#1: What Expected Credit Loss (ECL) Really Means — Through My Journey in Credit

I didn’t start my career in risk modelling.

I’m an accountant by training. My early years were spent in accounting and finance roles, working on things like OPEX and CAPEX planning, budgets, and variance tracking. Numbers, yes. But not credit models.

Statistically, my toolkit was basic. Mean, median, standard deviation. Some idea of correlation, regression, and probability from textbooks. Nothing fancy. No machine learning. No advanced econometrics.

So when I say this, I mean it honestly:

If I could understand ECL and work with it comfortably, almost anyone in finance can.

ECL looks complex mainly because of how it’s presented, not because of what it’s trying to do.

At its core, ECL is built around one very practical question:

If this borrower fails to pay in the future, how much will I realistically lose?

Everything else exists only to answer this question in a structured, regulator-acceptable way.

Before we talk about ECL, let’s talk about provisioning

When I was in core finance roles, provisioning wasn’t a complicated idea. It was almost common sense.

If a business knows that certain losses happen regularly, it should recognise those losses in advance, not wait for a specific event.

Think of a trader.

Every year, some stock gets damaged. Some becomes obsolete. Some gets stolen. It doesn’t happen to every item, and you don’t know exactly which item it will be. But over time, it happens with predictable regularity.

If the trader’s history tells him that losses are around, say, 5% each year, good accounting would expect him to provide for it every year.

He doesn’t wait for the damage to show up item by item. He recognises the expected loss based on experience.

That’s what provisioning is at its most basic level.

Now, lending is similar in spirit, but harder in execution.

A lender also knows a basic truth:

Some borrowers will not repay in full.

But credit losses are messy:

• you don’t know which borrower will default,

• you don’t know when it will happen,

• and even after default, you may recover some money.

So provisioning in lending can’t be as simple as “5% of inventory”.

It needs a method. And that’s exactly what ECL is trying to be: a structured way to do provisioning for a business where losses are real, but uncertain.

Why ECL exists at all

Before ECL, the system relied on the incurred loss approach.

Under incurred loss, losses were recognised largely when a loan visibly went bad. Missed payments, restructuring, NPA classification. In India, provisioning was also guided by regulator-prescribed percentages for different asset classifications.

This worked fine in calm times. It was practical and easy to implement.

But anyone who has tracked portfolios closely knows the hidden issue:
risk builds before it shows.

The irony is that under such rule-based provisioning, a conservative bank and a high risk-taking NBFC could end up providing at broadly the same rate, despite having very different underlying risk profiles.

By the time stress shows up in the books, the deterioration has already been forming quietly for months, sometimes years.

After the 2008 crisis, regulators across the world accepted that this approach was too backward-looking. IFRS 9 globally, and Ind AS 109 in India, were introduced to force lenders to recognise losses earlier using forward-looking information.

So ECL is not some “fancy model requirement”.

It’s simply the accounting world trying to apply the same provisioning logic we use everywhere else, to lending.

The three building blocks that never change

No matter how advanced the model or how thick the policy document, ECL always rests on three ideas.

Probability of Default (PD)
What are the chances that this borrower will default over a given time period?

Loss Given Default (LGD)
If default happens, how much money do I actually lose after recoveries, collateral, and time delays?

Exposure at Default (EAD)
How much will be outstanding when default happens, considering EMIs, utilisation, and prepayments?

Put together:

ECL = PD × LGD × EAD

This single line carries the entire framework.

Lifetime ECL and the reality of credit behaviour

One concept that confused me early on was lifetime ECL.

Why look beyond the next 12 months?

Because credit risk does not move linearly. A loan can look perfectly fine for a year and then deteriorate rapidly due to changes in income, business cycles, or external shocks.

Lifetime ECL exists to capture that delayed risk. It forces you to acknowledge stress when it is emerging, not when it has already materialised into default.

Stages are just a way to increase honesty

Ind AS 109 divides loans into three stages, but in practice they’re just a way to scale conservatism.

Stage 1: Risk looks stable. Recognise 12-month ECL.
Stage 2: Risk has meaningfully increased. Stop assuming business as usual. Recognise lifetime ECL.
Stage 3: Default has occurred. Lifetime ECL continues, with changes in income recognition.

Stages are not accounting tricks. They are a discipline to stop optimism from creeping in as risk rises.

The missing piece most beginners overlook: macroeconomic overlays

This is where ECL moves beyond borrower-level analysis.

Credit losses don’t occur in isolation. They are heavily influenced by the broader environment: interest rates, inflation, employment, commodity prices, and liquidity conditions.

Ind AS 109 explicitly requires ECL to be forward-looking, which means incorporating macroeconomic information.

In practice, this means adjusting PDs, and sometimes LGDs, for scenarios like:

• economic slowdown or recovery

• rising interest rates increasing EMI stress

• sector-specific downturns

• changes in household income or MSME cash flows

Macro overlays don’t need to be overly complex to be meaningful. Even a simple scenario-based adjustment can significantly improve realism.

ECL is not about heavy maths

You don’t need to be a statistician to understand or even build an ECL model.

You need:

• logical thinking about default risk over time

• a realistic view of recoveries

• a clean exposure schedule

• and an honest macro lens

I came into ECL from an accounting and planning background, not a modelling one. The clarity came from understanding the why, not memorising formulas.

That’s also how I teach it: plain language, simple Excel, and real-world examples.

A final thought

ECL is often made to sound intimidating. It isn’t.

It’s a framework that forces lenders to think clearly, early, and honestly about risk.

If you’re in finance, credit, or accounting, learning ECL changes how you see loan books. It connects underwriting, portfolio performance, and macro reality into one picture.

And if someone like me, with an accounting-first background, could make that transition, you absolutely can too.

If you want to go deeper and build a full ECL model step by step in Excel, I cover this practically in my live course. Details are available on the course page.