Build liability-relative judgement through 20 original written practice tasks and 60 worked concepts covering insurance and pension ALM. Practise surplus sensitivities, liquidity analysis, capital-framework comparisons, scenario design, derivatives and risk transfer. Coverage follows the supplied November 2026/March 2027/July 2027 syllabus, applicable to the next administration as of October 16, 2026; no later version is assumed. These are representative exercises, not official questions or a full mock examination. Written tasks are a practice choice because the supplied evidence does not establish the examination's response format. Syllabus citations establish topic relevance, while technical explanations use established subject knowledge. Model answers and rubrics make calculations and reasoning assessable.
What this guide covers
Twenty original CP 351 practice tasks and 60 concepts covering ALM objectives, risk measurement and hedging.
Objectives of Asset Liability Management
1. Economic surplus and liability-relative performance
Economic surplus is the difference between consistently valued assets and liabilities. Its sensitivity depends on both sides of the balance sheet, so an attractive asset return can still accompany deteriorating financial capacity. ALM translates the institution's obligations into investment objectives and constraints, including payment timing, capital resilience and acceptable surplus volatility. The chosen valuation basis matters: market-consistent surplus and accounting equity need not move together. Apply surplus analysis by identifying the relevant liability benchmark, revaluing both sides under common scenarios and separating changes caused by markets from those caused by experience or management actions.
Worked example: An insurer's assets rise from $200 million to $208 million while liabilities rise from $180 million to $191 million. Assets earn a positive return, yet surplus falls from $20 million to $17 million because obligations increase faster.
Common mistake: A positive investment return does not establish successful ALM when liability values or payment demands have increased by more.
Official syllabus reference: [1]
2. Economic value, earnings and liquidity objectives
ALM objectives can concern economic value, reported earnings, regulatory capital or the ability to make payments. These measures respond differently to the same transaction because recognition rules, discounting conventions and cash timing differ. A hedge may reduce economic exposure while creating earnings volatility if its valuation changes are recognized differently from the hedged liability. Liquidity measures track accessible cash rather than net present value. A useful objective hierarchy specifies the primary measure and constrains the others explicitly. Conflicting measures should be reconciled through a decision framework rather than combined into an unexplained single target.
Worked example: A receive-fixed swap offsets an insurer's liability sensitivity to falling rates. Its changing fair value appears immediately in one reporting measure, while the liability responds differently. Management assesses the hedge's economic benefit alongside the resulting earnings variability and collateral needs.
Common mistake: Reducing volatility in one reporting measure does not guarantee lower economic risk or improved ability to meet cash demands.
Official syllabus reference: [1]
3. Maturity transformation and rollover exposure
Maturity transformation arises when asset and liability cash-flow horizons differ. Funding long assets with short or callable liabilities creates rollover exposure: the institution may need new funding before assets mature. Conversely, backing long liabilities with short assets creates reinvestment uncertainty because future proceeds must be invested at unknown rates. Average maturity alone is insufficient when cash flows contain options or concentrations. Assess the mechanism through dated inflows and outflows, renewal assumptions and liquidation capacity. A spread earned from maturity transformation compensates for exposure; it does not eliminate the possibility that funding or reinvestment conditions deteriorate.
Worked example: A fund invests in ten-year private loans using contracts renewable annually. Even if the loans perform as expected, a renewal decline next year can force asset sales or replacement funding well before the loans repay their principal.
Common mistake: Matching total asset and liability values does not resolve a mismatch between contractual payment dates or renewal conditions.
Official syllabus reference: [1]
4. Forced-sale feedback and confidence-sensitive liabilities
A liquidity event can become a solvency event through forced-sale feedback. Withdrawals require cash; selling illiquid assets quickly may realize discounts; those losses weaken capital or confidence; and weakened confidence can trigger further withdrawals. Contractual triggers, concentrated counterparties and common policyholder incentives can accelerate the loop. The relevant lesson from institutional failures is to connect these mechanisms rather than stress each independently. Analysis should distinguish estimated asset value from realizable proceeds over a stated time window and include the consequences of management actions on ratings, funding terms and subsequent liability behaviour.
Worked example: An institution faces $30 million of withdrawals but holds only $10 million of accessible cash. Selling loans with estimated value of $25 million yields $20 million under time pressure, creating a realized loss that may worsen confidence among remaining contract holders.
Common mistake: Assuming asset sales occur at unstressed model values can conceal the feedback connecting liquidity pressure with capital deterioration.
Official syllabus reference: [1]
5. Risk taxonomy and causal risk mapping
A risk taxonomy groups exposures into categories such as market, credit, insurance, liquidity, operational and model risk. Categories improve accountability, but the same event can propagate across several categories. Causal mapping records the initiating event, transmission channels and affected financial measures, avoiding both omissions and double counting. For ALM, identify underlying drivers separately from their consequences: a rate rise is a driver, increased surrender is a behavioural response, and forced-sale loss is a consequence. Use the taxonomy to assign ownership while retaining integrated scenarios that capture the connected balance-sheet effects.
Worked example: A rating downgrade changes collateral terms and activates institutional withdrawal rights. The map links credit deterioration to liquidity demands, asset disposals and potential market losses. Separate owners can manage each control while the combined scenario measures the institution's total exposure.
Common mistake: Placing an event in one risk category does not mean its effects stop at that category's organizational boundary.
Official syllabus reference: [1]
6. Surrender values as embedded policyholder options
A surrender option lets a policyholder terminate a contract in exchange for its specified surrender value. Its economic attraction depends on continuation value, competing yields, surrender charges, taxation and individual needs. When market rates rise, guaranteed surrender values may decline less than backing bond values, creating a potentially adverse combination for the insurer. Exercise is behavioural rather than automatically optimal, so valuation and liquidity models require realistic assumptions about response strength and timing. Distinguish the contractual exercise payment from the probability of exercise, and evaluate how both affect liability duration and cash requirements.
Worked example: A policy promises a $50,000 surrender payment while its backing bond falls to $46,000 after rates increase. If competing savings yields attract the holder, surrender converts a future obligation into an immediate payment exceeding the asset's current sale value.
Common mistake: Treating surrender solely as a fixed lapse percentage misses how the option's attraction changes with market and contract conditions.
Official syllabus reference: [1]
7. Guaranteed annuity conversion options
A guaranteed annuity conversion option allows accumulated funds to purchase retirement income on predetermined terms. Its value increases when those terms are more favourable than the current cost of providing the promised annuity. Lower interest rates and improved survival can both increase that cost, creating exposure to interacting financial and demographic drivers. Exercise may also involve selection because people expecting longer lifetimes can find annuitization more attractive. Evaluate the guarantee using its actual conversion basis, eligible dates and benefit form. A fund-value projection alone does not measure the additional liability created by favourable guaranteed terms.
Worked example: A retirement contract converts each $100,000 into $7,000 of annual lifetime income. If current pricing would provide only $5,500 on otherwise comparable terms, exercising the guarantee creates additional value for the holder and an additional obligation for the insurer.
Common mistake: The account balance is not the complete liability when a guaranteed conversion rate can purchase benefits above current market terms.
Official syllabus reference: [1]
8. Pension lump-sum elections as embedded options
A pension contract may let members exchange future pension payments for an immediate lump sum. When contractual conversion terms adjust differently from the economic cost of the pension, this election creates an embedded option. Its exercise changes both liability value and payment timing, so ALM objectives should include financing member choices alongside funding scheduled pensions. Projections should reflect conversion terms, election dates and plausible member behaviour. Exercise need not follow financial value alone: health, household circumstances and preferences also matter. An annuity-only projection can therefore understate settlement costs and near-term cash requirements.
Worked example: An annuity-only projection values each member’s pension at $160,000, but members can elect a fixed $180,000 lump sum immediately. If 100 members elect cash, settlement requires $18 million, exceeding the projected annuity value of $16 million by $2 million and accelerating funding needs.
Common mistake: Assuming a lump-sum election merely changes payment format overlooks differences between contractual conversion terms and economic pension values, as well as accelerated cash demands.
Official syllabus reference: [1]
9. Risk appetite translated into ALM limits
Risk appetite states how much uncertainty an institution is willing and able to bear in pursuit of its objectives. ALM makes it operational through measurable limits, escalation rules and decision authority. Useful limits can address surplus sensitivity, cash coverage under specified stresses, concentration and hedge implementation capacity. A limit requires a defined metric, valuation basis, horizon and treatment of management actions. Limits should also reflect interactions: a strategy that reduces rate risk may consume liquidity. Governance connects breaches to concrete choices, while periodic review checks whether the metrics still represent the institution's obligations and resources.
Worked example: A board approves a maximum modelled surplus loss under a specified curve stress and a separate minimum cash buffer for collateral. A proposed swap overlay passes the surplus test but fails the cash test, prompting a smaller overlay or additional liquidity.
Common mistake: A broad statement that risk should remain moderate is not an operational limit without a measure, horizon and decision rule.
Official syllabus reference: [1]
Measuring Risks from Assets and Liabilities
10. Cash-flow ladders and cumulative funding gaps
A cash-flow ladder places expected asset inflows and liability outflows into time buckets. The cumulative balance reveals when the institution needs funding, even if the total horizon ends with a surplus. Separate contractual flows from contingent withdrawals, collateral and discretionary actions. Bucket choice matters because a positive monthly total can hide a payment shortfall early in the month. Incorporate settlement dates and accessible opening cash, then stress both the amounts and timing. A ladder measures funding requirements under its assumptions; it does not establish that projected asset sales or new borrowing will be executable.
Worked example: A fund starts with $3 million cash, pays $8 million in week one and receives $7 million in week three. Its month-end balance is positive, but the ladder identifies a $5 million funding need before the receipt arrives.
Common mistake: Netting all inflows and outflows over a long horizon can hide an earlier cash shortfall that still requires funding.
Official syllabus reference: [1]
11. Modified duration, DV01 and dollar exposure
Modified duration approximates the proportional change in value for a small yield change, with fixed cash flows and a specified yield convention. DV01 converts this sensitivity into currency per basis point: approximately value × modified duration × 0.0001. Under the positive-loss convention, value change is minus DV01 times the rate increase in basis points. Surplus exposure depends on asset DV01 minus liability DV01. Always state whether DV01 is signed or reported as a positive magnitude. This local measure does not capture convexity, curve shape changes or exercise-driven changes in cash flows.
Worked example: A $50 million bond portfolio with modified duration six has approximate DV01 of $30,000. A 25-basis-point yield increase reduces its value by about $750,000 before convexity, providing a direct dollar measure for comparison with liability sensitivity.
Common mistake: Equal durations do not imply equal dollar sensitivities when asset and liability present values differ materially in size.
Official syllabus reference: [1]
12. Convexity and second-order surplus effects
Convexity measures the curvature of value with respect to yield. A second-order approximation adds 0.5 × convexity × squared yield change to the duration term. For ordinary fixed cash flows, positive convexity improves value relative to the duration-only estimate in either direction. Surplus convexity depends on dollar convexities on both sides, so individually positive convexities can still produce negative surplus curvature. Embedded options can alter the sign and magnitude. Use consistent rate units and valuation conventions, and recognize that larger shocks may require full revaluation because higher-order and behavioural effects become important.
Worked example: Assets and liabilities each equal $100 million and have equal duration, but their convexities are 60 and 90. For a one-percentage-point parallel move, the second-order surplus effect is approximately negative $150,000 regardless of the move's direction.
Common mistake: Matching duration removes only first-order exposure; it does not make unequal asset and liability convexities disappear under larger shocks.
Official syllabus reference: [1]
13. Key-rate durations and interpolation conventions
Key-rate duration measures value sensitivity to a localized change at a selected curve maturity. Implementations specify how surrounding rates move through interpolation, so the node labels alone do not fully define the measure. A vector of key-rate sensitivities reveals exposures to slope and curvature that aggregate duration hides. Under compatible definitions, node sensitivities can approximately sum to parallel sensitivity. Apply the same curve, bump size and interpolation convention to assets, liabilities and hedges. Comparisons across systems can otherwise show apparent mismatches caused by methodology rather than genuine differences in economic exposure.
Worked example: An asset portfolio has excess ten-year sensitivity and deficient two-year sensitivity relative to liabilities. Its aggregate duration nearly matches, yet a curve steepening can generate a substantial loss. The key-rate vector identifies where the offsetting exposures reside.
Common mistake: Two key-rate reports are not automatically comparable merely because they use the same maturity labels on their columns.
Official syllabus reference: [1]
14. Forward-bucket sensitivities
Forward-bucket sensitivities perturb forward rates over specified future intervals rather than selected spot or par-rate nodes. Because discount factors depend on accumulated forward rates, changing one interval affects cash flows beyond that interval. This representation can align risk measurement with future reinvestment periods or segmented hedging decisions. It is a different coordinate system for curve risk, so numerical bucket exposures need not resemble key-rate exposures directly. Specify whether the bumped quantity is an instantaneous or period forward rate and preserve the curve transformation when mapping instruments into the same sensitivity framework.
Worked example: A liability paid in year twenty depends on discounting through every preceding forward interval. Raising forwards only between years five and ten changes that payment's present value even though no liability payment occurs within the bumped interval itself.
Common mistake: A forward bucket affects discounting beyond its ending date; it is not limited to cash flows paid inside that interval.
Official syllabus reference: [1]
15. Effective duration for option-sensitive liabilities
Effective duration estimates sensitivity by revaluing the instrument under upward and downward curve shocks, allowing modelled cash flows to respond. A common symmetric formula is [V_down − V_up]/[2V_initial × shock size]. This differs from modified duration applied to fixed cash flows. For insurance liabilities, the revaluation may update surrender, guarantee exercise or credited rates. The result depends on the behavioural and management rules embedded in the valuation. Test multiple bump sizes when nonlinearities matter, and distinguish a change in discount rates from a simultaneous change in all economic and behavioural assumptions.
Worked example: A liability worth $100 million is valued at $106 million after a 50-basis-point rate decrease and $96 million after an equivalent increase, with surrenders reprojected. Its effective duration is approximately ten, incorporating those modelled cash-flow responses.
Common mistake: Calling a measure effective duration does not make it behaviour-sensitive unless the shocked valuations actually update the relevant cash flows.
Official syllabus reference: [1]
16. Real-world probabilities and risk-neutral pricing
Real-world models describe prospective economic outcomes and are used for planning, earnings distributions and risk analysis. Risk-neutral models use pricing probabilities under which appropriately discounted traded gains satisfy the required martingale relationship. Their probabilities reflect market prices of risk rather than literal forecasts. The same future state can therefore receive different weights in the two models. Choose the measure according to the task: projecting how often an event occurs differs from valuing a hedgeable contingent payoff. Nontraded insurance risks require additional valuation assumptions; switching measures does not automatically supply a unique market price.
Worked example: A market model assigns a 70% real-world probability to an equity increase, but replication implies a 62.5% pricing probability. The first supports frequency projections; the second prices a replicable option consistently with the underlying and risk-free asset.
Common mistake: Risk-neutral probabilities are pricing weights, so interpreting them directly as management's best forecast of event frequencies confuses two different purposes.
Official syllabus reference: [1]
17. Consistent outer scenarios and inner valuations
An ALM projection can use real-world outer scenarios to generate future balance-sheet states and risk-neutral inner valuations to price remaining market-sensitive obligations at each state. Consistency requires the inner model to start from the outer state's yield curve, asset prices and relevant market conditions. Otherwise, measured gains can arise from incompatible valuation assumptions rather than economics. The construction also requires care with information timing: future states cannot influence earlier decisions. Nested valuation is computationally demanding, so approximations should be tested where guarantees become valuable or market relationships differ materially from the calibration region.
Worked example: A real-world scenario reaches year five with low yields and a depressed equity market. The remaining guarantee is valued using those year-five conditions, rather than restarting from today's curve and equity level, which would misstate the projected liability.
Common mistake: Using real-world outer paths and risk-neutral inner paths does not ensure consistency unless their state variables and valuation dates are aligned.
Official syllabus reference: [1]
18. Stressed asset monetization and liquidity capacity
Liquidity capacity depends on how much cash an asset can generate within the required period after discounts, settlement delays and operational constraints. Market value alone does not establish this amount. Classify assets by credible monetization routes such as maturity, sale or secured financing, and avoid counting the same holding through multiple routes simultaneously. Stress both price and execution capacity because they can deteriorate together. Distinguish unrestricted holdings from pledged or otherwise unavailable assets. A liquidity assessment should identify the sequence of actions and the cash each produces before the relevant payment deadline.
Worked example: A portfolio contains $40 million of bonds, but $15 million is pledged and the remaining bonds incur a 10% stressed sale discount. Before settlement delays, only $22.5 million is potentially available from selling the unrestricted holdings.
Common mistake: Adding market values of all holdings overstates liquidity when assets are pledged, slow to settle or costly to sell quickly.
Official syllabus reference: [1]
19. Collateral calls as contingent cash outflows
Derivative collateral converts changing market values into cash or eligible-asset requirements under the contract's terms. Variation margin generally tracks current exposure, while initial margin addresses potential future exposure under the relevant arrangement. An economically effective hedge can still require collateral when its value falls, even if the offsetting liability gain is noncash. Model thresholds, eligible collateral, settlement timing and potential changes in requirements using actual contract assumptions. Netting is available only within applicable arrangements. Compare projected calls with accessible resources rather than total assets, and avoid treating unrealized gains elsewhere as immediately spendable cash.
Worked example: A pension rate hedge loses $12 million as yields rise, while liabilities decline by a similar amount. Funding remains broadly stable, but the derivative requires a cash transfer. The liability reduction does not itself provide money for that transfer.
Common mistake: An offsetting liability valuation gain does not automatically fund a derivative collateral call that must be settled in cash.
Official syllabus reference: [1]
20. Swiss Solvency Test and changing capital requirements
FINMA’s Swiss Solvency Test assesses Swiss insurers through an economic, risk-sensitive capital framework. Its solvency ratio compares risk-bearing capital with target capital. Economic asset and liability valuations affect available resources, while market, credit and insurance exposures influence required capital, including their dependence. ALM analysis should examine both sides of this ratio under stress: a shock can reduce capital resources while increasing target capital. Comparing an SST ratio with another jurisdiction’s capital ratio requires examining valuation rules, risk calibration, diversification recognition and capital eligibility. Numerical ratios alone do not establish equivalent resilience.
Worked example: A Swiss insurer has CHF 240 million of risk-bearing capital and CHF 200 million of target capital: its SST ratio is 120%. Under a stipulated stress, these become CHF 210 million and CHF 220 million. The ratio falls to approximately 95.5%, reflecting changes in both components.
Common mistake: Holding target capital fixed during every stress can conceal deterioration in the SST ratio when adverse conditions simultaneously weaken resources and increase measured risk.
Official syllabus reference: [1]
21. Expected shortfall and discrete tail allocation
Expected shortfall averages losses in the worst specified fraction of the distribution. For a continuous distribution it can be expressed through tail quantiles; for discrete outcomes, the calculation may require including only part of the probability mass at the boundary. Simply averaging losses strictly greater than VaR can select the wrong tail probability. Expected shortfall captures severity beyond a quantile and supports coherent aggregation under standard integrability conditions. Its estimation is nevertheless sensitive to sparse extreme observations and model assumptions. Define the target tail probability explicitly and check that the included probability weights sum to it.
Worked example: Losses of $0, $5 and $20 million have probabilities 70%, 20% and 10%. The worst 20% includes all the $20 million outcome and half the $5 million probability mass, producing expected shortfall of $12.5 million.
Common mistake: Averaging only outcomes strictly above VaR can miscalculate expected shortfall when a discrete probability mass straddles the tail boundary.
Official syllabus reference: [1]
22. Regression hedge ratios and residual variance
A minimum-variance hedge chooses the quantity of a traded instrument that best offsets changes in a target exposure under an estimated joint distribution. If surplus change is hB − L, the unconstrained optimal quantity is Cov(L,B)/Var(B). Residual variance equals Var(L) minus the covariance squared divided by Var(B). The method captures statistical relationships rather than contractual replication. Its usefulness depends on stable estimates and a suitable observation horizon. Transaction costs, liquidity constraints and tail objectives can justify a different position, while changing regimes can weaken an apparently strong historical fit.
Worked example: Liability changes have covariance 12 with unit hedge gains, and hedge variance is four. The minimum-variance position is three units. If liability variance is 40, residual variance remains four, showing that the hedge substantially reduces but does not eliminate variability.
Common mistake: A high historical fit does not establish exact replication or guarantee that the same hedge ratio will work in stressed markets.
Official syllabus reference: [1]
23. Principal components of yield-curve movements
Principal component analysis decomposes the covariance matrix of observed curve changes into orthogonal statistical factors. In many datasets, leading components resemble level, slope and curvature movements, although their shapes and explained variance depend on the sample and representation. Portfolio exposure to a component combines maturity sensitivities with that component's loading vector. Hedging leading components can reduce dimensionality, but omitted components remain sources of risk. PCA maximizes explained historical variance rather than economic relevance or tail protection. Re-estimation and stress comparisons help detect whether the selected factors remain appropriate for the liability profile.
Worked example: A portfolio is neutral to the first component, which resembles a parallel shift, but has substantial exposure to the second component's steepening pattern. A low aggregate duration therefore coexists with meaningful sensitivity to the observed slope factor.
Common mistake: The first few principal components explain historical variance, but they need not capture every economically important stress for a particular liability.
Official syllabus reference: [1]
24. Credit-spread risk versus default loss
Credit-spread risk concerns value changes when required compensation over a reference curve changes. Default risk concerns failure to receive promised payments, with losses depending on exposure and recovery. Spreads reflect more than expected default loss, including risk premiums and liquidity conditions, so spread widening need not imply an immediate default. In ALM, examine whether liability valuation moves with the same spread component and whether that relationship is economically justified. Spread duration measures local price exposure, while default scenarios model discontinuous cash-flow loss. Treating the two as identical can misstate both valuation volatility and ultimate payment capacity.
Worked example: A corporate bond loses value when its spread widens even though all coupons remain current. If held to maturity and fully repaid, that mark-to-market loss differs from default loss; meanwhile, liquidity needs may still force the insurer to realize it.
Common mistake: A spread hedge does not necessarily remove default exposure, and a forecast of low defaults does not eliminate spread-driven valuation losses.
Official syllabus reference: [1]
25. Currency translation and net liability exposure
Foreign-exchange exposure arises when asset and liability values respond differently to exchange-rate changes. Measure the net position in each currency before deciding whether an asset is exposed relative to obligations. A foreign asset can hedge a liability denominated in that same currency, even while both fluctuate in the reporting currency. Cash-flow dates, derivative maturities and local interest rates create additional dimensions beyond current net value. State the exchange-rate quotation explicitly, since reversing the quote reverses the apparent direction. Distinguish currency risk from the underlying asset's credit, equity or interest-rate exposure.
Worked example: A plan reports in dollars but owes €20 million and holds €18 million of matching assets. Its net position is a €2 million liability. Euro appreciation increases the dollar cost of that shortfall even though the existing euro assets also appreciate.
Common mistake: Classifying every foreign asset as an unwanted currency exposure ignores whether it offsets liabilities payable in the same currency.
Official syllabus reference: [1]
26. US life risk-based capital measures
US life risk-based capital is an NAIC framework implemented through applicable state arrangements, combining prescribed risk components and aggregation rules with capital resources. Components address areas including asset risk, insurance risk, interest-rate and market risk, and business risk. Total adjusted capital is compared with specified RBC control-level measures. A reported RBC percentage must identify its denominator because conventions using authorized control level and company action level are not interchangeable. RBC provides a regulatory capital perspective rather than a complete economic balance-sheet assessment. Analyse ALM actions through both their RBC consequences and their effects on actual risks and liquidity.
Worked example: An insurer reports total adjusted capital of $180 million and authorized control level RBC of $60 million. Its ratio on that stated basis is 300%. A second percentage using a different control-level denominator cannot be compared without conversion and context.
Common mistake: An RBC ratio is ambiguous without its denominator, and its percentage cannot be ranked directly against a foreign capital ratio.
Official syllabus reference: [1]
27. Solvency II own funds, SCR and MCR
Solvency II assesses insurers through an economic balance-sheet approach with eligible own funds and distinct capital requirements. The Solvency Capital Requirement captures specified risks using a standard formula or an approved internal-model approach, while the Minimum Capital Requirement is a separate regulatory measure. Own funds are classified by quality, and eligibility restrictions affect how much can cover each requirement. Liability valuation and risk mitigation rules influence measured capital resilience. An SCR coverage ratio divides eligible own funds by the SCR; it does not substitute for MCR assessment, liquidity analysis or investigation of the insurer's underlying exposures.
Worked example: An insurer has €240 million of own funds eligible for SCR coverage and an SCR of €160 million, producing 150% coverage. Evaluating a derivative overlay also requires checking its effect on market risk, counterparty exposure and available cash.
Common mistake: SCR coverage and MCR coverage are distinct measures; passing one assessment does not make their denominators or eligibility rules equivalent.
Official syllabus reference: [1]
28. Canadian LICAT total and core capital perspectives
Canada's Life Insurance Capital Adequacy Test, administered by OSFI for applicable federally regulated life insurers, compares capital resources with a base solvency buffer. The total ratio incorporates available capital, surplus allowance and eligible deposits relative to that buffer. The core ratio focuses on higher-quality resources, using Tier 1 capital and specified portions of the other elements. The framework addresses financial and insurance risks with prescribed aggregation and eligibility rules. Calculate each measure from the applicable definitions rather than importing another jurisdiction's numerator. ALM evaluation should examine changes in both required resources and their quality, alongside economic and liquidity effects.
Worked example: A hypothetical LICAT assessment shows that a proposed transaction increases total capital resources but contributes little Tier 1 capital. Management examines total and core measures separately, because an improvement in the broader resource measure may not resolve a higher-quality capital constraint.
Common mistake: A rise in total available resources does not guarantee an equivalent improvement in a measure focused on higher-quality capital.
Official syllabus reference: [1]
29. Inflation accumulation over long liability horizons
Inflation-linked liabilities accumulate changes in the relevant price index, often over decades. A payment with uncapped annual indexation grows multiplicatively, so its uncertainty depends on the entire inflation path rather than a single year's rate. Caps, floors, lags and index definitions can create option and basis exposures. Beyond liquid market horizons, extrapolation adds model uncertainty and limits direct hedging. Analyse nominal and real components separately, then test how alternative inflation paths affect payments and funding. A fixed nominal return assumption is insufficient when both benefit amounts and discount rates respond to economic conditions.
Worked example: A $10,000 annual benefit indexed at a scenario rate of 3% becomes approximately $18,061 after twenty increases. If the contract instead caps each year's increase, a path with alternating high and low inflation can produce a different result from the same average inflation.
Common mistake: Applying an average inflation rate can conceal path effects when annual caps, floors or indexation lags alter contractual benefit increases.
Official syllabus reference: [1]
30. Longevity trend risk and survival extrapolation
Longevity risk is the possibility that beneficiaries survive longer than assumed, increasing payments for lifetime obligations. Distinguish random fluctuations within a population from systematic changes in mortality trends that affect many lives together. A large portfolio diversifies much of the former but not the latter. Long-horizon projections require assumptions about age effects, future improvement and possible cohort patterns. Small annual differences can accumulate into substantial liability changes. Evaluate survival cash flows under alternative trend structures and parameter assumptions, recognizing that asset maturities and liquid longevity hedges may not extend across the full liability horizon.
Worked example: A pension plan doubles its membership while keeping similar demographics. Individual death fluctuations become more predictable proportionally, but an unexpectedly persistent mortality improvement still raises payments across the enlarged population. Portfolio size alone does not remove that systematic trend exposure.
Common mistake: Diversifying across more pensioners reduces individual sampling noise but does not eliminate uncertainty about future mortality improvement affecting everyone.
Official syllabus reference: [1]
Tools and Strategies to Manage ALM Risks
31. LDI hedge portfolios and return-seeking assets
Liability-driven investment starts with the risks of the obligations and constructs a portfolio that manages funding relative to them. A hedge portfolio targets exposures such as rates and inflation, while return-seeking assets support funding objectives through compensated risk. Derivative overlays can add liability sensitivity without allocating all assets to long bonds, but require collateral and governance capacity. The appropriate hedge ratio reflects the plan's resources and risk appetite rather than a universal target. Evaluate growth and hedging together because losses on growth assets can coincide with collateral demands or a worsening funding position.
Worked example: A plan holds a bond allocation for liability sensitivity and equities for expected growth, then adds a receive-fixed overlay. The overlay reduces rate mismatch while preserving equity exposure, but the plan sets aside liquid assets to support possible margin calls.
Common mistake: An LDI overlay changes risk exposure; it does not generate the assets needed to eliminate an existing funding deficit.
Official syllabus reference: [1]
32. Non-fixed-income assets for long-tail obligations
Long liabilities can extend beyond available fixed-income maturities, encouraging consideration of infrastructure, property or other non-fixed-income assets. Such assets may provide long-lived income or partial inflation sensitivity, but neither feature guarantees a contractual liability match. Cash flows can depend on economic activity, operating costs, financing and regulation. Illiquidity and uncertain valuations create additional constraints. Assess the economic source of returns, scenario-dependent cash flows and realizable sale proceeds rather than classifying assets by label alone. The strategic role may be return generation or partial risk offset, with residual mismatch explicitly retained and measured.
Worked example: An infrastructure investment earns revenue linked partly to a price index, but maintenance costs and user volumes also vary. It may support an inflation-sensitive pension portfolio while still exposing the plan to operating and demand risks absent from the pension promise.
Common mistake: Calling an asset long-lived or inflation-sensitive does not make its uncertain net cash flows equivalent to guaranteed indexed benefits.
Official syllabus reference: [1]
33. Multistage stochastic programming for ALM
Multistage stochastic programming chooses current and future investment decisions across a scenario tree while respecting cash-flow and balance-sheet constraints. The objective can combine expected value with penalties for shortfall, capital use or risk. Decisions at later nodes adapt to information then available, allowing the model to represent rebalancing and benefit payments. Results depend on scenario design, objective scaling and feasible trading assumptions. A mathematically optimal solution may be economically fragile if transaction costs or liquidity are understated. Examine binding constraints and alternative objectives to understand why the model chooses its allocation.
Worked example: A pension model selects today's allocation and next year's rebalancing after either strong or weak markets. It penalizes contribution shortfalls and requires benefit cash at every node, so the optimal strategy accounts for future funding pressure as well as expected returns.
Common mistake: An optimizer's allocation is conditional on its scenarios and constraints; mathematical optimality does not establish robustness to omitted economic conditions.
Official syllabus reference: [1]
34. Nonanticipativity and implementable decisions
Nonanticipativity requires decisions to depend only on information available when they are made. In a scenario tree, paths sharing the same history must have identical decisions until their information sets diverge. Without this restriction, an optimizer can exploit future knowledge by buying assets before favourable outcomes and avoiding them before losses. The resulting performance is unattainable. Apply nonanticipativity to trading, management actions and any adaptive liability decisions. When assessing a simulation strategy, check that decision rules use current state variables and permitted forecasts rather than realized future outcomes hidden in the implementation.
Worked example: Two scenarios are identical through year one but differ in year-two equity returns. Their initial equity allocation must be the same. Allowing a high allocation only on the path that later rises would give the model advance knowledge.
Common mistake: Different future scenarios do not justify different current decisions when the institution cannot yet distinguish which scenario will occur.
Official syllabus reference: [1]
35. Cash-flow dedication and reinvestment avoidance
Cash-flow dedication selects assets whose promised receipts cover specified liability amounts on their payment dates. Exact matching can avoid reinvestment risk for those payments because funds arrive when needed. Coupons received earlier than obligations require assumptions about accumulation unless they are otherwise allocated. Credit events, inflation-linked benefits, options and changing expenses can break the match. Compare contractual cash flows rather than relying solely on present value or duration. Dedication is most straightforward for known nominal payments, while uncertain insurance and pension cash flows usually require margins, scenarios or complementary risk-management techniques.
Worked example: A benefit fund buys zero-coupon bonds paying $3 million in year two and $5 million in year four to meet identical obligations. No intervening reinvestment is required for those payments, provided the promised proceeds arrive in full and on time.
Common mistake: Matching present values is not cash-flow dedication when asset receipts arrive on different dates from the payments they must fund.
Official syllabus reference: [1]
36. Local immunization and its conditions
Classical local immunization controls the effect of small parallel rate changes by matching present value and dollar duration, with suitable convexity conditions for the objective being protected. It is conditional on the assumed cash flows and yield-curve movement. Passage of time and market changes alter durations, so maintaining the position can require rebalancing. A surplus hedge and protection of a future accumulated amount involve related but distinct formulations. State the target explicitly, then assess the relevant first- and second-order terms. Nonparallel changes and embedded options require richer modelling than a single duration match.
Worked example: A fund matches the present value and dollar duration of a ten-year obligation using shorter and longer bonds. Small parallel shocks have limited first-order effect, but the fund still checks convexity and later rebalances as the bonds and obligation age.
Common mistake: A duration match established today is a local condition, not permanent protection against every rate path or future cash-flow change.
Official syllabus reference: [1]
37. Key-rate hedge systems and cross-node effects
A key-rate hedge represents each instrument by its vector of sensitivities and solves for positions that offset the target exposure vector. In matrix form, instrument sensitivities multiplied by hedge quantities equal the required sensitivity adjustment. Instruments often affect several nodes, so sizing one maturity independently can disturb another. Exact solutions require adequate independent instruments; otherwise optimization can minimize weighted residual exposure subject to costs or limits. Check units and signs before solving, then revalue the combined position under curve scenarios. Matching local sensitivities does not establish an exact hedge for nonlinear liabilities.
Worked example: A five-year swap also contributes some ten-year sensitivity, while a ten-year swap affects both nodes. Solving two simultaneous equations accounts for these cross-effects; dividing each target gap by only its main swap sensitivity would leave unintended residual exposures.
Common mistake: A hedge instrument named for one maturity can affect other curve nodes, so independent single-node sizing may fail.
Official syllabus reference: [1]
38. Convexity hedging and nonlinear instruments
A duration-neutral position can retain substantial curvature exposure, particularly when liabilities contain guarantees or surrender options. Convexity hedging seeks instruments whose second-order response offsets that mismatch. Options and option-embedded bonds can change curvature more directly than linear swaps, but their sensitivities vary with rates, volatility and time. The cost of acquiring favourable convexity must be weighed against the protection. Analyse delta and curvature jointly, then test larger scenarios with full revaluation. A local convexity match does not remove volatility exposure or higher-order changes in exercise behaviour.
Worked example: An insurer matches liability DV01 using swaps but remains exposed to falling rates that make a guarantee increasingly valuable. Purchasing an appropriate receiver swaption can add nonlinear protection, although its premium and changing sensitivity must be included in the assessment.
Common mistake: Adding more linear duration exposure does not necessarily address the changing sensitivity created by an embedded option's curvature.
Official syllabus reference: [1]
39. Gaussian copulas and marginal transformations
A copula separates marginal distributions from their dependence structure. A Gaussian copula generates correlated normal latent variables, transforms them to uniform ranks and then applies the chosen inverse marginal distributions. Consequently, nonnormal individual losses can still be joined by Gaussian dependence. For nondegenerate correlations, the Gaussian copula has zero asymptotic tail dependence, which does not mean zero finite-threshold joint loss probability. Calibration should consider the relevant joint outcomes rather than correlation alone. In ALM, use the construction to distinguish individual asset or liability risk assumptions from assumptions about simultaneous adverse experience.
Worked example: Equity losses follow a skewed distribution and credit losses have discrete jumps. A Gaussian copula can connect their ranks without changing either marginal. The resulting probability of both exceeding severe thresholds depends on the chosen dependence parameter and marginal cutoffs.
Common mistake: Gaussian copula dependence does not require Gaussian marginal losses, and zero asymptotic tail dependence does not mean extreme joint events cannot occur.
Official syllabus reference: [1]
40. Student-t copulas and shared tail shocks
A Student-t copula introduces dependence through correlated latent variables with a shared random scale. That scale can create clustered extremes even when the marginal distributions are held fixed. Its degrees-of-freedom parameter controls an important aspect of tail behaviour; lower values generally imply stronger tail dependence at a fixed correlation parameter. Matching a rank correlation with a Gaussian copula does not equate their joint-tail probabilities. Assess capital-relevant exceedances and sensitivity to calibration uncertainty. Heavier modelled tail dependence can be prudent in a scenario comparison without proving that the specification accurately describes the institution's risks.
Worked example: Two portfolio models use identical equity and credit marginals and matching Kendall correlation. The Student-t version produces more simultaneous severe losses because its shared scale generates joint extremes, reducing the apparent diversification benefit compared with the Gaussian version.
Common mistake: Equal marginal distributions and rank correlation do not establish equal joint-tail behaviour or equal diversification benefits for capital purposes.
Official syllabus reference: [1]
41. Coherent multivariable stress narratives
A multivariable stress combines market, insurance and operational changes through an economically interpretable narrative. Coherence means that the assumptions can plausibly occur together, including changes in correlations, policyholder behaviour and liquidity. This differs from mechanically combining every individual worst case. Revalue both assets and liabilities, then incorporate contractual triggers and feasible management responses in sequence. Separate the gross shock from the effects of actions to reveal where resilience comes from. Stress testing explores vulnerability; assigning likelihood requires additional evidence and should not be implied merely because the scenario has a detailed narrative.
Worked example: A credit deterioration scenario combines spread widening, reduced bond liquidity and downgrade-triggered withdrawals. The insurer first revalues holdings, then estimates cash demands and stressed sales, capturing a connected mechanism rather than treating the three effects as independent unrelated losses.
Common mistake: Combining every standalone maximum loss can create an incoherent scenario and obscure the actual mechanisms that threaten the balance sheet.
Official syllabus reference: [1]
42. Reverse stress testing and failure boundaries
Reverse stress testing starts with a defined adverse outcome and works backward to identify conditions that produce it. Outcomes can include a capital floor breach, inability to meet payments or failure of a business strategy. The method highlights combinations that ordinary forward scenarios may overlook. Define the boundary precisely and distinguish reaching it from breaching it. Multiple constraints can bind at different points, so solve capital and liquidity boundaries separately. The resulting scenario is a vulnerability statement, not an estimated event probability, unless a separate calibrated distribution supports that inference.
Worked example: A fund's surplus remains above its stated floor until spreads widen 180 basis points, but accessible cash is exhausted after 110 basis points because collateral requirements increase. The earlier liquidity boundary determines the first operational vulnerability in that stress family.
Common mistake: Finding a shock that causes failure does not reveal how likely that shock is or establish that it is the only failure path.
Official syllabus reference: [1]
43. Scenario weights, reduction and decision stability
Scenario-based ALM uses outcomes and weights to approximate an uncertain economic environment. Weights determine expected values and probability-based risk measures, while unweighted stresses answer different questions. Scenario reduction can improve computational efficiency, but preserving averages alone may remove paths that drive funding shortfalls or collateral pressure. Evaluate whether the reduced set preserves relevant distributions, dependence and decision results. A stable asset allocation across reasonable scenario sets provides stronger evidence than a single optimized answer. Extreme paths can be retained deliberately for constraints even when they contribute little to average outcomes.
Worked example: A model reduction keeps expected returns nearly unchanged but deletes paths with simultaneous equity losses and collateral calls. The optimized portfolio then holds less cash, revealing that matching means did not preserve the liquidity features governing the decision.
Common mistake: A reduced scenario set can preserve average returns while losing the tail paths that determine whether the proposed strategy remains feasible.
Official syllabus reference: [1]
44. Parameter uncertainty and sensitivity surfaces
Parameter risk arises because model coefficients, trends and dependence inputs are estimated imperfectly. Sensitivity analysis changes these inputs to reveal their effect on outputs and decisions. Varying one parameter at a time can identify local drivers but miss interactions, so joint ranges or structured experiments may be needed. Statistical uncertainty and judgmental uncertainty should be distinguished rather than assigned artificial precision. Focus on decision stability: a small parameter change that reverses an allocation recommendation is important even if the reported liability movement seems modest. Parameter testing does not validate the model's structural assumptions.
Worked example: An allocation model prefers illiquid credit when expected excess return is 1.8%, but switches heavily toward bonds at 1.5%. The narrow change reveals that the recommendation depends strongly on an uncertain return estimate rather than a robust liability advantage.
Common mistake: Precisely estimating parameters within a chosen model does not resolve uncertainty about whether that model's structure is appropriate.
Official syllabus reference: [1]
45. Model validation through benchmarks and use limits
Model validation examines conceptual soundness, implementation and fitness for a stated use. Benchmark comparisons can reveal coding errors or dependence on modelling choices, while backtesting compares past predictions with observed outcomes where data permit. Neither a good historical fit nor agreement with another model proves reliability outside the tested range. Document limitations in terms of affected outputs and decisions, such as guarantee valuation in very low rates. Validation findings should lead to correction, additional margins, alternative analyses or explicit restrictions. A model can be acceptable for one application and unsuitable for another.
Worked example: A proxy liability model matches average values but understates a guarantee near its exercise boundary. Validation permits routine central projections while requiring full valuation for stress capital and hedge sizing, where the proxy's local error changes the decision.
Common mistake: Passing implementation checks does not establish that the model is conceptually suitable for every valuation, projection or stress application.
Official syllabus reference: [1]
46. Dynamic policyholder behaviour and management rules
Dynamic ALM models let policyholder actions and management decisions respond to evolving conditions. Credited-rate rules, bonus declarations, surrender behaviour and asset sales can materially alter future cash flows. The rules must respect contractual commitments and information available at the decision date. Model interactions explicitly: changing credited rates may reduce withdrawals but compress margins, while asset sales may meet cash needs but realize losses. Avoid assuming management acts instantly or without cost. Test alternative response speeds and behaviours because a strategy that works under perfect adaptation may fail when actions are delayed or less effective.
Worked example: An insurer raises credited rates after market yields increase to reduce surrender incentives. The model captures lower withdrawals alongside higher future crediting costs, instead of treating the management action as a costless reduction in liquidity risk.
Common mistake: A management action should not be modelled only through its benefit while ignoring its costs, delays and effects on other exposures.
Official syllabus reference: [1]
47. Counterparty limits and collateralized hedge design
An over-the-counter hedge can reduce market risk while creating exposure to the counterparty’s failure. Management can combine counterparty limits, enforceable close-out netting and collateral agreements to control this exposure. Netting offsets covered positive and negative replacement values; collateral supports the remaining claim, subject to eligibility, valuation and access conditions. Controls should also address wrong-way risk, where the counterparty weakens when the hedge becomes valuable. Collateralization requires liquidity planning because the insurer may itself need to post collateral. Evaluate hedge protection, residual counterparty exposure and funding demands together when selecting contractual terms.
Worked example: Two swaps with one counterparty have replacement values of positive $9 million and negative $4 million. Enforceable close-out netting leaves $5 million of exposure. Immediately accessible eligible collateral of $3 million reduces simplified unsecured exposure to $2 million, before subsequent market movements or settlement costs.
Common mistake: Treating collateralized hedges as free of counterparty risk ignores exposure between collateral exchanges, uncertain recovery timing and possible deterioration in collateral value.
Official syllabus reference: [1]
48. Protective puts and fair-premium economics
A protective put combines an asset holding with an option to sell the underlying exposure at a specified strike. The terminal put payoff is max(strike minus underlying level, zero), scaled by notional. Protection depends on strike, maturity and matching the underlying exposure. At a fair purchase price, paying premium exchanges cash for an equally valued option asset, leaving immediate economic surplus unchanged. Subsequent economic gain is the option's new value minus its initial value. Terminal strategy performance includes the premium's financing effect, even though the premium was not an immediate fair-value loss.
Worked example: A put costs $2 million and later pays $9 million. Its appreciation is $7 million, while the initial purchase merely substituted an option asset for cash. Terminal portfolio analysis includes the $2 million expenditure without recording it twice.
Common mistake: Counting the premium as an immediate economic loss and later counting the entire payoff as appreciation misstates fair-value changes.
Official syllabus reference: [1]
49. Collars and the cost of surrendered upside
A collar combines a protective put with a written call on the same underlying exposure. The call premium can fund some or all of the put premium, but the written call limits participation above its strike. A zero-net-premium collar therefore has an economic cost through surrendered upside and possible collateral requirements. Match notionals and maturities carefully, then evaluate the combined payoff rather than each option separately. For ALM, the structure can reduce downside risk while affecting future funding recovery in strong markets. Its suitability depends on both the liability profile and the institution's return needs.
Worked example: A pension fund buys protection below an equity index level of 90 and sells upside above 115, starting from 100. The premiums offset initially, but a terminal index level of 130 leaves the fund's covered upside capped at the call strike.
Common mistake: Zero net premium does not mean free protection; the institution pays economically by giving up specified upside and accepting contractual obligations.
Official syllabus reference: [1]
50. Currency forwards and maturity alignment
A currency forward fixes an exchange of specified currency amounts at a future date. It can hedge a known foreign-currency payment without converting funds immediately, but its value and collateral requirements change before settlement. The forward rate reflects the applicable interest-rate relationship and market basis rather than simply a forecast of the future spot rate. Align currency, amount and settlement date with the exposure. Rolling short forwards against long liabilities retains future pricing and rollover uncertainty. A currency hedge removes only the specified exchange-rate exposure, leaving the underlying asset or liability risks intact.
Worked example: A dollar-based insurer must pay €10 million in one year and contracts to buy that amount forward on the payment date. The exchange rate is fixed for the payment, but a separate euro bond investment can still suffer credit losses.
Common mistake: Hedging a foreign asset's currency translation does not also hedge its credit risk, equity risk or interest-rate sensitivity.
Official syllabus reference: [1]
51. Forward rates from discount factors
A forward interest rate is derived from discount factors under specified compounding and day-count conventions. For a simple-accrual period from T1 to T2, the forward rate is [P(0,T1)/P(0,T2) − 1] divided by the accrual fraction. The relationship follows from no-arbitrage between investing to T2 directly and investing to T1 then locking the later period. It is a market-implied contract rate, not necessarily an unbiased forecast. In modern applications, distinguish the discount curve from projection curves where relevant. Consistent conventions are essential when valuing forwards, swaps and liability cash flows.
Worked example: One-year and two-year discount factors are 0.96 and 0.91. With a one-year simple accrual period, the implied forward rate is 0.96/0.91 minus one, approximately 5.495%, regardless of management's separate forecast for next year's interest rate.
Common mistake: A market-implied forward rate follows pricing relationships and should not automatically be interpreted as the most likely future spot rate.
Official syllabus reference: [1]
52. Receive-fixed swaps and liability duration
An interest-rate swap exchanges fixed and floating interest cash flows on a notional amount. Receiving fixed generally adds positive duration relative to receiving floating, so the position commonly gains value when the relevant curve falls. It can increase asset-side sensitivity to match long liabilities without purchasing the full notional amount of bonds. Notional is a scaling amount rather than initial market value or economic exposure. Valuation depends on projection and discounting conventions, while collateral can create liquidity demands. Match the swap's sensitivity profile to the liability rather than selecting maturity solely from its label.
Worked example: A pension plan's liabilities gain $8 million in a falling-rate scenario while existing assets gain $5 million. A receive-fixed overlay designed to gain approximately $3 million can reduce the funding mismatch, although opposite rate moves may require collateral.
Common mistake: Swap notional is not the amount initially invested and does not directly state the position's market value or DV01.
Official syllabus reference: [1]
53. Receiver and payer swaptions
A swaption grants the right to enter a swap on specified terms. A receiver swaption permits receiving fixed and becomes more attractive when the relevant swap rate falls below the strike; a payer swaption benefits from higher rates relative to its strike. These instruments can hedge asymmetric liability exposures while preserving the choice not to enter an unfavourable swap. Premium reflects volatility, time and other contract features. Exercise style and physical or cash settlement matter. Select direction by the liability's adverse scenario, then evaluate whether expiry and underlying swap tenor align with the period of exposure.
Worked example: An insurer's annuity guarantee becomes more costly when rates fall before retirement elections. A receiver swaption expiring near the election date can gain in that scenario, with its underlying swap tenor chosen to reflect the resulting annuity payment horizon.
Common mistake: A receiver swaption protects against falling relevant swap rates; confusing receiver and payer direction can reinforce the original liability exposure.
Official syllabus reference: [1]
54. Bond futures and conversion-factor DV01
Bond futures transfer exposure to eligible deliverable bonds under contract-specific delivery rules. A common local hedge approximation divides the cheapest-to-deliver bond's DV01 by its conversion factor to estimate futures DV01. The estimate can change when the deliverable choice or bond sensitivity changes. Long futures add rate sensitivity; short futures reduce it under the usual positive-DV01 convention. Daily settlement creates cash flows before final delivery. Use contract-specific values, not generic face amounts, for sizing, and reassess the hedge when market movements alter the delivery economics or the target liability sensitivity.
Worked example: A deliverable bond has DV01 of $72 per basis point and conversion factor 0.8, giving approximate futures DV01 of $90. Offsetting excess asset DV01 of $45,000 per basis point requires selling about 500 contracts under that approximation.
Common mistake: Dividing the target DV01 by a futures contract's face amount ignores the sensitivity and delivery mechanics needed for hedge sizing.
Official syllabus reference: [1]
55. Basis risk between hedge and liability drivers
Basis risk arises when a hedge and the exposure respond to related but different drivers. Examples include government versus swap curves, different inflation indices or asset spreads that do not track liability discount spreads. A hedge can neutralize the common factor while leaving the difference exposed. Identify the economic source of the basis and measure its behaviour in ordinary and stressed conditions. Historical correlation may weaken when liquidity or credit conditions change. A precise hedge ratio for the wrong driver can therefore leave substantial residual risk despite apparently matched aggregate sensitivities.
Worked example: A liability is indexed to one consumer-price measure, while the available inflation swap references another. The swap offsets broad inflation changes, but differences in index composition can leave payments and hedge receipts moving apart over time.
Common mistake: A high correlation between the hedge index and liability index does not guarantee a stable relationship during the relevant stress.
Official syllabus reference: [1]
56. Hedge rebalancing thresholds and transaction costs
Hedge exposures change as markets move, time passes and liability cash flows evolve. Rebalancing restores the target but incurs transaction costs, market impact and potentially collateral movements. A threshold policy trades when exposure leaves an approved range rather than continuously restoring an exact match. The range balances residual risk against implementation cost and capacity. Evaluate it through scenario-based trading paths, including stressed liquidity and discrete price jumps. A policy needs observable triggers and executable trade rules. Backtests should include the timing and costs of decisions rather than assuming trades occur at idealized prices.
Worked example: A plan permits its liability hedge ratio to vary within an approved band. Small market moves generate no trades; a larger shift triggers adjustment. Simulation compares the resulting funding variability and trading costs with those of daily exact rebalancing.
Common mistake: Frequent rebalancing is not automatically better when trading costs, market impact and available liquidity outweigh the reduction in small mismatches.
Official syllabus reference: [1]
57. Reinsurance structures and retained exposure
Reinsurance transfers specified insurance cash-flow risks under contractual terms while creating a claim on the reinsurer. Proportional arrangements share defined premiums and claims, whereas excess-of-loss structures respond above an attachment point subject to limits. Analyse exclusions, timing, reinstatements and experience features because they determine what risk remains. Economic risk transfer, accounting recognition and regulatory capital credit are related but distinct assessments. Reinsurance can reduce adverse claim exposure without supplying immediate cash at the original payment date. Evaluate recoverability and counterparty concentration alongside the gross-to-net reduction in modelled losses.
Worked example: A contract reimburses annual covered claims above $20 million up to a $15 million limit. If covered claims total $42 million, the recovery is $15 million, leaving $27 million with the cedant before considering settlement timing and other contract terms.
Common mistake: A large gross-to-net reduction in modelled claims does not eliminate counterparty risk, exclusions or the need to fund payments before recovery.
Official syllabus reference: [1]
58. Insurance-linked securities and trigger basis
Insurance-linked securities transfer specified insurance-event exposure to capital-market investors, often through funded structures. The payout trigger may reference the sponsor's actual losses, an industry measure or a parameter describing the event. Indemnity triggers can align more closely with sponsor losses but require loss verification; nonindemnity triggers may settle differently and create basis risk. Analyse attachment, exhaustion, maturity and collateral arrangements. Funding can reduce some counterparty concerns without making every structure risk-free. For ALM, compare the timing and amount of protection with the actual obligation rather than relying on the instrument's headline coverage.
Worked example: A sponsor suffers $80 million of claims, but a parametric security pays according to measured event intensity. If the intensity threshold is not reached, the sponsor can receive little protection despite substantial claims, illustrating trigger basis risk.
Common mistake: A funded insurance-linked security does not necessarily reimburse actual losses when its contractual trigger measures something different from those losses.
Official syllabus reference: [1]
59. Longevity swaps, buy-ins and buy-outs
A longevity swap typically exchanges expected survival-linked payments for payments reflecting specified longevity experience, isolating a demographic risk component. A pension buy-in purchases an insurance asset paying covered benefits while the plan retains obligations to members. A buy-out generally involves transferring covered benefit obligations to an insurer, subject to the applicable arrangement and legal requirements. These structures differ in upfront funding, retained investment risk, counterparty exposure and benefit coverage. Compare actual contract terms rather than labels. Survival reference populations, indexation and settlement provisions determine how closely the protection follows the plan's payments.
Worked example: A plan can retain its investment portfolio and use a longevity swap, or fund a buy-in that supplies covered pension receipts. The former leaves more investment management with the plan; the latter requires substantial premium funding while member obligations remain with it.
Common mistake: A buy-in is an asset supporting the plan's payments and should not automatically be treated as a discharge of member obligations.
Official syllabus reference: [1]
60. Risk budgets and marginal hedge value
A risk budget allocates acceptable exposure across risks or strategies in relation to institutional objectives. Marginal hedge analysis asks how an additional unit of protection changes surplus risk, capital use, expected return and liquidity consumption. The cheapest premium is not necessarily the best value when it leaves important basis risk or requires costly collateral. Evaluate incremental effects on the whole portfolio because diversification makes standalone risk reductions nonadditive. Budget constraints can justify partial hedging where full protection is expensive or impractical. Document the retained exposure and test whether the strategy remains acceptable when costs or dependence assumptions change.
Worked example: Two hedges reduce modelled surplus loss by the same amount. One has a lower premium but creates large stressed margin calls; the other costs more initially but fits the institution's limited liquidity resources. Marginal analysis values both protection and resource use.
Common mistake: Selecting the lowest quoted hedge cost can be misleading when basis exposure, collateral demands and effects on total portfolio risk differ.
Official syllabus reference: [1]