employers or clients must be reported.

Confidentiality: information obtained in the course of work must not be disclosed without authority, subject to legal and regulatory obligations and the public interest.

Professional standards and TASs (Technical Actuarial Standards) require proportionate, transparent and comprehensible work; the actuary must document data, assumptions and methods.

Whistleblowing: where an actuary believes a breach of the Code or a legal requirement has occurred, they must raise it through the appropriate channel, escalating to the IFoA if unresolved.

CPD and record-keeping: actuaries must maintain competence and keep records sufficient for another actuary to review the work.

Ethical decision test: would the actuary be comfortable if the decision, and its reasoning, were disclosed publicly to policyholders and the regulator?

Lloyd's operates as a subscription market: risks are syndicated across multiple syndicates, each taking a share, with a lead syndicate setting terms; the chain of security runs through members, syndicates and central assets.

The underwriting cycle links premium rates to prior loss experience and capital availability: soft markets follow surplus capital and produce rate erosion; hard markets follow capital depletion (large catastrophe or investment losses) and produce rate hardening.

Combined ratio = loss ratio + expense ratio; a ratio below 100% indicates an underwriting profit before investment income. Calendar-year and accident-year ratios differ from ultimate ratios because of reserve development.

Capital intensity differs by line: catastrophe-exposed property and liability lines consume more capital per unit of premium than short-tail, high-frequency personal lines, so returns must be judged on return on allocated capital, not loss ratio alone.

Distribution economics matter: broker, agency, MGA and direct channels carry different acquisition cost ratios and different levels of underwriting control; delegated authority requires explicit underwriting guidelines and audit rights.

Reinsurance and retrocession capacity, plus Insurance-Linked Securities and catastrophe bonds, expand peak-risk capacity; pricing of these instruments feeds back into primary rates for catastrophe-exposed business.

Demand-side drivers include exposure growth (sum insured inflation, urban concentration), legal and social inflation, and coverage penetration; supply-side drivers include capital, reinsurance pricing and regulatory capital rules.

Market cycles are not synchronised across lines or territories; specialty lines can harden while personal motor remains soft, so portfolio mix drives the group result more than any single market view.

Solvency II structure: own funds must exceed the Solvency Capital Requirement (SCR), which itself must exceed the Minimum Capital Requirement (MCR); breach of the SCR triggers supervisory intervention, breach of the MCR triggers withdrawal of authorisation.

SCR is calibrated to a 1-in-200 one-year value-at-risk of basic own funds; it may be computed by the standard formula or an internal model subject to supervisory approval.

Two capital tiers apply: Tier 1 (ordinary share capital, retained earnings, surplus funds) is highest quality; Tier 3 (deferred tax assets, net deferred tax) is limited in its eligible amount to cover SCR and MCR.

Risk margin under Solvency II is the cost of holding non-hedgeable SCR (and MCR) capital for the run-off of obligations; cost-of-capital rate is 6% per annum in the standard calibration.

Prudential conduct splits: prudential regulation covers capital, governance and reporting; conduct regulation covers fair treatment of customers, product oversight and governance, and claims handling standards.

Regulatory reporting includes quantitative reporting templates (QRTs), the Solvency and Financial Condition Report (SFCR) and the Regular Supervisory Report (RSR); the ORSA is the insurer's own forward-looking assessment of capital adequacy.

Cross-border and group supervision require equivalence assessment of third-country regimes; group solvency is calculated on a consolidated or aggregation-and-deduction basis.

Regulatory change risk (capital rules, IFRS 17, climate disclosure, operational resilience rules) is a strategic risk requiring scenario planning, not merely a compliance cost.

Technical premium = expected loss cost + expected expense + cost of capital + risk margin, adjusted for reinsurance cost and investment income; the pure premium equals frequency times average severity.

Burning cost = historical losses (net of excess) divided by historical exposure, adjusted for trend and development; it is the standard starting point for excess-of-loss and catastrophe layers.

Experience rating credibility: credibility factor Z is a function of claim count and between-claim variance; the credibility-weighted estimate is Z times the insured's own experience plus (1 - Z) times the class rate.

Deductibles reduce frequency cost more than proportionally to the deductible amount because small claims are removed; co-insurance and aggregate limits cap the tail and must be reflected in the loss cost.

Policy terms that change risk: basis of valuation (reinstatement vs indemnity), claims-made vs occurrence trigger, retroactive dates, exclusions and warranties; each alters expected loss and must be priced.

Underwriting authority must be delegated with explicit limits (line size, aggregate, classes) and monitored through bordereaux review; pricing adequacy is tested by rate monitoring and technical pricing versus written rate.

Exposure measures differ by line: payroll for employers' liability, turnover for business interruption, vehicle-years for motor, total insurable value for property; wrong exposure bases distort loss ratios.

Adverse selection arises when insureds know more about their risk than the insurer; risk classification, mandatory coverage and anti-selection loadings mitigate but do not eliminate it.

Chain-ladder: ultimate = latest cumulative paid or incurred divided by the cumulative development factor; assumes past development patterns persist and applies best to stable, mature portfolios.

Bornhuetter-Ferguson: ultimate = reported to date + expected ultimate loss ratio times unearned exposure times (1 - proportion reported); it blends expectation with experience and is more stable for immature years.

Mack chain-ladder gives a distribution-free estimate of prediction error; bootstrap and Bayesian methods give full predictive distributions needed for capital.

Reserve risk drivers: case reserve adequacy, IBNR versus IBNER split, claim handling speed, legal environment, inflation (including social inflation) and changes in mix or limits.

Case reserves are set claim by claim; IBNR covers claims incurred but not reported and development on known claims; total reserves equal case plus IBNR plus ULAE.

Capital modelling aggregates risk distributions using copulas or correlation matrices; the sum of standalone capital overstates diversified capital, and the diversification benefit must be evidenced.

Catastrophe models produce an exceedance probability curve from hazard, vulnerability and financial modules; model uncertainty and non-modelled perils require explicit loading.

Back-testing and validation: compare previous model outputs with realised experience, test sensitivity to assumptions, and document limitations for the model use test.

Quota share: cedant cedes a fixed percentage of premium and losses; proportional reinsurance reduces both volatility and expected profit, and the ceding commission offsets acquisition cost.

Surplus treaty: cession is proportional to the amount by which the risk exceeds the cedant's retention, subject to line capacity; it is exposure-based rather than a flat percentage.

Excess of loss: reinsurer pays losses above the cedant's retention up to the limit; the cedant's retained loss is capped per occurrence and the layer price depends on expected loss and exposure to large events.

Aggregate stop loss and cat covers protect the annual result rather than a single risk; reinstatements (free or paid) restore limit and must be priced into the layer cost.

Reinsurance credit risk: recoverables are an asset subject to counterparty default; collateral, trust funds and rating requirements mitigate, and Solvency II applies a counterparty default risk charge.

Risk transfer adequacy: finite or financial reinsurance that transfers insufficient risk may be treated as financing, not insurance, for accounting and regulatory capital purposes.

Capital management levers: reinsurance purchase, dividend policy, capital raising (equity or debt), retrocession and risk transformation via ILS; each trades expected return against capital relief.

Optimal retention balances the cost of reinsurance against the reduction in capital and volatility; retention should be set where marginal capital relief per unit of premium equals the marginal cost of cover.

IFRS 17 measurement: the liability for incurred claims (LIC) and liability for remaining coverage (LRC) equal fulfilment cash flows plus the contractual service margin (CSM); the CSM is unearned profit released over coverage.

Under IFRS 17, the general measurement model uses a risk adjustment for non-financial risk and a discount rate; the premium allocation approach is a simplified option for short-duration contracts.

Insurance revenue under IFRS 17 excludes investment components and is recognised as coverage is provided; insurance service result = insurance revenue less incurred claims and expenses.

Best-estimate liabilities use discounted expected cash flows including claim handling costs; the risk margin or risk adjustment is added separately, and the discount rate must reflect the timing of claims.

Profit emergence differs by basis: accident-year underwriting result, calendar-year result including reserve development, and IFRS 17 insurance service result; comparisons across bases are invalid without reconciliation.

Key performance metrics: combined ratio, return on capital, loss ratio, expense ratio, reserve development as a percentage of opening reserves, and investment yield versus discount rate.

Discounting matters for long-tail liability: a rise in discount rates reduces the liability and creates a one-off gain, while a fall increases it; the effect is amplified by duration.

Segment reporting and embedded value analysis require allocation of capital, expenses and investment income to lines of business to reveal true profitability.

Risk taxonomy for a general insurer: underwriting, reserving, catastrophe, market (interest rate, credit spread, equity, property), credit, liquidity, operational, cyber, climate and strategic risk.

Three lines of defence: business units own risk (first line), risk and compliance functions oversee (second line), internal audit provides independent assurance (third line).

Risk appetite is expressed as quantitative tolerances (e.g. probability of breaching SCR, earnings-at-risk, maximum cat loss) plus qualitative statements; it must be cascaded to underwriting and investment limits.

ORSA requires a forward-looking assessment of capital adequacy over the business plan horizon, including stress and reverse stress tests, and a statement of overall solvency needs.

Board responsibilities: approve risk appetite, strategy and capital plan; ensure the Actuarial Function opines on technical provisions, reinsurance adequacy and underwriting policy.

Operational risk controls include segregation of duties, delegated authority limits, business continuity and outsourcing oversight; cyber risk needs scenario analysis and aggregation with other risks.

Climate risk splits into physical risk (acute and chronic perils) and transition risk (policy, technology, litigation); both require scenario analysis and disclosure.

Model risk management: independent validation, documentation, governance of model changes, and a model inventory with owners and review cycles.

Frame the problem: identify the decision, the relevant time horizon, the materiality threshold and the stakeholders before selecting methods; method choice follows the decision, not the reverse.

Triangulate: compare multiple methods (chain-ladder, Bornhuetter-Ferguson, frequency-severity, expected loss ratio) and explain divergences rather than averaging blindly.

Distinguish data limitations from model limitations: sparse data calls for credibility weighting and expert judgement; structural change (new scheme, new law) calls for scenario analysis.

Test sensitivity to key assumptions (development factors, loss ratios, discount rates, cat model version) and report the range, not a single point estimate.

Document expert judgement explicitly, with rationale, impact and review date, so that it can be challenged and audited.

Beware of spurious precision: present results to a number of significant figures appropriate to the uncertainty; a reserve with 20% uncertainty does not need four decimal places.

Consider second-order effects: reinsurance recoveries, tax, capital and dividend capacity all change when the reserve or pricing estimate changes.

Structure for the audience: conclusion first, then key assumptions, then evidence; non-technical readers need the decision and the range of outcomes, not the derivation.

State uncertainty explicitly: give a central estimate, a range and the drivers of the range, and separate model uncertainty from parameter uncertainty.

Use plain language for technical terms on first use, and avoid jargon such as 'IBNR' without a one-line explanation for non-actuarial readers.

Charts should show the distribution or range, not just a point; label axes, state units and avoid misleading truncated scales.

Peer review and sign-off: material results should be reviewed by another actuary, and the review scope and conclusions documented.

For regulatory and board papers, link results to capital, risk appetite and the ORSA, and flag any breach or near-breach of tolerances.

Answer the question asked: if asked for a reserve, give the reserve and its uncertainty; if asked for advice, state the recommended action and its consequences.

The Actuaries' Code is built on the principles of integrity, competence and care, impartiality, and compliance; it applies to all actuarial work, including work outside a formal actuarial role.

Actuaries must act with integrity, be honest about their competence and limitations, and not mislead; they must decline work outside their competence or obtain expert input.

Conflicts of interest must be identified, disclosed and managed; independence of judgement is required, and undue pressure from employers or clients must be reported.

Confidentiality: information obtained in the course of work must not be disclosed without authority, subject to legal and regulatory obligations and the public interest.

Professional standards and TASs (Technical Actuarial Standards) require proportionate, transparent and comprehensible work; the actuary must document data, assumptions and methods.

Whistleblowing: where an actuary believes a breach of the Code or a legal requirement has occurred, they must raise it through the appropriate channel, escalating to the IFoA if unresolved.

CPD and record-keeping: actuaries must maintain competence and keep records sufficient for another actuary to review the work.

Ethical decision test: would the actuary be comfortable if the decision, and its reasoning, were disclosed publicly to policyholders and the regulator?