DataStride Blog Start learning free
FinTech

IFRS 9 for FinTech Analysts: A Step-by-Step Guide

Ayush Naik · 5 min read

IFRS 9, the International Financial Reporting Standard 9, tells us how to classify, measure, and impair financial instruments. If you work in FinTech, this framework is really important. It directly affects how financial institutions book revenue, handle risk, and report their financial health, especially when it comes to setting aside money for loan losses.

What is IFRS 9 and Why Does it Matter?

IFRS 9 replaced IAS 39, bringing in a more forward-looking way to account for financial instruments. It has three main parts: classification and measurement, impairment, and hedge accounting. For FinTech pros, the impairment part, especially the Expected Credit Loss (ECL) model, usually has the biggest impact. This model makes companies recognise potential credit losses much earlier than before, based on future predictions, not just losses that have already happened.

Why should you care? FinTech constantly innovates with lending, payments, and investment products. Understanding IFRS 9 helps you make sure financial products are valued correctly and that risks are properly provisioned for. This directly impacts a company's profit, how much capital it needs, and how much investors trust it. For example, a lending platform must accurately predict potential defaults on its loans to follow IFRS 9. This then shapes its lending strategies and pricing.

The Three Pillars of IFRS 9

Let's break down the main parts:

1. Classification and Measurement

This pillar decides how financial assets and liabilities appear and are valued on a company's balance sheet. It's not just about what you own or owe, but how you account for it. For financial assets, IFRS 9 uses a principle-based approach, with two main tests:

  • Business Model Test: How does the company manage its financial assets to create cash flows? Is it to collect contractual cash flows, sell them, or both?
  • Contractual Cash Flow Characteristics Test (SPPI Test): Do the contract terms of the financial asset lead to cash flows on specific dates that are only payments of principal and interest (SPPI) on the outstanding principal?

Based on these tests, financial assets usually fall into one of three groups:

  • Amortised Cost (AC): For assets held to collect contractual cash flows, where cash flows are SPPI. Think of a regular loan where you expect to get principal and interest until it matures.
  • Fair Value Through Other Comprehensive Income (FVOCI): For assets held both to collect contractual cash flows and for selling, where cash flows are SPPI. This might be for certain debt instruments.
  • Fair Value Through Profit or Loss (FVTPL): For all other assets, or if the SPPI test isn't met. This is often the default for stock investments or complex derivatives.

For financial liabilities, classification is usually simpler, often measured at amortised cost, unless they're held for trading or marked as FVTPL.

2. Impairment: The Expected Credit Loss (ECL) Model

This is probably the biggest change IFRS 9 brought. Instead of waiting for a loss to happen (the 'incurred loss' model of IAS 39), IFRS 9 requires companies to estimate and recognise potential credit losses before they occur. This forward-looking approach uses an 'expected credit loss' model. It considers many possible outcomes, the time value of money, and reliable information about past events, current conditions, and future economic predictions.

The ECL model works in three stages:

  • Stage 1: Performing Loans (12-month ECL): For financial instruments where credit risk hasn't significantly increased since the start, you recognise a 12-month ECL. This is the part of lifetime ECL that comes from defaults that could happen within 12 months after the reporting date.
  • Stage 2: Underperforming Loans (Lifetime ECL - Not Credit-Impaired): If credit risk has significantly increased since the start, but the asset isn't yet credit-impaired, you recognise a lifetime ECL. This covers all possible defaults over the financial instrument's expected life.
  • Stage 3: Credit-Impaired Loans (Lifetime ECL - Credit-Impaired): When a financial asset becomes credit-impaired (e.g., it has defaulted), you also recognise a lifetime ECL. Interest revenue is then calculated on the net carrying amount (gross carrying amount minus loss allowance).

Deciding what counts as a 'significant increase in credit risk' and 'credit-impaired' often needs advanced modelling and data analysis. This is where data science and machine learning are super important in FinTech.

3. Hedge Accounting

Hedge accounting is a complex area. It tries to match how you account for hedging instruments (like derivatives) with the items they hedge (like a loan or a future transaction). The goal is to make profit or loss less volatile. IFRS 9 simplifies some parts of hedge accounting, making it more aligned with how companies manage risk. While it's important, it's usually not a daily concern for many FinTech analysts compared to classification and impairment.

A Practical Example: Calculating ECL for a Loan Portfolio

Let's look at a simple example for a FinTech lending platform. You have a portfolio of small business loans and need to calculate the ECL for reporting. To keep it simple, we'll focus on one loan in Stage 1.

Imagine a loan with these details:

  • Principal: ₹1,00,000
  • Interest Rate: 10% p.a.
  • Term: 1 year (monthly payments)
  • Probability of Default (PD) in the next 12 months: 2%
  • Loss Given Default (LGD): 40% (meaning if default happens, 40% of the outstanding amount is lost)
  • Exposure at Default (EAD): We'll assume this is the current outstanding principal for simplicity, ₹1,00,000.

You can calculate the 12-month ECL as:

12-month ECL = PD (12-month) * LGD * EAD

Let's use Python to calculate this:

principal = 100000  # Loan principal
interest_rate = 0.10 # 10% annual interest
term_years = 1

# ECL parameters for Stage 1 (12-month ECL)
pd_12_month = 0.02  # Probability of Default in next 12 months (2%)
lgd = 0.40          # Loss Given Default (40%)
ead = principal     # Exposure at Default (simplified to principal)

ecl_12_month = pd_12_month * lgd * ead

print(f

Topic inspired by ListenData.

More in FinTech