IFRS in Banks: How a Financial Institution's Reporting Differs

Why bank reporting stands apart

A bank has a different balance sheet structure: its assets are mostly financial instruments, and its liabilities are funds raised from customers and creditors. As a result, the standard approaches familiar to a trading company barely work here.

An additional layer is regulatory requirements. A financial institution prepares IFRS financial statements and reports to the regulator at the same time, and these two data sets are built on different logic even though they rely on the same accounting records.

Hence the heightened attention to judgements. A significant share of the figures in bank financial statements are estimates rather than facts, and reporting quality depends on how well supported and reproducible those estimates are.

Classification and measurement of financial instruments

The key fork in applying IFRS to banks is how an instrument is classified. This determines whether it is measured at amortised cost or at fair value, and where remeasurement results go: to profit or loss or to other comprehensive income.

Classification is based on the business model for managing the assets and on the characteristics of the contractual cash flows. A mistake at this step carries through the whole set of statements, so the decision is made at initial recognition and documented.

• loans and receivables from customers
• debt securities in the bank's portfolio
• equity instruments and investments
• derivatives and hedging transactions
• borrowed funds and issued liabilities

Expected credit losses: the central topic of bank reporting

International standards require an allowance to be recognised not when a loss has already occurred, but in advance — based on expected credit losses. The model relies on the probability of default, loss given default and exposure at default.

A separate element is a significant increase in credit risk since initial recognition: when it occurs, the loss measurement horizon is extended. Defining the criteria for such an increase is a matter of professional judgement, and it is what generates the most debate.

The model requires forward-looking information: macroeconomic scenarios and their weightings. This makes the allowance sensitive to assumptions, so their justification and consistent application are scrutinised particularly closely.

Risk disclosures

Bank financial statements contain an extensive block of disclosures without which the balance sheet figures say little. Users need to understand portfolio concentration, asset quality and the institution's ability to meet its obligations.

Disclosures are built not as boilerplate text but as an extension of management procedures: if a bank manages liquidity risk in a certain way, the financial statements should reflect that.

Practice shows that disclosures written in isolation from risk management are immediately visible. They are generic, do not change from year to year and do not explain changes in the portfolio.

• credit risk and loan portfolio quality
• liquidity risk and maturity structure
• market risk, including currency and interest rate risk
• fair value and the levels of its hierarchy
• related-party transactions

Dual reporting: the standards and the regulator

Differences between IFRS financial statements and regulatory reporting are normal: they have different purposes. The regulator looks at stability and capital adequacy, the standards at a faithful presentation for external users.

A problem arises when the differences are not explained. The institution must be able to show the bridge from one data set to the other, otherwise any difference looks like an accounting error.

That is why it is better to reconcile the data sets regularly rather than assembling the reconciliation at the reporting date. This way differences are found while they are still easy to explain.

What most often raises questions in the audit

In practice, most findings concern not arithmetic but the justification of judgements: why these particular criteria for credit risk deterioration were chosen, where the macro scenarios came from, how the model's performance was validated.

The second group of questions concerns data. The expected loss model requires portfolio history, and if the data is incomplete or export formats have changed, the result becomes hard to defend.

• undocumented judgements on instrument classification
• unsupported assumptions in the expected loss model
• incomplete historical portfolio data
• boilerplate risk disclosures unrelated to actual procedures
• unexplained differences from regulatory reporting

How we work with financial institutions

Leader Audit is an audit firm. Over 13 years of practice we have carried out more than 220 audits of companies of various profiles in Uzbekistan.

Work with a financial institution begins with an analysis: what instruments are in the portfolio, how the provisioning model works, what data is available. We then help bring the methodology and disclosures to a state where they withstand an audit, and support the preparation of the financial statements through to issue.