Calculateur IFRS 9 de pertes de crédit attendues (ECL)

Suppose you lend a friend 20,000 TRY this time. As collateral, your friend leaves you the keys to a car worth 8,000 TRY if they cannot repay.

Ask yourself three questions in order:

1. How much is at risk?
   If your friend cannot pay one day, the amount owed to you at that moment is exactly 20,000 TRY. Your exposure is 20,000 TRY — before considering collateral, this answers “how large is the debt in the bad scenario?”
2. What is the probability of non-payment?
   You know your friend — work is a bit unstable, but generally reliable. Say there is a 15% chance they cannot repay. That probability is estimated from how often similar people historically failed to pay.
3. If they do not pay, how much do we lose?
   If they default, you still have the car as collateral and can sell it for 8,000 TRY. So 8,000 of the 20,000 is recovered from collateral and the remaining 12,000 is the true loss. Loss rate: 12,000 / 20,000 = 60%.

Now multiply the three:

20,000 TRY (exposure) × 15% (probability of default) × 60% (loss rate) = 1,800 TRY

So you already know that, on average, you may never see 1,800 TRY of this 20,000 receivable — and you book it that way. Banks do the same for millions of loans, one by one, looking at collateral, payment history and economic outlook.

Start with a simple example

Suppose you lend 1,000 TRY to each of 10 friends. All say “I will repay.” But life tells you that probably 1 of those 10 will never repay — maybe they lose their job, maybe something else goes wrong.

Ask yourself: how much money do you really have?

• On paper: you have 10,000 TRY receivable.
• In expected reality: you will probably get only 9,000 TRY back.

That is exactly what Expected Credit Loss (ECL) is: estimating a loss that has not yet happened but is likely, and recognising it in advance.

The same story for companies

Think of a company — trade receivables from credit sales, or a bank’s loan book. The books show “receivables.” In real life:

• Some customers pay late.
• Some never pay (bankruptcy, shutdown, disappearance).
• Some loans become non-performing over time.

If the company books “I will collect everything,” it looks richer than it is, misleading itself, banks, partners and investors.

ECL prevents that illusion: it answers “How much of this receivable/loan will we really not collect?”, expenses that amount, and sets it aside as an allowance (provision).

Why is it mandatory?

It is not optional — it is required by TFRS 9 / IFRS 9. The aim is simple:

Financial statements should show the real situation, not a polished picture.

After the 2008 crisis revealed hidden losses behind “healthy” bank balance sheets, the world moved to recognising probable losses before they crystallise. ECL is that lesson.

How is it calculated? (Briefly)

There are two simple approaches:

1. Trade receivables: apply historical non-collection rates by aging bucket (e.g. “60% of >90 day balances were never collected”) and adjust for forward-looking economic expectations.

2. Loans: ask three questions — how much exposure, what probability of default, and what loss given default after collateral.

Another example: you lend 10,000 TRY and take a watch worth 6,000 TRY as collateral.

• Exposure: 10,000 TRY.
• Probability of default: say 10%.
• Loss given default: after selling the watch for 6,000, only 4,000 is truly lost — a 40% loss rate.

Multiply: 10,000 × 10% × 40% = 400 TRY.

Putting the watch and car examples side by side makes the point clear: the higher the collateral value, the lower the loss rate, because more value is recoverable.

Loans are also split into 3 stages: performing loans use a 12-month horizon; significantly increased risk or credit-impaired loans use lifetime ECL.

What does this achieve?

• The company does not fool itself about how much cash it really has.
• Banks, investors and partners get reliable figures.
• Surprise losses decline — losses appear gradually instead of as a shock.
• Audit and compliance requirements under TFRS 9 are met.

In short

ECL prepares financial statements with realistic collectibility — just as you want to know the real money in your pocket after lending to friends, companies must know their true financial position.