Forecast-based valuation

The Group has significant unrecognized deferred tax assets in respect of tax losses. There is inherent uncertainty involved in forecasting future taxable profits, which determines the extent to which deferred tax assets are or are not recognized. The recoverability of the recognized deferred tax asset is dependent on the future profitability of the UK business in particular.

Our procedures included:

• Control design and observation: Evaluating the internal review and challenge process the Group undertakes during the development of the operational plan and by the Group Risk Function during the approval process, and the extent to which the Group considered alternative outcomes.

• Historical comparisons and our sector experience: Comparing the expected growth rates to the Group’s approved operating plan and assessing the accuracy of that forecasting process in the past. We considered whether projected margins are achievable with reference to the business’s recent performance and operating plans, as well as our own industry knowledge.

• Sensitivity testing: Sensitivity testing of taxable profits to assumptions such as the Combined Operating Ratio.

• Own tax expertise: With the support of our own tax specialists and their knowledge of tax legislation, we also assessed the extent to which projected profits were taxable, in particular the Group’s assumptions about how accumulated tax losses and other associated tax attributes can be utilized within the Group against UK taxable profits.

• Assessing transparency: Assessing adequacy of the Group’s disclosures in respect of the assumptions supporting the deferred tax asset valuation and recognition.

Processing errors

Many financial reporting controls depend on the correct functioning of operational and financial IT systems. If these systems or controls fail, a significant risk of error in reported financial information can arise from the failure to transfer data appropriately between systems or inappropriate changes being made to financial data or systems.

This is an area of significant risk in our audit due to the complexity of the IT infrastructure, particularly in the UK and Scandinavia, where there are legacy systems which require increased manual inputs, relative to more automated processes. This risk is heightened in the current year given the Group is undertaking IT transformation activities, including system implementations and transitions, leading to changes to processes and increasing the risk that controls do not continue to operate effectively.

• Controls design and observation: Testing general IT controls around system access, change management and computer operations within specific applications pertinent to the financial statements by assessing if appropriate policies are in place and adhered to by inspecting supporting evidence. Where general IT controls are not operating effectively, we addressed the increased risk of financial statement misstatement by extending the scope of our work. This included assessing the operation of controls over changes or transactions being recorded in the systems and testing manual compensating controls, such as reconciliations between systems and other information sources, through re-performance or inspection.

• Extended scope: Where general IT controls and compensating manual controls did not operate to mitigate a risk, we performed additional substantive testing, such as using extended sample sizes and performing data analysis routines over impacted accounts.

Subjective valuation

Insurance liabilities represent the single largest liability for the Group. Valuation of these liabilities is highly judgmental because it requires a number of assumptions to be made with high estimation uncertainty such as loss ratios, estimates of the frequency and severity of claims and, where appropriate, the discount rates for longer tail classes of business by territory and line of business. The determination and application of the methodology and performance of the calculations are also complex.

A margin is added to the actuarial best estimate of insurance liabilities to make allowance for specific risks identified in assessment of the best estimate. The appropriate margin to recognize is a subjective judgement and estimate taken by the directors, based on the perceived uncertainty and potential for volatility in the underlying claims.

Certain lines of business have greater inherent uncertainty, for example those where claims emerge more slowly over time, or where there is greater variability in claim settlement amounts. This includes Abuse, Asbestos and Deafness classes; UK Professional and Financial Risk Classes; the Danish Workers Compensation class; the Swedish Personal Accident classes; the Canadian General Liability class; and classes of business affected by emerging industry issues such as the impact of Periodic Payment Orders (which are akin to annuities with longevity and inflation risk) on UK motor business.

Completeness and accuracy of data

The valuation of insurance liabilities depends on complete and accurate data about the volume, amount and pattern of current and historical claims since they are often used to form expectations about future claims. If the data used in calculating insurance liabilities, or for forming judgements over key assumptions, is not complete and accurate then material impacts on the valuation of insurance liabilities may arise.

With the assistance of our actuarial specialists, we performed the following procedures:

• Sector experience and Benchmarking: Comparison of assumptions, reserving methodologies and estimates of losses to expectations based on the Group’s historical experience, current trends and benchmarking to our own industry knowledge including information relating to forthcoming legislation that may affect claims settlement speed or amount.

• Sensitivity analysis: Review of sensitivity analysis over key judgements and assumptions, such as the discount rates for longer tail classes of business and Periodic Payment Order projections.

• Independent re-performance: Independent re-projection of the reserve balances for certain classes of business. The determination of which classes to re-project was based on risk assessment and consideration of the evidence available from other alternative data analysis procedures.

• Control design and observation: Evaluation of the governance around the overall Group reserving process, including the scrutiny applied by the Local, Regional and Group Reserving committees, as well as Group level actuarial reviews. We assessed qualifications and experience of those responsible and examined the output of the reviews to assess the scope and depth of these processes. Our evaluation of the methodologies and key assumptions for the most significant and subjective classes of business enabled us to assess the quality of challenge applied through the Group’s reserving process.

• Margin evaluation: Evaluation of the appropriateness of the Reserve Committee’s recommended margin to be applied to the actuarial best estimate. We reviewed the Director’s approach to, and analysis performed, in setting the margin. In particular we considered the allowance for uncertainties inherent in the data and assumptions in developing the actuarial best estimate through inquiry with the Directors.

In addition to the above the audit team performed procedures to assess the completeness and accuracy of data:

• Control observation and re-performance: Evaluation and testing of key controls designed to ensure the integrity of the data used in the actuarial reserving process (including both current and prior year case reserve data). The controls included reconciliations between data in the actuarial reserving systems and data in the policy administration systems. We tested controls through inspecting or re-performing the Group’s reconciliations.

In addition, we tested the operation of controls over the valuation of individual claims reserves, such as large loss review controls, internal peer reviews and claims approvals. We assessed whether such controls were performed in line with the Group’s policies and with reference to underlying claims data.

• Tests of detail: Re-performance of reconciliations between the claims data recorded in the policy administration systems and the data used in the actuarial reserving calculations to ensure the integrity of the data used in the actuarial reserving process. We also compared samples of claims case reserves to appropriate documentation, such as reports from loss adjusters in order to test the controls over the valuation of individual claims reserves. We used data analysis to focus our procedures on significant or anomalous claims.

Refer to page 84 (Corporate Governance Statement), page 114 (Group accounting policies), page 131 (critical accounting estimates and judgements) and page 156 (note 22 to the consolidated financial statements).

The group has deferred tax assets of € 373 million that primarily arise due to historical operating losses. A key judgement is whether there is convincing evidence of sufficient future taxable profits against which those losses can be utilized.

This judgement relies on the assessment of the probability, the timing and the sufficiency of future taxable profits, which in turn is based on assumptions concerning future economic conditions and business performance.

The Group’s considerations in respect of the recognition of the deferred tax asset are outlined in Note 2, which provides an overview of the key assumptions underpinning the financial projections. We have focused on this area due to the level of judgement involved.

We assessed the forecast of taxable profits which inform management’s decision to recognize a deferred tax asset in respect of tax losses arising from historic operating losses.

We understood and tested key controls over the production and approval of the medium term plan and considered the appropriateness of the growth rate used to extrapolate those profits over the period beyond the detailed plan.

We considered whether the projections provide convincing evidence that sufficient taxable profits will be available to utilize unused tax losses. We assessed the relevant macroeconomic assumptions and growth assumptions underlying the projections in the context of economic consensus forecasts. We also evaluated the other key assumptions for reasonableness by reference to historic performance, future plans and external data as appropriate.

We have also considered the appropriateness of the disclosures included in the financial statements.

The IT framework of the group incorporates a number of IT systems which have been in place for many years.

We have focused on IT controls and in particular deficiencies in the IT control environment, as deficiencies in access controls over a number of applications on certain systems could have a significant impact on financial reporting controls and systems.

We involved our IT audit specialists to update our understanding of the group’s IT environment and of changes made to it during 2016.

To the extent required for our audit, we assessed and tested the design and operating effectiveness of IT controls over financial reporting systems relating to access security, IT operations and change control management, including mitigating controls where relevant.

Where deficiencies affected specific applications within the scope of our audit we extended our programme of work to test compensating controls and to perform other procedures as considered relevant.

A number of projects are ongoing across DLG aimed at improving the customer experience and enhancing operational efficiency. Certain costs relating to these projects are capitalized on the balance sheet, with the carrying amount of intangible assets at year-end totaling £298m (2015: £314m). We have identified a key risk that costs relating to these projects are incorrectly capitalized or that the impairment charge recognized in the year of £39m (2015: £4m) has not been accurately determined in line with IAS 36. Significant judgements and estimates have been made in calculating the future economic benefit of these assets, including the expected future cash flows and the discount rate used.

In addition, we have identified a risk that the transformation programme, which includes the outsourcing and offshoring of finance and operations, could impact the financial reporting control environment, in particular where processes and controls have changed operation and location. There is a risk that these changes will cause management stretch which could impact on the effectiveness of existing financial reporting internal controls, particularly in relation to adequate segregation of duties.

We have assessed the design and implementation and tested the operating effectiveness of key controls over the capitalization of costs and the impairment of intangible assets. These included senior management approval for costs to be capitalized against projects having analyzed the detailed business plan and assessed feasibility. We have tested the capitalization criteria of key projects by assessing the inputs and noting whether these are applicable expenses in accordance with IAS 38. We have tested the impairment charge by performing the following:

• testing the completeness and accuracy of the data used in the model;
• testing the accuracy and mechanics of the model and assessing whether this was compliant with IAS 38;
• challenging management’s expected future cash flows by performing benchmarking analysis and engaging our IT consultants to assess the IT architecture; and
• engaging our valuation specialists to assess the appropriateness of the discount rate applied to the asset.

We have assessed the design and implementation and tested the operating effectiveness of key business processes that have been subject to change. For example, we have tested the operating effectiveness of controls at a number of outsourced locations, involving site visits by the audit team. When testing the controls that had been impacted by the transformation programme our testing focused on determining that there remained an appropriate level of review and segregation of duties.

Key observations:
We completed our procedures on the impairment of the intangibles and have not identified any additional impairment. We have observed that the discount rate used was in the middle of a reasonable range and that the expected future cash flows were reasonable. In addition, we completed our controls testing and did not identify any significant control weaknesses.

1) We have identified the estimation of the ultimate number of large bodily injury claims as being inherently uncertain. Management also exercises significant judgement in assessing the impact of market factors which could materially impact the valuation of excess bodily injury claims and hence the insurance reserves. On 27 February 2017 the Lord Chancellor announced a reduction in the Ogden discount rate used to value lump sum injury claim settlements from 2.5% to minus 0.75%. This change in discount rate has a significant impact on the valuation of large bodily injury claims, as detailed on page 137.

We have gained a detailed understanding of the end to end reserving process, including the process for adjusting the reserves as a result of the change in the Ogden discount rate and assessed the design and implementation of selected controls. We have tested the operating effectiveness for actuarial data reconciliations and key management review controls over the reserving process. Our work included attendance at the December Loss Ratio Committee and January Reserve Review Committee.

We have tested the completeness and accuracy of the underlying data used in the Company’s actuarial calculations and the actuarial data used by our Deloitte actuarial specialists in performing their audit through performing reconciliations on the data back to the financial ledger.

Having done this, we worked with those specialists to:

• challenge the suitability of the methodology and assumptions used in estimating the ultimate number of large bodily injury claims by comparing it to industry benchmarks;
• challenge the approach to and the assumptions used in adjusting reserves for the change in the Ogden discount rate;
• assess whether the reserving methodology has been applied consistently across periods;
• evaluate prior year reserve releases and emerging trends for consistency with management’s calculations.

Key observations:
We have determined that the estimate for the ultimate value of large bodily claims to be within a reasonable range. We observed that the frequency and severity assumptions were set towards the prudent end of a reasonable range.

2) The Group is required to settle a proportion of large bodily injury claims as PPOs rather than lump sum payments. The valuation of PPOs has a material impact on the financial statements, with these liabilities totaling £983m (2015: £991m) on a discounted gross basis as detailed in note 2. PPO liabilities are sensitive to the choice of the propensity, severity and discount rate assumptions used. These assumptions require significant management judgement which increases the susceptibility of the balance to material misstatement.

In addition, the Lord Chancellor’s announcement of a reduction in the Ogden discount rate is likely to impact the propensity of a claim becoming a PPO going forward which increases the level of uncertainty in the setting of this assumption at year end.

We have assessed the design and implementation of controls surrounding the propensity of a claim to become a PPO, the discount rate used and the severity assumption applied. In addition we tested the operating effectiveness of a direct and precise business control, performed weekly, over the completeness of the PPO listing; this is a key data input which has a material impact on the PPO assumptions and hence the valuation.

We have worked with our actuarial specialists to:

• challenge management’s assumptions by performing market benchmarking, sensitivity analysis and reviewing historical trends to assess the reasonableness of long term future expectations, including consideration of the change in the Ogden discount rate;
• assess whether the assumptions were consistent in the period or, based on emerging market data, that the assumptions needed to be reconsidered;
• consider the suitability of the methodology and model used in estimating the propensity, severity and discount rate; and
• test the model of a sample of individual PPO’s through re-performance using a Deloitte developed model.

Key observations:
We have determined that key assumptions used in the calculation of the PPO claims reserve are within a reasonable range. We observed that the investment yield component of the discount rate was towards the optimistic end of a reasonable range, but that the propensity and severity assumptions were set towards the prudent end of a reasonable range.

We have identified a key risk for investments relating to the valuation of loan portfolios that are not held at fair value totaling £502m (2015: £388m) as detailed in note 40. During 2016 the Group has invested in a new commercial real estate loan portfolio and has increased investment in its private placement bond portfolio, with investment in infrastructure debt remaining stable at £337m (2015: £330m). These debt instruments are carried at amortised cost and represent a higher credit risk relative to the majority of DLG’s investment portfolio. As a result, there is a risk that the Group does not recognize an impairment provision on these loans when required by IAS 39.

We have assessed the design and implementation and tested the operating effectiveness of the key controls that mitigate the risk over the valuation of investments not held at fair value. Our work included attendance at the year-end impairment review meeting in order to assess the operation of a key management review control.

We have consulted with our credit and valuation specialists to challenge the methodology used in the assessment of credit risk within the investment portfolios that are held at amortized cost against best practice. In addition we performed the following:

• independently calculated the fair value for a sample of assets to identify any significant decreases in fair value below book cost;
• traced a sample of interest payments to bank during the year to test for default or delinquency in interest payments; and
• challenged management on loans of interest where indicators could point to issuer financial difficulty, obtaining evidence to assess whether the position taken by management is reasonable.

Key observations:

We completed our procedures and did not note any indicators of material impairment within the loan portfolios.

See notes 2.13, 2.21, 3 and 26 of the consolidated financial statements for disclosures of related accounting policies, judgments and estimates.

Total claims incurred but not reported for the year ended 31 December 2016 are £1.59 billion. The methodologies and assumptions utilized to develop incurred but not reported reserves involve a significant degree of judgment as there is generally less information available with the related claims. The liabilities are based on the estimated ultimate cost of all claims incurred but not settled at a given date, whether reported or not, together with the related claims handling costs. In addition, classes of business where there is a greater length of time between initial claim event and settlement (such as professional indemnity and other liability classes) also tend to display greater variability between initial estimates and final settlements. A range of methods may be used to determine these provisions.

We focused on this area as the underlying methods include a number of explicit and implicit assumptions relating to the expected settlement amounts and settlement patterns of claims and are subject to complex calculations which include application of management’s judgement.

We evaluated whether the Group’s actuarial methodologies were consistent with those used in the industry and with prior periods. No inconsistencies were noted.

As historical claims data is a key input into the actuarial reserving process, we tested the completeness, accuracy and reliability of the underlying data utilized by management, to support the actuarial valuation.

As part of our risk assessment, we stratified the loss reserves based on the inherent nature of the business class, the size of the class relative to the total reserves, exposure to adverse market development, sensitivity to significant assumptions and the degree of prior year reserve movement.

In order to challenge management’s assumptions and methodologies, we were assisted by our actuarial specialist team members who performed independent re-projections on selected classes of business, particularly focusing on the largest and most volatile reserves as these were considered higher risk. For these classes we compared our independent claims reserve estimates, to those booked by management, and sought to understand any significant differences.

In relation to catastrophe reserves, we also tested the point estimate to specific notifications received confirming event data.

For the remaining classes we performed targeted testing of key indicators and diagnostics to identify and follow up any anomalies and assessed whether there was any audit evidence that was inconsistent with conclusions based on our knowledge from the independent projections noted above. No material differences were identified in our procedures performed.

Risk of fraud in revenue recognition

ISAs (UK & Ireland) presume there is a risk of fraud in revenue recognition because of the pressure management may feel to achieve market expectations. In this regard we focused on transactions which included a judgmental element in their calculation, typical within the life insurance industry, as set out in note 2 to the financial statements.

For the Group, revenue includes fees and commission income on investment contracts and investment advice, premiums on insurance business, and investment return.

We focused specifically on the following:

• Estimates/assumptions made by management that have a direct impact on revenue, as extended by Auditing Practices Board Practice Note 20 – ‘The audit of insurers in the United Kingdom’ (‘PN20’), for example the amortization profile of Deferred Acquisition Costs (‘DAC’) and Deferred Income (‘DIR’).
• Journals posted impacting revenue in addition to the source system determined balance.

We also considered the recognition of income or costs which may have a close relationship to earned insurance premiums and fees, in accordance with PN20, such as reinsurance costs. The Group fully reinsures the UK insurance risk of its closed book of protection business and therefore we focused on whether the £31.5m of premium income associated with this business that was passed on to the reinsurer was complete.

Investment return relates largely to investment contract policyholders and third party holdings in consolidated unit trusts and so has not been included as an area of focus.

We assessed the critical accounting estimates and judgements as set out in note 2 to the financial statements that had a direct impact on revenue. Specifically we:

• substantively tested the deferred income and acquisition costs and the amortization of DIR and DAC, including assessing the future profitability of the products to which the income and acquisition costs related to ensure that profitability was sufficient to support the carrying value of the deferred balances; and
• confirmed substantively the classification of the Group’s products between insurance and investment business to check that insurance product revenue was appropriately included in the consolidated statement of comprehensive income and investment business (except for fees related to investment contract management) was excluded.

We confirmed that there were no new reinsurance arrangements during the year and agreed a sample of premiums ceded to the cash payment to the reinsurer.

Our work on the above areas of judgement was supported by controls testing and substantive procedures over all material revenue streams including:

• reconciling fees on investment business to confirmatory documentation provided by the asset custodian, State Street;
• testing internal controls over the accuracy and occurrence of revenue recognized in the financial statements;
• obtaining and reading the two International Standard on Assurance Engagements (‘ISAE’) 3402 ‘Assurance Reports on Controls at a Service Organization’ reports issued by International Financial Data Services (‘IFDS’) relevant to the Group, in particular focusing on the controls designed to prevent and detect fraud operating over the Group administration systems owned and operated by IFDS;
• testing a sample of journal entries posted throughout the year to revenue accounts that met specific criteria to identify unusual or irregular items.

There were no issues in the ISAE 3402 reports that impacted our audit scope. We also performed testing of controls and substantive procedures on the IFDS policy administration system ‘Salas’ and at Capita in Ireland, where ISAE 3402 reports were not available.

Overall, we noted no material exceptions in our testing and found the judgements taken by the Directors to be reasonable.

The Group financial statements include c. £70bn of investments. A large number of these are straight forward, vanilla instruments and as such do not require judgement in calculating the valuation of the investments.

However c. £2bn of the investments are in derivatives and investment properties, which require management to use estimates and judgements in order to calculate the year end valuation. Due to the magnitude of the balances and the level of judgement involved, this was an area of focus for our audit.

Financial assets including derivatives

SJP outsources investment custodian and valuation activities for financial assets, including derivatives, to State Street. Our audit procedures therefore focused on the evidence available over these outsourced processes.

We obtained and read the International Standard on Assurance Engagements (‘ISAE’) 3402 ‘Assurance Reports on Controls at a Service Organization’ for State Street’s Global Fund Accounting and Custody operations, which provided a description of the systems and controls in place and the results of testing of the operational effectiveness of those controls.

Where appropriate we placed reliance on the controls described in the ISAE 3402 report over the valuation and existence of the financial investments within the portfolio.

We independently re-priced a sample of derivative investments as at year end. We agreed our independent prices to those provided by State Street.

Investment properties

The investment property portfolio is managed by Orchard Street, with title deeds held by DLA Piper and regular valuations performed by CBRE.

We reconciled the listing of properties valued by CBRE to details provided by Orchard Street and also agreed the total valuation to that recorded in the general ledger.

We engaged our in house real estate valuation experts to review the methodology and key assumptions used by CBRE in valuing the portfolio. We also agreed factual inputs to the calculations (eg. Rental income) to tenancy agreements.

We noted no material exceptions in our procedures.

The Central Fund contributions (recognized as revenue in the Central Fund reporting unit), members’ subscription fees (recognized as revenue within the Corporation of Lloyd’s reporting unit) and the market modernization levy (recognized as revenue within the Corporation of Lloyd’s reporting unit) represent 63% of the revenue of the Society.

The market modernization levy is a new revenue stream for 2016.

We focused on these components of revenue because they are recognized based on management’s estimate of the future premium of each syndicate underwriting year. Although not complex, this estimate involves subjectivity with regards to assumptions on the estimation of future premium.

The future premium is estimated based on historical development trends. From this analysis, management selects an extrapolation factor, which is then applied to the current year.

There is also a risk of error in terms of the calculation of the estimated revenue as this is a manual calculation.

(Refer also to note 2 to the financial statements)

We considered the appropriateness of Lloyd’s revenue recognition policy for the market modernization levy based on the contractual terms between Lloyd’s and its syndicates and found it to be appropriate and in line with the IFRS accounting framework.

We evaluated the reasonableness of the assumptions applied in the estimation of the future premium for the three revenue streams. As future premium is estimated based on the development trend of historical premium, we traced a sample of the historical premium data to the market system which records audited market results. We tested Information Technology General Controls over the market system, where no exceptions were noted. Based on our work on the assumptions, we found that management’s analysis supported the extrapolation factor that was selected.

We tested the accuracy of the application of the assumptions to the underlying data and recomputed the total Central Fund contributions, members’ subscription fees and market modernization levy.

We found management’s assumptions to be reasonable. No exceptions were noted in our testing of underlying data or our recalculation of revenue.

We focused on this area because it represents 77% of the total assets of the Society and so the valuation of financial investments has a significant impact on the financial statements.

We focused in particular on the non-publicly traded investments in the Society’s portfolio, which includes investments held as statutory insurance deposits by Additional Securities Limited (‘ASL’), and other non-publicly traded investments in the Corporation and Central Fund. These investments are hard to value because quoted prices are not readily available.

(Refer also to note 16 to the financial statements)

For these hard to value investments we assessed both the methodology and assumptions used by management in the calculation of the year end values as well as testing the governance controls that the Council of Lloyd’s has in place to monitor these processes. The testing included performing the following procedures:

• assessing the methodology and assumptions used within the valuation models;
• the assumptions used against appropriate benchmarks and investigating significant differences; and

• testing the operating effectiveness of the valuation controls over assets managed by third parties.

Our work and the evidence provided identified no significant issues.

We have identified a risk in revenue recognition in respect of the profit commission class of income. This revenue line contributed £54.3m to group profit before tax in the year (2015: £85.4m) and is recorded on the basis of management’s year end calculations, which are subject to the same underlying estimation as the insurance claims reserves. There is, therefore, a greater risk associated with this class of transactions than more automated revenue streams which are recorded at source through a large number of individually insignificant transactions.

Amongst the most significant inputs to the calculations are the loss ratios resulting from the gross insurance reserves booked in the financial statements, which is discussed as a separate risk, above. This is because the setting of reserves will affect the claims incurred in a given underwriting year, thus driving the loss ratios which determine the amount of profit commission to be recognized in the year.

Refer to page 57 in the audit committee report where this is included as a significant issue, note 5c and note 3 in the financial statements which refer to this matter

In addition to the procedures outlined above in respect of the valuation of insurance reserves, we reviewed the profit commission calculation in detail, and compared the calculation logic and the inputs included by management to the terms of the relevant co-insurance contracts in force at the balance sheet date. On the basis of this understanding of the contractual terms, we independently formed an expectation for the recorded income using audited loss ratios and expenses data and compared this to that actually recorded by management.

Following the announcement of the new Ogden rate, we reassessed the profit commission calculation to test the loss ratios being used in the final calculations had been adjusted for the impact of the Ogden rate change.

We have also assessed the design and implementation and tested the operating effectiveness of the internal review controls performed by management, which are designed to identify any errors in the inputs or mechanical workings of the calculations.

Finally, we vouched the settlement of profit commissions during the year to cash receipts on bank statements and used this to reconcile the income recorded to the movement in the balance sheet position associated with these arrangements.

The group’s gross insurance claims reserves total £2,031 million (2015: £1,725 million). The judgements which are made by management in determining the valuation of incurred but not reported (“IBNR”) claims reserves are by far the most significant, in terms of their impact on the Group’s financial position. Setting these claims reserves is an inherently subjective exercise and small changes in underlying assumptions may have a material impact on the overall year end result reported.

We have identified a key risk of misstatement in respect of management’s selection of assumptions and estimates underpinning the incurred but not reported (“IBNR”) claims reserves, specifically around periodic payment orders (“PPOs”) and the ultimate severity of large bodily injury claims arising in the UK Car Insurance business. These particular claims result in higher individual claims reserves and are more judgmental in terms of the development of the ultimate losses which will be incurred – due to the longer-term nature of the Group’s exposure – than the lower value attritional losses which are more common in the normal course of business.

Following the conclusion of the government’s consultation on the Ogden discount rate used by courts to calculate the value of lump sum claims to be paid for long-term damages, management needed also to consider the interaction of this change with the assumed propensity of PPO settlements. The timing of the announced rate change from 2.5% to minus 0.75%, on 27 February 2017, meant that there was a high degree of uncertainty associated with projecting the ultimate impact of this change; given that no subsequent experience has yet emerged.

Refer to page 56 in the audit committee report where this is included as a significant issue, note 5d and note 3 in the financial statements which refer to this matter.

We have assessed the design and implementation and tested the operating effectiveness of the key controls which management performs in relation to insurance reserving. This included testing controls over the data provided to the group’s external actuarial expert, the internal challenge of that expert’s work, and the appropriate governance oversight in determining the key assumptions for both the actuarial best estimates and the additional margin applied above the best estimate reserve.

We tested the completeness and accuracy of the underlying claims and exposure data used in the actuarial calculations by performing reconciliations of the relevant data back to audited financial information.

We completed procedures to assess the competence and objectivity of management’s external actuarial expert and involved our own Deloitte actuarial experts to review and provide challenge on the methodology applied and the key assumptions and judgements taken in determining the gross insurance reserves noted above as significant risks.

Our challenge focused on benchmarking the claims severity assumptions adopted by the group against our wider industry experience to assess suitability, evaluating the reserve releases and development trends from previous accident years, to verify the robustness of the reserving policy adopted, and understanding the year-on-year consistency in determining the reserve margins.

Finally, we have reviewed the specific adjustments made to refine the actuarial models for the change in Ogden discount rate and considered the related changes in PPO propensity assumptions.

For the finite lived intangible assets we performed the following procedures:

• We agreed data to underlying documentation on a sample basis.
• Understood and tested the governance process in place to determine the finite lived intangible assets, including testing the associated financial reporting control framework.
• We examined the new amortization and impairment methodology to check it is in accordance with IFRS requirements.
• We tested management’s application of the methodology by re-performing the amortization and impairment calculations as appropriate.
• We assessed the disclosures in the year end accounts.

Based on the work performed, we consider the valuation of the finite lived intangibles to be appropriate.

Given the ongoing market volatility and macroeconomic uncertainty, investment valuation continues to be an area of inherent risk. The risk is not uniform for all investment types and is greatest for the following, where the investments are hard to value because quoted prices are not readily available:

• Commercial mortgage loans (UK Life).
• Equity release and UK securitized mortgage loans (UK Life).
• Structured bond-type investments (France Life).
• Collateralized loan obligations and non-recourse loans (UK Life).

For these hard to value investments we assessed both the methodology and assumptions used by management in the calculation of the year end values as well as testing the governance controls that the Directors have in place to monitor these processes.

The testing included performing the following procedures:

• We agreed data inputs to underlying documentation.
• We evaluated the methodology and assumptions in particular, yield curves, discounted cash flows, property growth rates and liquidity premium used within the valuation models.
• We compared the assumptions used against appropriate benchmarks and investigated significant differences.
• We tested the operation of data integrity and change management controls for the models.
• We used our valuation experts to perform independent valuations, where applicable.

Based on the work performed, we considered the assumptions used by management to be appropriate.

For UK Life (including Friends Life) insurance contract liabilities, the Directors’ valuation of the provisions for the settlement of future claims involves complex and subjective judgements about future events, both internal and external to the business, for which small changes in assumptions can result in material impacts to the valuation of these liabilities.

The work to address the valuation of the UK Life (including Friends Life) insurance contract liabilities included the following procedures:

• We tested on a sample basis the underlying data to source documentation.
• Using our actuarial specialist team members, we applied our industry knowledge and experience and we compared the methodology, models and assumptions used against recognized actuarial practices.
• Understood and tested the governance process in place to determine the insurance contract liabilities, including testing the associated financial reporting control framework.
• We tested the key judgements and controls over the preparation of the manually calculated components of the liability. We focused on the consistency in treatment and methodology period-on-period, across life insurance funds and with reference to recognized actuarial practice.
• We used the results of an independent PwC annual benchmarking survey of assumptions to further challenge the assumption setting process by comparing certain assumptions used relative to the Group’s industry peers.
• Further testing was also conducted on the Annuitant Mortality, Credit Default and Expense assumptions as set out below.

Based on the work performed, we consider that the assumptions used to be in line with recognized market practices and, where appropriate, industry peers.

Annuitant Mortality Assumptions

Annuitant mortality assumptions require a high degree of judgement due to the number of factors which may influence mortality experience. The differing factors which affect the assumptions are underlying mortality experience (in the portfolio), industry and management views on the future rate of mortality improvements and external factors arising from developments in the annuity market.

In addition to the procedures above, in respect of the annuitant mortality assumptions:

• We understood and tested the governance process in place to determine the annuitant mortality methodology and assumptions.

• We tested the methodology used by management to derive the assumptions with reference to relevant rules and actuarial guidance and by applying our industry knowledge and experience.

• We assessed the results of the experience investigations carried out by UK Life (including Friends Life) management for the annuity business to determine whether they provided support for the assumptions used by management.

• We compared the mortality assumptions selected by UK Life (including Friends Life) against those used by their peers.

Based on the work performed and the evidence obtained, we consider the assumptions used for annuitant mortality to be reasonable.

Credit default risk assumptions for corporate bonds and commercial and equity release mortgage assets

UK Life (including Friends Life for corporate bonds only) has substantial holdings in asset classes with significant credit risk, notably, corporate bonds and commercial and equity release mortgages.

Management use an active approach to setting the assumptions. For both corporate bonds and mortgages a long term deduction from the current market yield is made and a supplementary provision is held to cover the risk of higher short term default rates.

In respect of the credit default assumptions :• We understood and tested the governance process in place to determine the credit default risk methodology and assumptions.

• We tested the methodology and credit risk pricing models used for commercial and equity release mortgages by management to derive the assumptions with reference to relevant rules and actuarial guidance, and by applying our industry knowledge and experience.

• We validated significant assumptions used by management against market observable data (to the extent available and relevant) and our experience of market practices.

Based on the work performed, we consider the allowance for credit default risk to be appropriate.

Expense Assumptions

Future maintenance expenses and expense inflation assumptions are used in the measurement of insurance and participating investment contract liabilities and any associated reinsurance assets. The assumptions used require significant judgement.

• We understood and tested the governance process in place to determine the maintenance expense and expense inflation assumptions.

• We tested the methodology used by management to derive the assumptions with reference to relevant rules and actuarial guidance and by applying our industry knowledge and experience, including assessing the appropriateness of a service fee approach to allowance for internal service company arrangements.

• We validated significant assumptions used by management against past experience, market observable data and our experience of market practices.

Based on the work performed, we consider the assumptions for expense risk to be appropriate.

The estimation of non-life insurance contract liabilities involves a significant degree of judgement. The liabilities are based on the best-estimate ultimate cost of all claims incurred but not settled at a given date, whether reported or not, together with the related claims handling costs. A range of methods, including stochastic projections, may be used to determine these provisions. Underlying these methods are a number of explicit or implicit assumptions relating to the expected settlement amount and settlement patterns of claims. This includes assumptions relating to the settlement of personal injury lump sum compensation amounts. Regulators across the globe continue to focus on reserving adequacy for non-life insurers, particularly in the current market. Given their size in relation to the consolidated Group and the complexity of the judgements involved our work focused on the liabilities in the UK General Insurance and Canada markets.

In the UK General Insurance and Canada markets, we assessed the Directors’ calculation of the non-life insurance liabilities by performing the following procedures:

• We tested on a sample basis the underlying data to source documentation.
• Using our actuarial specialist team members, we applied our industry knowledge and experience and we compared the methodology, models and assumptions used against recognized actuarial practices.
• Understood and tested the governance process in place to determine the insurance contract liabilities, including testing the associated financial reporting control framework.
• Our actuarial specialist team members performed independent re-projections on selected classes of business, particularly focusing on the largest and most uncertain reserves. For these classes we compared our re-projected claims reserves to those booked by management, and sought to understand any significant differences.
• For the remaining classes we evaluated the methodology and assumptions, or performed a diagnostic check to identify and follow up any anomalies.

Based on the work performed we consider the non-life insurance contract liabilities to be appropriately valued.

As described in Note 3.1, impairment is assessed based on discounted cash flow projections. For the Bupa Care Services UK and New Zealand cash generating units, cash flow forecasts require a high level of judgement in respect of fee rate and cost of care. In the Bupa Chile and Quality HealthCare cash generating units, key assumptions are revenue growth and gross margin, particularly in respect of assets such as clinics and hospitals which are under development, the discount rate and terminal growth rate. These two businesses were acquired more recently and there is a low level of headroom in the impairment calculations. For intangible assets subject to impairment tests, cash flow forecasts are sensitive to expected benefits to be derived from the assets, and the period over which they will be earned. For intangible assets subject to amortization there is a high level of judgement when determining whether indicators of impairment exist.

For assets subject to impairment tests our procedures included challenging the cash flow forecasts and the underlying assumptions, based on our understanding of the relevant business and the sector and economic environment in which it operates. We compared forecasts to business plans and also previous forecasts to actual results to assess the performance of the business and the accuracy of forecasting. We challenged the forecast periods utilized in the models and performed sensitivity testing using different forecast periods. We compared the Group’s assumptions to externally derived data as well as our own sector knowledge in relation to key inputs such as the projected cash flows for these cash generating units, terminal growth rates, cost inflation and discount rates and applied sensitivities in evaluating the Group’s assessments. Where external valuation specialists were used, we considered the external valuation report and assessed the valuer as an independent expert.

Our own valuation specialists assisted us in evaluating the assumptions and methodologies used by the Group, in particular those relating to terminal growth rates and discount rates, and in evaluating these assumptions with reference to valuations of similar businesses. For intangible assets subject to amortization, we considered indicators of impairment, focusing in particular on the extent to which assets are still being utilized and the levels of customer attrition and operating margin compared to the assumptions applied when the assets were acquired. We assessed whether the Group’s disclosures over the goodwill impairment review, including the disclosures regarding the sensitivity of the outcome of the impairment reviews to changes in key assumptions were appropriate.

DAC represents 2 per cent of the Group’s total assets and involves judgements in the identification of the acquisition costs that may be deferred, the appropriateness of the deferral methodology adopted and the assessment of the recoverability of the asset.

The DAC associated with the US business, which represents 90 per cent of the total DAC, involves the greatest judgement in terms of measurement and recoverability. The amortization and recoverability assessment of the US DAC asset is related to the achieved and projected future profit profile. This involves making assumptions about future investment returns and the consequential impact on fee income.

We used our own actuarial specialists to assist us in performing our audit procedures in this area, which included:

• Evaluating the appropriateness of the Group’s deferral policy by comparing it against the requirements of relevant accounting standards;
• Evaluating whether costs incurred are deferred in accordance with the Group’s deferral policy; and
• Assessing the calculations performed including the appropriateness of the assumptions used in determining the estimated future profit profile and the extent of the associated adjustment necessary to the amortization of the DAC asset. We compared the estimated future profits to the carrying value of the DAC asset to assess recoverability. Our work included assessing the reasonableness of assumptions such as the projected investment return by comparing against the Group’s investment portfolio mix and market return data.

We also considered the adequacy of the Group’s disclosures about the degree of estimation involved in the valuation of DAC.

The Group’s investment portfolio represents 90 per cent of the Group’s total assets. The valuation of the portfolio involves judgement in selecting the valuation basis for each investment and further judgement in determining the appropriate valuation for harder to value investments.

The areas that involved significant audit effort and judgement in 2016 were the valuation of illiquid positions within the financial investments portfolio representing 2 per cent of the Group’s total assets. These included unlisted equity, unlisted debt securities, certain derivatives and loans such as commercial mortgage loans and bridge loans. For these positions a reliable third party price was not readily available and therefore involved the application of expert judgement in the valuations adopted.

We used our own valuation specialists and pricing services to assist us in performing our procedures in this area. Our procedures included:

• Assessing the availability of quoted prices in liquid markets;

• Assessing whether the valuation process is appropriately designed and captures relevant valuation inputs;

• Testing whether associated controls in respect of the valuation process are operating properly;

• Performing our own independent price checks from our own pricing services using external quotes for liquid positions and, where available, for illiquid positions;

• Assessing pricing model methodologies and assumptions against industry practice and valuation guidelines;

• Evaluating the valuation assessment performed by the Group in order to identify any potential impairment in relation to loans; and

• Performing our own assessment of loan files to understand the performance of the loans. We examined the existing and prospective investee company cash flows in order to evaluate whether loans can be serviced or refinancing may be required and considered the impact on impairment testing performed.

We also assessed whether the Group’s disclosures in relation to the valuation of investments are compliant with the relevant accounting requirements and appropriately present the sensitivities in the valuations based on alternative outcomes.

The Group has significant policyholder liabilities representing 85 per cent of the Group’s total liabilities. This is an area that involves significant judgement over uncertain future outcomes, mainly the ultimate total settlement value of long term policyholder liabilities. Economic assumptions, including investment return, credit risk and associated discount rates, and operating assumptions including mortality, morbidity, expenses and persistency (including consideration of policyholder behavior) are the key inputs used to estimate these long term liabilities.

Additionally:

• In the US, the valuation of the guarantees in the variable annuity business is a complex exercise as it involves exercising significant judgement over the relationship between the investment return attaching to these products and the guarantees contractually provided to policyholders and the likely policyholder behavior in response to changes in investment performance.

• In the UK, the valuation of the policyholder liabilities in relation to the annuity business requires the exercise of significant judgement over the setting of mortality and credit risk assumptions.

We used our own actuarial specialists to assist us in performing our procedures in this area.

Key procedures included assessing the Group’s methodology for calculating the policyholder liabilities and their analysis of the movements in policyholder liabilities during the year, including consideration of whether the movements are in line with the assumptions adopted by the Group, our understanding of developments in the business and our expectation derived from market experience.

Our procedures in the US included:

• Considering the appropriateness of the assumptions used in the stochastic models for the valuation of the variable annuity guarantees.

• Assessing assumptions for investment mix and projected investment returns by reference to company specific and industry data and for future growth rates by reference to market trends and market volatility.

• Assessment of assumptions of policyholder behavior, including consideration against relevant company and industry historical data.

Our procedures in the UK included:

• Considering the appropriateness of the mortality assumptions used in the valuation of the annuity liabilities by reference to company and industry data on historical mortality experience and expectations of future mortality improvements, including evaluation of the choice of the Continuous Mortality Investigation (‘CMI’) model and the parameters used in relation to this.

• Considering the appropriateness of the credit risk methodology and assumptions by reference to industry practice and our expectation derived from market experience.

We utilized the results of KPMG benchmarking of assumptions and actuarial market practice to inform our challenge of management’s assumptions in both areas noted above.

Our work on the policyholder liability adequacy test included assessing the reasonableness of the projected cash flows and challenging the assumptions adopted in the context of company and industry experience data and specific product features. We also performed test work to ensure the appropriateness of changes made to the policyholder liability reserving models during the year. We considered whether the Group’s disclosures in relation to the assumptions used in the calculation of policyholder liabilities are compliant with the relevant accounting requirements and appropriately represent the sensitivities of these assumptions to alternative scenarios and inputs.

Within the life businesses in Emerging Markets and Old Mutual Wealth, judgement is required over the variety of uncertain future outcomes affecting policyholder liabilities, including the estimation of economic assumptions, such as investment return, discount rates, and operating assumptions, such as expense, tax assumptions, mortality and persistency and the policy for creating and releasing discretionary margins held.

Our procedures included testing the design, implementation and operating effectiveness of key controls over the measurement and management of the Group’s calculation of insurance liabilities and evaluating the appropriateness of methodologies and assumptions used. We involved our own internal actuarial specialists to assist us in challenging certain assumptions used and the process followed for setting and updating these assumptions, particularly around investment return, discount rates, tax, mortality and persistency assumptions. This included assessing the data used in management’s analysis prepared to set the assumptions, in the context of our own industry knowledge, external data and our views of experience to date, an understanding of which was enhanced through our attendance at the Group’s own internal Independent Review Committee meetings. We also assessed whether the disclosures made relating to the long-term policyholder liabilities is consistent with IFRS and with the methodologies applied by management.