Refer to Note 2(k), Note 3, Note 26 and Note 27 to the consolidated financial statements.

As at 31 December 2019, loans and advances to customers and other parties of CITIC Bank amounted to RMB4,001.18 billion (approximately HK$ 4 , 466 . 70 billion), and the corresponding allowance for impairment losses was RMB115.96 billion (approximately HK$129.45 billion); investments in financial assets of CITIC Bank amounted to RMB1,559.79 billion (approximately HK$1,741.27 billion), and the corresponding allowance for impairment losses was RMB8.41 billion (approximately HK$9.39 billion).

The balances of loss allowances for the loans and advances to customers and other parties and investments in financial assets represent the management’s best estimates at the balance sheet date of expected credit losses (“ECL”) under HKFRS 9: Financial Instruments expected credit losses models.

The management assesses whether the credit risk of loans and advances to customers and other parties and investments in financial assets have increased significantly since their initial recognition, and apply a three-stage impairment model to calculate their ECL. For corporate loans and advances and investments in financial assets classified into stages 1 and 2, and all personal loans, the management assesses loss allowance using the risk parameter modelling approach that incorporates key parameters, including probability of default, loss given default, exposure at default and discount rates. For corporate loans and advances and investments in financial assets in stage 3, the management assesses loss allowance by estimating the cash flows from the loans.

The measurement models of ECL involves significant management judgments and assumptions, primarily including the following:

• • Criteria for determining whether or not there was a significant increase in credit risk, or a default or impairment loss was incurred;

• • Economic indicators for forward-looking measurement, and the application of economic scenarios and weightings;

• • Management overlay adjustments due to significant uncertain factors not covered in the models;

• • The estimated future cash flows for corporate loans and advances and investments in financial assets in stage 3.

CITIC Bank established governance processes and controls for the measurement of ECL.

For measuring ECL, the management adopted complex models, employed numerous parameters and data inputs, and applied significant management judgments and assumptions. In addition, the exposures of the loans and advances to customers and other parties and investments in financial assets, the loss allowance and provision accrued involve significant amounts. In view of these reasons, we identified this as a key audit matter.

We evaluated and tested the design and operational effectiveness of the internal controls relating to the measurement of ECL for loans and advances to customers and other parties, and investments in financial assets, primarily including:

• • Governance over ECL models, including the selection, approval and application of modelling methodology; and the internal controls relating to the ongoing monitoring and optimization of the models;

• • Internal controls relating to significant management judgments and assumptions, including the review and approval of model selections, parameters estimation, identification of significant increase in credit risk, defaults or credit-impaired loans, and forward-looking measurement;

• • Internal controls over the accuracy and completeness of key inputs used by the models;
  • Internal controls relating to estimated future cash flows and calculations of present values of such cash flows for corporate loans and advances and investments in financial assets in stage 3;
  • Internal controls over the information systems for model-based measurement.

The substantive procedures we preformed primarily included:We reviewed the modelling methodologies for ECL measurement, and assessed the reasonableness of the models selection, key parameters estimation, significant judgements and assumptions in relation to the models. We tested whether or not the measurement models reflect the modelling methodologies documented by the management on a sample basis.

We selected samples, in consideration of the financial information and non-financial information of the borrowers, relevant external evidence and other factors, to assess the appropriateness of the management’s identification of significant increase in credit risk, defaults and credit-impaired loans.

For forward-looking measurement, we reviewed the management’s model analysis of their selection of economic indicators; economic scenarios and weightings employed, assessed the reasonableness of the prediction of economic indicators and performed sensitivity analysis of economic indicators, economic scenarios and weightings.

In addition, we assessed the reasonableness of the selection of significant uncertain factors, its application and measurement in the management overlay adjustments, and examined the accuracy of the relevant mathematical calculations.

We examined major data inputs to the ECL models on selected samples, including historical data and data at the measurement date, to assess their accuracy and completeness. We also performed walk-through testing and reconciliation of the transmission of major data inputs between the models’ measurement engines and the information systems, to verify their accuracy and completeness.

For corporate loans and advances and investments in financial assets in stage 3, we examined, on a sample basis, forecasted future cash flows prepared by CITIC Bank based on financial information of borrowers and guarantors, latest collateral valuations and other available information together with discount rates in supporting the computation of loss allowance.

Based on our procedures performed, in the context of the inherent uncertainties associated with measurement of ECL for loans and advances to customers and other parties and investments in financial assets, the models, key parameters, significant judgements and assumptions adopted by management and the measurement results were considered acceptable.

Refer to Note 2(g), Note 3, Note 27 and Note 50 to the consolidated financial statements.

As at 31 December 2019 , consolidated and unconsolidated structured entities, including non- principal guaranteed wealth management products (“WMPs”) issued and managed by CITIC Bank, are disclosed in Note 27(a) and Note 50(b) to the consolidated financial statements.

Management’s decision on whether or not to consolidate structured entities was based on an assessment of the power of CITIC Bank, its variable returns and the ability to exercise its power to influence the variable returns from these structured entities.

We focused on the consolidation assessment and judgement made by management involving the structured entities for non-principal guaranteed WMPs during our audit as whether or not to consolidate these entities involved significant judgement.

We evaluated and tested the design and operating effectiveness of management’s relevant controls over the consolidation of structured entities for non- principal guaranteed WMPs. These controls primarily included the review and approval of the contractual terms, the results in variable return calculations, and the consolidation assessment conclusions for these structured entities.

We selected samples of structured entities for non- principal guaranteed WMPs and performed the following tests:

• • assessed the contractual rights and obligations of CITIC Bank in light of the transaction structures, and evaluated the power of CITIC Bank over the structured entities;
  • performed independent analysis and tests on the variable returns from the structured entities, including but not limited to commission income and asset management fees earned, retention of residual income, and, if any, liquidity and other support provided to the structured entities;
  • assessed whether CITIC Bank acted as a principal or an agent through analysis of the scope of the decision-making authority of CITIC Bank, its remuneration entitlement, other interests CITIC Bank held, and the rights held by other parties;
  • evaluated and examined on the appropriateness of disclosures relating to structured entities in the consolidated financial statements.

Based on the audit procedures performed above, we considered that management’s assessment and disclosure on the consolidation of structured entities with non-principal guaranteed wealth management products were appropriate in all material respects.

Refer to Note 2(k), Note 3 and Note 50(d) to the consolidated financial statements.

During the year ended 31 December 2019, CITIC Bank entered into transactions which involved transfers of financial assets including securitisation transactions, structured transfers on assets usufruct, and transfers of non-performing loans.

Management analysed the contractual rights and obligations of CITIC Bank in connection with such transfers, and assessed the extent to which the associated risks and rewards of ownership were transferred by using models to determine whether the de-recognition criteria were met. Where necessary, CITIC Bank assessed whether it had relinquished its control over the transferred financial assets to determine whether the de-recognition criteria were met.

The de-recognition assessment relating to the transfer of financial assets involved significant judgement from management, and as such, we focused our audit on the de-recognition of these financial assets.

We evaluated and tested the design and operating effectiveness of the relevant controls over transfers of financial assets, including the review and approval of the structure designs and contractual terms of the transactions, the approval of models for testing the transfer of risks and rewards of ownership and the key parameters and assumptions used in the models, as well as the review and approval of management’s assessment results.

We selected samples and read through transaction agreements to assess the contractual rights and obligations of CITIC Bank, and whether CITIC Bank transferred the rights to receive contractual cash flows from the financial assets, or the transfer qualified for the “pass through” of those cash flows, to independent third parties.

We also assessed the appropriateness of the models, the parameters and assumptions, the discount rates, and the variability factors. We also tested the mathematical accuracy of the calculations.

For financial assets where CITIC Bank neither retained nor transferred substantially all of the risks and rewards associated with their ownership, we analysed whether CITIC Bank had relinquished its control over these financial assets, and if CITIC Bank had a continuing involvement in these transferred financial assets.

Based on the procedures performed above, we found management’s assessment over de-recognition of transferred financial assets acceptable.

Refer to Note 3 and Note 9 to the consolidated financial statements.

An impairment assessment was undertaken on the Sino Iron Project (“the Project”) by management as at 30 June 2019, with an update assessment on 31 December 2019.

Management has assessed the recoverable amount of the Project using the Fair Value Less Cost of Disposal (“FVLCD”) method as consistent with the approach taken previously. As a result, management has determined that no further impairment charge is required.

In the impairment assessment, the most significant areas of judgement applied by management relate to:

• • The production profile of the Project (including production rates as the Project ramps up, ore grades and operating and capital expenditures);
  • Iron ore prices (inclusive of base price and premium on product grade);
  • The discount rate adopted in the valuation;
  • Foreign exchange rates, particularly between Australian and United States dollars.

As the impairment assessment involves significant assumptions and judgements, we regard this as a key audit matter.

In evaluating management’s valuation of the Project we undertook the following procedures:

• • Evaluated the reasonableness of management’s judgement in relation to non-current asset impairment indicators, cash flow forecasts and the adoption of the FVLCD model;
  • Assessed whether management had included all appropriate assets and liabilities in the cash generating unit with appropriate consideration of tax impact;
  • Compared assumptions adopted in cash flow forecasts on production, future capital and operating expenditure with approved Life of Mine Plans, operating budgets and, where applicable, actual performance outcomes achieved to date;
  • With the support of our valuation experts, benchmarked key market related assumptions included in the valuation model, being base price, foreign exchange rates and the discount rate, against external market data; assessed the forecast premium on product grade with actual premiums achieved to date; and validated the competence and objectivity of the third party experts utilised by management to develop these assumptions;
  • Performed sensitivity analysis on the key assumptions.

Based on the above procedures, we found the assumptions and judgements applied by management to be reasonable and consistent with the audit evidence we obtained.