Actuarial Principles and Insurance Mathematics

Chapter 13
Asset–Liability Management, Solvency and Financial Reporting

Post Graduate In Insurance Management
Actuarial Principles and Insurance Mathematics   •   Post Graduate

Asset–Liability Management, Solvency and Financial Reporting

Introduction

An insurer promises cash flows whose amount, timing and sensitivity to economic conditions are uncertain. It finances those promises with premiums, investment assets and shareholders’ capital. Asset–liability management (ALM) coordinates these elements so that obligations can be met while acceptable profitability, liquidity and solvency are maintained.

ALM cannot be separated from solvency assessment or financial reporting. Economic mismatches affect regulatory capital; management actions affect both future solvency and reported earnings; and accounting measurement may reveal volatility that differs from economic volatility. Effective insurance management therefore requires an integrated view of economic value, cash-flow adequacy, regulatory capital and accounting presentation.

Learning Objectives

After studying this chapter, readers should be able to:

  • explain the purpose and scope of insurance ALM;
  • measure interest-rate, liquidity, currency and other mismatches;
  • apply duration, convexity, cash-flow matching and immunisation techniques;
  • construct and interpret deterministic and stochastic ALM projections;
  • calculate diversified solvency capital and project solvency ratios;
  • distinguish regulatory, economic and accounting balance sheets;
  • reconcile insurance liability movements, profit and equity;
  • evaluate management actions, model limitations and governance requirements.

1. The Integrated Insurance Balance Sheet

For economic analysis, available capital or economic surplus at time t is commonly defined as:

Ct = At - Lt - Ot

where:

  • At = market-consistent value of assets at time t;
  • Lt = economic value of insurance liabilities;
  • Ot = value of other liabilities;
  • Ct = available economic capital or surplus.

All amounts must be measured on a consistent valuation basis and in the same currency. A positive surplus does not by itself establish solvency: the surplus may be highly volatile, illiquid or insufficient relative to required capital.

Three related balance-sheet perspectives should be distinguished.

PerspectiveMain purposeTypical valuation basis
Economic balance sheetRisk and value managementMarket-consistent or best-estimate economic values
Regulatory balance sheetPolicyholder protection and capital complianceRules prescribed by the supervisor
Financial-reporting balance sheetGeneral-purpose reportingApplicable accounting standards, such as IFRS 9 and IFRS 17

Differences may arise from discount rates, risk margins, contract boundaries, asset classification, recognition timing and allowance for future profits.

Integrated asset liability management cycle Assets andliabilities Economic andstress scenarios ALM metrics andoptimisation Solvency andliquidity Financialreporting Managementactions
Integrated asset liability management cycle
Figure: Integrated asset liability management cycle

Figure: ALM is a feedback process linking exposures, scenarios, solvency, reporting and management actions.

2. Principal Asset–Liability Risks

2.1 Interest-rate risk

Asset and liability values respond differently to changes in discount rates. Life and annuity liabilities may have very long duration, while general insurance liabilities may be shorter but sensitive to inflation and claim-settlement patterns.

For cash flows CFt payable at times t, present value under a single annual effective yield y is:

PV = ∑t=1n CFt(1+y)-t

This formula assumes that the same yield is appropriate for every maturity. In practice, each cash flow is discounted using a term-structure rate appropriate to its maturity, currency and valuation basis.

2.2 Liquidity risk

Liquidity risk is the possibility that an insurer cannot meet cash obligations when due without excessive cost. It can arise even when economic capital is positive. Sources include catastrophe claims, collateral calls, policy surrenders and illiquid investments.

Useful measures include projected net cash flows, liquid-asset coverage ratios, collateral requirements and the time required to monetise assets under stressed conditions.

2.3 Other mismatches

  • Currency risk: assets and liabilities are denominated in different currencies.
  • Inflation risk: claims or expenses rise faster than asset income.
  • Credit risk: issuers or reinsurers default or suffer spread widening.
  • Equity and property risk: market values decline when capital is needed.
  • Optionality risk: policyholder behaviour changes with market conditions.
  • Basis risk: the hedge instrument and liability respond to different risk drivers.

3. Duration, Convexity and Immunisation

3.1 Modified duration

Modified duration measures the first-order sensitivity of value to a small yield change:

Dmod = -\frac{1}{P}\frac{∂P}{∂y}

Consequently:

∆P ≈ -DmodP∆y

Here P is present value in currency units, Dmod is measured in years, and ∆y is a decimal yield change. The approximation assumes a small, parallel shift in the relevant yield curve and stable cash flows.

For assets and liabilities, the approximate change in surplus is:

∆C ≈ -(DAA-DLL)∆y

where DA and DL are modified durations and A and L are current values. The quantity DAA-DLL is a currency-duration gap, measured in currency-years.

Worked Example: Correct Duration-Gap Calculation

An insurer has assets of ₹600 million with modified duration 5.5 years and liabilities of ₹570 million with modified duration 4.5 years. Rates increase by 75 basis points, so ∆y=0.0075.

DAA-DLL=(5.5×600)-(4.5×570)=3,300-2,565=735

Thus:

∆C≈-735×0.0075=-₹5.5125\text{ million}

Opening surplus is ₹30 million, so the approximate post-shock surplus is ₹24.4875 million. The calculation is a first-order estimate; it does not capture convexity or non-parallel yield-curve movements.

3.2 Convexity

For larger yield movements, a second-order approximation is preferable:

\frac{∆P}{P}≈-Dmod∆y+\frac{1}{2}K(∆y)2

where K is convexity, measured in years squared under the stated yield convention. Convexity improves the approximation when cash flows are fixed and the price–yield relationship is smooth.

3.3 Immunisation and cash-flow matching

Local immunisation against small parallel rate changes normally requires:

  1. equality of asset and liability present values;
  2. equality of asset and liability durations;
  3. preferably, asset convexity at least as high as liability convexity.

These conditions do not guarantee protection against non-parallel curve movements, defaults, changing policyholder behaviour or cash-flow revisions. Key-rate durations, which measure sensitivity at selected maturities, are more useful where yield-curve shape changes are material.

Cash-flow matching instead constructs asset receipts to meet liability payments by date and amount. It reduces reinvestment risk but may be costly or impossible where liabilities are uncertain.

4. Stochastic ALM

Deterministic stress tests examine selected scenarios. Stochastic ALM models a probability distribution of future outcomes and captures interactions among markets, claims, policyholder behaviour, expenses and management actions.

4.1 Scenario dynamics

A simplified asset recursion is:

At+1=At(1+Rt+1)+Pt+1-Bt+1-Et+1+It+1

where Rt+1 is the portfolio return, Pt+1 premiums received, Bt+1 benefits and claims paid, Et+1 expenses and It+1 external capital introduced. All cash flows refer to the same projection interval.

Liabilities may be projected as:

Lt+1=f(Lt,rt+1,qt+1,st+1,πt+1,ht+1)

where the function reflects interest rates r, demographic or claim variables q, surrender behaviour s, inflation π and management or policyholder actions h.

Real-world scenarios are generally used for business planning, solvency distributions and strategy. Risk-neutral scenarios are used where market-consistent valuation requires discounted expected cash flows consistent with observable market prices. The two probability measures are not interchangeable.

4.2 Risk measures

For a loss variable X, value at risk at confidence level α is:

VaR_α(X)=inf\{x:Pr(X≤x)≥α\}

Expected shortfall is the mean loss in the worst 1-α proportion of outcomes, with an appropriate quantile-integral definition where distributions have atoms. Expected shortfall considers tail severity, whereas VaR identifies a quantile but not the magnitude of losses beyond it.

Worked Stochastic Example

An insurer begins with assets of ₹1,000 million and liabilities of ₹920 million, giving capital of ₹80 million. One-year scenarios are:

ScenarioProbabilityAsset valueLiability valueClosing capital
Benign60%1,050.0938.4111.6
Adverse30%1,000.0975.224.8
Severe10%880.01,058.0−178.0

Expected closing capital is:

E[C1]=0.60(111.6)+0.30(24.8)+0.10(-178)=₹56.6\text{ million}

The estimated insolvency probability is 10%. Defining capital loss as X=80-C1, losses are −₹31.6 million, ₹55.2 million and ₹258 million. Therefore, under the stated discrete quantile definition, VaR90\%=₹55.2 million. The worst 10% consists solely of the severe scenario, so 90% expected shortfall is ₹258 million.

4.3 Monte Carlo estimation

With N simulated scenarios, the estimated probability of breaching required capital is:

\widehat{Pr}(Ct<St)=\frac{1}{N}∑i=1N1(Ct(i)<St(i))

where St is required capital and 1(·) is an indicator. Simulation error, parameter uncertainty, dependency assumptions and insufficient tail scenarios must be assessed. Reverse stress testing complements simulation by identifying scenarios capable of making the business model unviable.

5. Solvency Capital and Its Projection

5.1 Required and available capital

A solvency ratio is commonly expressed as:

Solvency\ ratiot=\frac{Eligible\ own\ fundst}{Required\ capitalt}×100\%

The precise definitions of eligible funds, required capital, risk margin and intervention thresholds depend on the jurisdiction. Under risk-based regimes, required capital may reflect a high-quantile one-year loss in basic own funds.

5.2 Diversification

A correlation-based aggregation formula is:

SCRbasic=\sqrt{∑i=1m∑j=1mρijSCRiSCRj}

Here SCRi is the stand-alone capital for risk module i, and ρij is the prescribed or estimated correlation. The correlation matrix must be valid and positive semidefinite. Linear correlation may inadequately represent extreme dependence.

Worked Example: Diversified Capital

Suppose underwriting, market and credit capital requirements are ₹90 million, ₹70 million and ₹25 million. Correlations are 0.25 between underwriting and market risk, 0.25 between underwriting and credit risk, and 0.50 between market and credit risk.

SCRbasic=\sqrt{902+702+252+2(0.25)(90)(70)+2(0.25)(90)(25)+2(0.50)(70)(25)}

The quantity inside the square root is ₹19,650 million squared, giving:

SCRbasic=₹140.18\text{ million}

If operational-risk capital of ₹12 million is added and no adjustment is assumed, total required capital is ₹152.18 million. Eligible own funds of ₹180 million produce a solvency ratio of approximately 118.3%.

5.3 Multi-period solvency projection

A planning model may project eligible capital as:

Ct+1=Ct+Πt+1-Dt+1+Kt+1

where Π is comprehensive profit or loss relevant to eligible capital, D is dividends or distributions, and K is new eligible capital. Required capital must be projected separately as exposures, volatility and diversification change.

StageEligible capital (₹m)Required capital (₹m)Ratio
Year 0180152.18118.3%
Year 1: profit 18, dividend 5193158122.2%
Year 2 after ₹45m stress loss14817087.1%
After ₹25m capital issue and reinsurance reducing requirement by ₹20m173150115.3%

The example demonstrates that capital adequacy depends on both the numerator and denominator. Reinsurance may lower required capital but introduces cost, counterparty risk and basis risk. Capital projections should include adverse scenarios, dividend constraints and credible management-action timing.

6. Financial Reporting and Reconciliation

6.1 Insurance-contract measurement

Under IFRS 17, the general measurement model combines fulfilment cash flows and a contractual service margin (CSM). Fulfilment cash flows include probability-weighted estimates of future cash flows, discounting for timing and financial risk, and a risk adjustment for non-financial risk.

For a profitable group at initial recognition, the CSM prevents recognition of unearned future profit. For onerous groups, the loss is recognised immediately, subject to the standard’s detailed requirements. Subsequent reporting distinguishes liability for remaining coverage from liability for incurred claims.

Insurance revenue represents services provided during the period; it is not simply premium cash received. Insurance finance income or expense reflects discounting and financial-assumption effects. IFRS 9 classification of supporting assets influences whether economic asset–liability offsets are also visible in profit or loss or other comprehensive income.

6.2 Liability roll-forward

A general reconciliation identity is:

Closing\ liability=Opening\ liability+Increases-Reductions

The component classification must follow the applicable accounting standard. A reconciliation is not merely arithmetic: it explains the sources of movement and connects the balance sheet to profit, other comprehensive income and cash flows.

Worked Financial-Statement Reconciliation

Assume an illustrative insurer reports the following insurance-liability movements, in ₹ million:

MovementAmount
Opening insurance liability800
Premium cash receipts increasing the net liability+160
Insurance revenue reducing remaining coverage−110
Incurred claims and service expenses+95
Claim payments−100
Insurance finance expense+32
Adverse current-service experience+8
Closing insurance liability885

The closing value is verified by 800+160-110+95-100+32+8=885.

Insurance service result is revenue of ₹110 million less service expenses of ₹103 million, comprising ₹95 million plus ₹8 million of adverse current-service experience. It is therefore ₹7 million. If investment income is ₹42 million, insurance finance expense is ₹32 million and non-attributable overhead is ₹15 million, profit is:

Profit=110-103+42-32-15=₹2\text{ million}

Suppose opening assets are ₹930 million. Assets then reconcile as:

Closing\ assets=930+160+42-100-15-5=₹1,012\text{ million}

where ₹5 million is a dividend. With closing insurance liabilities of ₹885 million and other liabilities of ₹20 million, closing equity is:

Equity=1,012-885-20=₹107\text{ million}

This agrees with the equity roll-forward:

Closing\ equity=110+2-5=₹107\text{ million}

The example ignores tax and other comprehensive income. Its purpose is to demonstrate the internal reconciliation among liability movements, profit, cash movements and equity.

7. Illustrative Integrated Case: Aster Life and General Insurance

Background

Aster is a fictional composite insurer with long-duration annuity liabilities and short-tail general insurance obligations. Rapid interest-rate changes have reduced asset values, claims inflation has increased liability estimates, and management plans a dividend despite a modest solvency buffer.

Problem

The insurer’s analysis identifies four concerns:

  1. asset duration exceeds liability duration, making surplus vulnerable to rising rates;
  2. severe scenarios generate collateral and claim-payment pressure;
  3. projected growth raises required capital faster than retained earnings;
  4. accounting profit is volatile because asset and liability finance effects are presented differently.

Analysis

Aster constructs an integrated model containing monthly cash flows, key-rate durations, credit transitions, inflation-linked claims and dynamic surrender behaviour. Real-world Monte Carlo scenarios are used for solvency planning, while market-consistent scenarios support valuation.

Management evaluates the following actions:

  • shorten asset duration using bond sales or interest-rate derivatives;
  • purchase inflation-linked assets for claims-sensitive portfolios;
  • arrange contingent liquidity and collateral facilities;
  • increase catastrophe reinsurance;
  • reduce new-business volume in capital-intensive products;
  • retain earnings by limiting dividends;
  • align asset classification and liability finance presentation where accounting standards permit.

Approach and Outcome

Aster chooses a combination rather than a single hedge. Key-rate duration gaps are reduced, liquid assets are increased, catastrophe reinsurance is purchased and the proposed dividend is halved. A solvency projection shows that diversification and reinsurance improve the capital ratio, but the board also reviews reinsurance default stress and renewal-price risk.

Financial reporting teams reconcile projected economic results to accounting earnings. They explain that lower accounting volatility does not necessarily mean lower economic risk and that genuine risk reduction may not immediately increase reported profit.

Learning Points

  • ALM decisions should be tested simultaneously against value, liquidity, solvency and reporting objectives.
  • Management actions must be operationally feasible and modelled with realistic delays.
  • Hedging one metric can worsen another; for example, illiquid long-duration assets may improve duration matching but weaken stressed liquidity.
  • Accounting presentation should inform decisions without replacing economic analysis.

8. Governance, Limitations and Contemporary Developments

ALM governance normally includes board-approved risk appetite, exposure limits, independent model validation, documented assumptions and escalation procedures. Actuarial, investment, finance, risk and treasury functions must use controlled and reconcilable data.

Major limitations include parameter uncertainty, unstable correlations, sparse catastrophe data, behavioural uncertainty and simplified management-action rules. Models may create false precision when severe events lie outside historical experience. Results should therefore be accompanied by sensitivities, stress tests and qualitative judgement.

Ethically, assumptions should not be manipulated to justify dividends, suppress capital needs or defer loss recognition. Conflicts between sales targets and policyholder security require explicit governance. Model changes and expert judgements should be transparent and auditable.

Technology supports daily cash-flow projection, automated balance-sheet reconciliation and large-scale simulation. Machine learning may improve lapse, claim and liquidity forecasts, but opaque models require explainability, bias testing, drift monitoring and human oversight.

Contemporary ALM increasingly addresses climate-related physical and transition risks, persistent inflation, cyber accumulation, collateral liquidity and interactions between IFRS 17 and IFRS 9. These developments favour integrated models rather than separate actuarial, investment and accounting analyses.

Best-Practice Principles

  1. Use consistent cash flows and valuation dates across assets and liabilities.
  2. Examine key-rate and cash-flow mismatches, not only aggregate duration.
  3. combine deterministic, stochastic and reverse stress tests.
  4. Project both eligible capital and required capital over the business-planning horizon.
  5. Model liquidity and collateral separately from solvency.
  6. Reconcile economic, regulatory and accounting results.
  7. Validate dependencies, tail assumptions and management actions.
  8. Communicate uncertainty rather than presenting model outputs as exact forecasts.

Chapter Summary

Asset–liability management coordinates investment assets, insurance obligations, liquidity and capital. Duration and convexity measure sensitivity to interest rates, while cash-flow matching and key-rate analysis address timing and yield-curve risks. Stochastic ALM extends deterministic analysis by generating distributions of capital, solvency and liquidity outcomes.

Solvency assessment compares eligible own funds with risk-based required capital, allowing for diversification subject to model and regulatory constraints. Multi-period projections are essential because profits, dividends, growth, reinsurance and market shocks affect both available and required capital.

Financial reporting provides another view of performance. Insurance-liability roll-forwards must reconcile with revenue, service expenses, finance effects, cash flows and equity. Integrated governance recognises that economic value, regulatory solvency, liquidity and accounting profit are related but not identical.

Key Terms

Asset–liability management: Coordinated management of assets, liabilities, liquidity and capital.

Available capital: Assets exceeding recognised liabilities on a specified valuation basis.

Basis risk: Risk that a hedge and the hedged exposure respond differently.

Cash-flow matching: Selection of assets whose cash receipts correspond to liability payments.

Contractual service margin: Unearned profit recognised as insurance services are provided under IFRS 17.

Convexity: Second-order measure of value sensitivity to yield movements.

Duration: First-order measure of sensitivity to interest-rate changes.

Economic balance sheet: Balance sheet based on economically consistent asset and liability values.

Eligible own funds: Capital resources permitted to cover regulatory capital requirements.

Expected shortfall: Average loss within the specified worst tail of a distribution.

Fulfilment cash flows: Discounted, risk-adjusted estimates of cash flows required to fulfil insurance contracts.

Immunisation: Construction of a portfolio designed to protect surplus against specified interest-rate changes.

Key-rate duration: Sensitivity to yield changes at a selected maturity point.

Liquidity risk: Risk of being unable to meet cash obligations when due without excessive cost.

Risk adjustment: Compensation required for bearing non-financial uncertainty in insurance cash flows.

Solvency capital requirement: Risk-based capital required under an applicable regulatory framework.

Solvency ratio: Eligible own funds divided by required capital.

Stochastic ALM: ALM using probabilistic scenarios for assets, liabilities and management actions.

Value at risk: Loss quantile at a stated probability level and horizon.

Review Questions

Short Answer Questions

  1. What is the primary objective of asset–liability management?
  2. Distinguish liquidity risk from insolvency risk.
  3. What assumptions underlie a modified-duration approximation?
  4. Why are real-world and risk-neutral scenarios not interchangeable?
  5. What is the contractual service margin?

Long Answer Questions

  1. Explain how duration, convexity and key-rate duration contribute to interest-rate risk management.
  2. Discuss the design, uses and limitations of stochastic ALM models.
  3. Explain how diversification is reflected in a risk-based solvency-capital calculation.
  4. Compare economic, regulatory and financial-reporting balance sheets.
  5. Discuss governance and ethical issues in setting ALM and solvency assumptions.

Analytical / Application Questions

  1. Design an ALM framework for an annuity insurer exposed to falling interest rates, longevity improvement and policyholder options.
  2. Evaluate the effects of reinsurance, dividend restriction and a capital issue on a stressed solvency projection.
  3. Construct a reconciliation linking insurance-liability movements, reported profit, asset cash flows and closing equity.

Multiple Choice Questions with Answers

Question 1

Which statement best describes the central purpose of insurance asset–liability management?

A. Maximising investment return without considering liabilities

B. Coordinating assets, liabilities, liquidity and capital within risk appetite

C. Eliminating every source of accounting volatility

D. Holding only risk-free short-term assets

Correct Answer: B. Coordinating assets, liabilities, liquidity and capital within risk appetite

Question 2

A liability pays ₹110 million in one year. Using an annual effective discount rate of 10%, what is its present value?

A. ₹90 million

B. ₹99 million

C. ₹100 million

D. ₹121 million

Correct Answer: C. ₹100 million

Question 3

Assets are ₹600 million with modified duration 5.5 years, while liabilities are ₹570 million with modified duration 4.5 years. Under a parallel yield increase of 0.0075, what is the approximate change in surplus?

A. A decrease of ₹5.5125 million

B. An increase of ₹5.5125 million

C. A decrease of ₹0.735 million

D. An increase of ₹30 million

Correct Answer: A. A decrease of ₹5.5125 million

Question 4

In the chapter's three-scenario stochastic example, what is the estimated probability of negative closing capital?

A. 5%

B. 10%

C. 30%

D. 40%

Correct Answer: B. 10%

Question 5

In that example, capital loss is defined as X=80-C1. What is VaR90\% under the stated discrete quantile definition?

A. −₹31.6 million

B. ₹24.8 million

C. ₹55.2 million

D. ₹258 million

Correct Answer: C. ₹55.2 million

Question 6

Under the stated correlations, what is the diversified basic capital requirement for stand-alone requirements of ₹90 million, ₹70 million and ₹25 million?

A. ₹128.00 million

B. ₹140.18 million

C. ₹152.18 million

D. ₹185.00 million

Correct Answer: B. ₹140.18 million

Question 7

If eligible own funds are ₹180 million and required capital is ₹152.18 million, what is the approximate solvency ratio?

A. 84.5%

B. 100.0%

C. 118.3%

D. 152.2%

Correct Answer: C. 118.3%

Question 8

Using the liability reconciliation 800+160-110+95-100+32+8, what is the closing insurance liability?

A. ₹853 million

B. ₹877 million

C. ₹885 million

D. ₹917 million

Correct Answer: C. ₹885 million

Question 9

If closing assets are ₹1,012 million, insurance liabilities are ₹885 million and other liabilities are ₹20 million, what is closing equity?

A. ₹87 million

B. ₹107 million

C. ₹127 million

D. ₹147 million

Correct Answer: B. ₹107 million

Question 10

Which pairing of scenario measure and application is most appropriate?

A. Risk-neutral scenarios for business-volume forecasting only

B. Real-world scenarios exclusively for market-consistent derivative pricing

C. Real-world scenarios for solvency planning and risk-neutral scenarios for market-consistent valuation

D. Risk-neutral and real-world scenarios are always interchangeable

Correct Answer: C. Real-world scenarios for solvency planning and risk-neutral scenarios for market-consistent valuation
Actuarial Principles and Insurance Mathematics
Chapter 13 of 13