Insights

IFRS 9: the two ways of calculating ECLs

Since IFRS 9 replaced IAS 39, entities have been getting to grips with new reporting requirements. We look at the methods and considerations along the way.

For a financial asset, the expected credit loss (ECL) is the difference between the contractual cash flows that are due to an entity and the cash flows that an entity expects to receive. 

The calculation of ECLs applies to financial assets that are measured under amortised cost or at fair value through other comprehensive income. These assets may be in the form of loans, debt securities or trade receivables.

Financial assets vary from entity to entity depending on the nature of the business and the products they provide. Some entities offer loan products that are long-term in nature and some may be secured on collateral. This is common for banks and consumer lending companies.

In other entities, such as manufacturing and retail companies, their most common financial asset may be trade receivables. These are amounts billed by companies to customers upon delivery of goods or services and are usually due within 12 months.

 

The approach fits the entity


With these different types and characteristics of financial assets, there is the question ‘How should entities calculate the ECLs for each type?’. IFRS 9 permits two approaches: the general approach and the simplified approach.

The general approach is complex. It usually involves, among other things, calculation of the probability of default, considering whether there have been significant increases in credit risk, and forward-looking macro-economic information.

The simplified approach involves the calculation of historical loss rates.

 

The general approach


The general approach is used by banks and other financial institutions that have longer-term financial assets. There are three functions that need to be considered:

  • Exposure at default (EAD). This is the amount of principal to which the calculated probability of default rate and the loss given default rate is applied. A repayment rate is calculated based on an historic analysis of repayments in the period to default. EAD = The principal amount outstanding x (1- the calculated repayment rate in the period to default).  
  • Probability of default (PD). This is an estimate of the likelihood of default over a given period. PD is determined based on the historical loss experience of an entity. This historic PD is then adjusted by a factor, determined by reviewing the historic relationship between key economic parameters such as GDP and unemployment and PD.  Forward-looking macro-economic information relating to, say, future GDP and/or unemployment is then considered and the calculated historical PD is adjusted.
  • Loss given default (LGD). This is an adjustment to the ECL calculation for post-default recoveries.  These can be in the form of cash repayments, proceeds from the realisation of security or sale of the debt to a third party. The LGD is based on an analysis of historical post-default recoveries. LGD = 1- the post-default recovery rate.

The calculation process

Once the three functions are determined, the ECL is calculated as EAD x PD x LGD. The calculation can be either for 12 months or based on the lifetime of the financial asset. This depends on whether there has been a significant increase in credit risk since the date of initial recognition. The credit loss that is calculated on a 12-month basis involves analysis of historical credit losses over 12 months.

But credit loss calculated over the lifetime of the financial asset is derived from historical losses over the life of the asset. The PD calculated on a lifetime basis will be higher than the PD calculated over 12 months. As such, the lifetime ECL will be higher than the 12-month ECL.

 

Three stages


Under IFRS 9, there are three stages of credit risk. Under each stage there is a different prescribed method of calculating the ECL (by using PDs calculated over different periods – 12 months or over the entire life of the financial asset) and recognising interest income:

  • Credit risk – Stage 1. There is no significant increase in credit risk from initial recognition. Only the ECLs within 12 months of a reporting date are calculated. Interest income is calculated on the gross carrying amount of the financial asset.  
  • Credit risk – Stage 2. There is a significant increase in credit risk from initial recognition. ECLs over the lifetime of the financial asset must be recognised. Interest income is calculated on the gross carrying amount of the financial asset.  Under IFRS 9, there is a rebuttable presumption that there is a significant increase in credit risk if a contractual repayment is more than 30 days past its due date.
  • Credit risk – Stage 3. This is where the financial asset has become credit impaired: the point when there is objective evidence of impairment as defined under IAS 39 (the predecessor to IFRS 9).  Examples might include evidence of significant financial difficulty of the debtor, or default. In terms of the ECL, like credit risk Stage 2, these are recognised on a lifetime basis.  Interest income, however, is recognised on the net carrying amount (the gross amount of the financial asset, less the calculated impairment).    

Let’s consider an example.

Company A has a two-year loan receivable from a customer with a gross carrying amount of £2 million and interest rate of 1% per annum payable in two annual instalments.

At a reporting date, Company A has assessed that there has been no significant increase in credit risk from initial recognition. The loan is therefore classified in terms of credit risk under Stage 1.

The PD within 12 months has been calculated based on historical data at 2% and LGD is also calculated based on historical data is 5%. The calculation of ECL would be:


EAD £PD LGDECL £Discount rate PV of ECLs £ 
1,010,0002%5%    1,0101% 1,00012-month ECL
Total ECL 1,000

IFRS 9 requires that ECLs are discounted to the reporting date applying the effective interest rate used at recognition. So, in the above example, the calculated ECL of £1,010 is discounted to £1,000.

If, at the reporting date, Company A has assessed that the loan has suffered a significant increase in credit risk from initial recognition, the loan would be classified in terms of credit risk under Stage 2.

The following calculation assumes that: the PD for loans in Stage 2 within the first 12 months has been calculated based on historical experience at 5%, and then 10% in the second 12 months; and that through a historical analysis of post-default recoveries of loans in Stage 2, the Company has calculated an LGD of 20% in the first 12 months and then 30% for the second 12 months:

EAD £PD LGDECL £Discount rate PV of ECLs £12-month ECL
Lifetime ECL
1,010,0005%20% 10,1001%10,000
1,010,00010%30% 30,3001%    29,703
Total ECL 39,703
In this case, with the loan in credit risk Stage 2, the ECL recognised in the financial statements of Company A would be on a lifetime basis, which in this case is two years. The total ECL charge in the profit and loss account would thus be £39,703. 


 

The simplified approach

Some entities – those with trade receivables, contract assets and lease receivables – do not calculate the PD and LGD separately, but instead use a loss rate approach. This is known as the simplified approach under IFRS 9.

For trade receivables that do not contain a significant financing component, the loss allowance should be measured as equivalent to lifetime ECLs. This is because they are very short-term in nature and are usually due within 12 months. So the 12-month ECL and lifetime ECL would be the same.

For trade receivables or contract assets which do contain a significant financing component, and for lease receivables, the entity can choose between the simplified approach and the general approach.

 

The role of a provision matrix


The loss rate approach allows the use of a provision matrix adjusted for current conditions and future expectations, based on available forward-looking information. The default rates in the provision matrix should be calculated by segmenting the loan portfolio into appropriate groupings, based on shared credit characteristics.

A provision matrix is simply a table that analyses the trade receivables into groupings and applies a calculated loss rate to each one. The groupings can be by product type, which can be sub-analysed into geographic regions. These groups are then, finally, split into aged bandings.

Here is an example of a possible provision matrix:

Product AProduct B
North regionSouth regionNorth regionSouth region
< 30 days past due> 30 days past due< 30 days past due> 30 days past due< 30 days past due> 30 days past due< 30 days past due> 30 days past due
Loss ratesLoss ratesLoss ratesLoss rates
3%5%4%7%2%4%5%8%
IFRS 9 does not provide any specific guidance on how to calculate loss rates. Let’s look at one method. It involves collecting historical data over a period in relation to sales, and losses suffered on those sales.

Let’s assume for Company B that the historical period over which data was collected is three years. The total sales in that period amounted to £3 million and the total losses (sales not paid and written-off) suffered on those sales amounted to £150,000.

To determine the loss rate, the sales receipts are observed moving through different ageing groupings, to determine a loss rate for each grouping as follows:

Ageing groupSales in period/ debtors at start of period £Cash received in the ageing group £Sales carried forward to next ageing group £Historical loss rate
0 days overdue3,000,0002,000,0001,000,0005% (150,000/3,000,000)
1 to 30 days overdue1,000,000400,000600,00015% (150,000/1,000,000)
31 to 60 days overdue600,000250,000350,00025% (150,000/600,000)
61 to 90 days overdue350,000100,000250,00043% (150,000/350,000)
Later than 90 days overdue250,000100,000150,00060% (150,000/250,000)
Sales not paid  150,000Written off

Forward-looking macro-economic information

Under IFRS 9, an entity must incorporate forward-looking information into the calculated historical loss rates. This adjustment involves judgment and may be complex.

But let’s assume that, through statistical analysis of historical data, the unemployment rate has a strong direct relationship with Company B’s loss rates. Let’s say that economists in the country of Company B have forecast unemployment rates to increase from 3% to 5% and Company B’s experience (derived from historical analysis) is that when unemployment increases by 2%, the losses increase by 5%.

We are applying a 4% increase (the mid-range of the economic forecast) to the above example (a 4% increase in unemployment would lead to a 10% increase in the loss rate). Thus, the historical loss rates for Company B are adjusted by forward-looking information as follows:

Ageing groupHistorical loss ratesAdjusted loss rates
0 days overdue5%5.5%
1-30 days overdue15%16.5%
31-60 days overdue55%27.5%
61-90 days overdue43%47.3%
Later than 90 days overdue60%66%
Now that Company B has derived the adjusted loss rates, these rates can be applied to the outstanding balances of trade receivables to determine the ECL for the current period.

Days past dueCurrent outstanding trade receivables balance £Adjusted loss ratesECL £
0 days overdue   15,0005.5%825
1-30 days overdue8,00016.5%1,320
31-60 days overdue2,00027.5%550
61-90 days overdue1,00047.3%473
Later than 90 days overdue50066%330
Total ECL3,498
In practice, there is unlikely to be any strong relationship between future macro-economic indicators and loss rates for trade receivables. This is because of the short-term nature of such receivables compared to the longer-term nature of economic forecasts. But companies need to consider the relationship between the loss rates and future macro-economic indicators to comply with IFRS 9.

Remember that there is no single method prescribed by IFRS 9 when calculating ECLs. But under IFRS 9 the measurement of ECLs must reflect an unbiased and probability-weighted outcome; the time value of money; and reasonable and supportable information that is available without undue cost or effort.