#11: One Loan, Three Years, and the Truth About Lending

Before you start reading, download this Excel model.

Keep it open beside you.

You don't need to understand every formula in it right now. Just look at two things:

the amortisation schedule
the simple P&L summary built on top of it

Amortisation Schedule

3 year P&L view

This article explains what those numbers mean.

Excel helps you see them.

Read both together, and lending will start making sense in a way it usually doesn't.

Most discussions on lending start in the wrong place.

They start with portfolios, ROE, growth, or regulations.
They talk about yields, spreads, and leverage.

But lending does not begin with a balance sheet.
It begins with one loan.

If you truly understand how one loan behaves over time, everything else in lending becomes easier to reason about. If you don't, no amount of ratios will save you.

So let's start there — with a single loan, and with the amortisation schedule you see in the Excel.

The product: a simple unsecured MSME loan

Assume a very common product:

Loan amount: ₹100
Loan Type: Unsecured
Borrower: MSME
Interest rate: 18% per year
Tenure: 36 months
Repayment: EMI

Nothing exotic. This is the kind of loan thousands of NBFCs give every day.

Now let's understand what actually happens once this loan is disbursed. You may open this excel model for better understanding.

Every section of this article refers back to this same loan.
Every number you see in the Excel flows from this.

What EMI really means (look at the schedule)

Scroll to the amortisation schedule in the Excel.

You'll notice three things immediately:

the EMI is constant every month
interest starts high and falls
principal repayment starts low and rises

This is what "equated" in EMI really means.

The total payment stays the same.
The composition changes.

Interest is always calculated on the outstanding loan amount.
As outstanding falls, interest automatically falls.

This is not a product feature.
It is basic arithmetic.

Month 1: connect the words to the numbers

Look at month 1 in the schedule.

Opening balance: ₹100
Interest at 18% (monthly): roughly ₹1.5
EMI: roughly ₹3.6
Principal repaid: the difference i.e. ₹2.1

So after just one EMI:

the borrower no longer holds ₹100
your outstanding exposure has already reduced to ₹97.9

From this point onward, the loan is shrinking every month.

This single fact drives almost everything that follows.

The first misconception the Excel quietly fixes

Many beginners instinctively think:

"I lent ₹100 at 18%, so I earn 18 every year."

But if you keep scrolling down the amortisation schedule, you'll see why that intuition is wrong.

You earn interest on the outstanding amount, not on what you originally disbursed.

And the outstanding keeps falling.

Over three years, something important happens

If you look at this loan over the full 36 months:

In the beginning, outstanding is close to ₹100
In the middle, it is somewhere around ₹55–60
Towards the end, it approaches zero

So for most of the loan's life, your exposure is much lower than ₹100.

This leads us to the most important concept in lending economics.

Average outstanding: pause here and really see it

Now look at the summary section in Excel that shows average outstanding.

Over the full three years, the average outstanding of this ₹100 loan is roughly ₹55.8, not ₹100.

This is the most important idea in lending economics.

Because:

interest income is earned on this average
operating costs sit on this average
credit losses hit this average
capital is tied up against this average

Disbursement is an event.
Outstanding is a state.

Lending economics lives in the state, not the event.

Why am I being so slow here

Before we talk about:

profitability
credit losses
opex
cost of debt
leverage

We need this picture to be clear.

A lending business is nothing more than many such loans stacked together.

If you don't understand the life of one loan, a portfolio will only hide the truth.

"Am I profitable?" is not a yes or no question

Now that we understand how one loan behaves over time, let's ask the question most people jump to too quickly:

Am I profitable on this loan?

At first glance, it feels obvious.

Loan amount: ₹100
Interest rate: 18%
Sounds attractive, right?

But lending profitability is not decided by the interest rate alone. It is decided by what survives after everything else takes its share.

So let's layer reality onto the same loan.

First, how much do I really earn?

From above, we established something critical:

This ₹100 loan does not stay at ₹100
Over 3 years, the average outstanding is roughly ₹55.8

Let's take ₹55.8 as a working number.

At 18% yield, your annual interest income is roughly:

18% × 55.8 ≈ ₹10.1 per year.

Not ₹18.
Not even close.

Now introduce operating costs

A lending business does not run on goodwill.

Even for a simple unsecured MSME loan, you incur costs for:

sourcing
credit underwriting
disbursement
collections
systems
compliance
people

Let's assume a very reasonable number:

Opex = 3% of average outstanding per year

That means:

3% × 55.8 ≈ ₹1.7 per year

This cost exists whether the loan performs or not.

So now your ₹10.1 of interest income is already down to:

₹10.1 − ₹1.7 = ₹8.4

Then comes credit loss. Always.

Even the best lenders have losses.

Some borrowers will default.
Some will restructure.
Some will settle late.

Let's assume a lifetime credit loss of 4% on this unsecured MSME loan.

Over 3 years, that is roughly:

~1.3% per year on average outstanding. For simplicity I have assumed even spread of credit cost over 3 years. However, in real life this is seldom the case.

So annual credit cost is approximately:

1.3% × 55.8 ≈ ₹0.7 per year

Now subtract this as well:

₹8.4 − ₹0.7 = ₹7.7

Readers can read more on credit cost here.

Pause and observe what just happened

Let's recap, slowly:

You lent ₹100
You charge 18%
Your average exposure is ~₹55.8
After opex and credit loss, you are left with ₹7.7 per year

This ₹7.7 is not profit yet.

This is what the loan generates before funding costs and before capital considerations.

Already, the original "18% yield" feels very far away.

Let's ignore taxes for now.

This is why "am I profitable?" is the wrong question

At this stage, asking "am I profitable?" doesn't make sense.

The better questions are:

Profitable after what?
Profitable for whom?
Profitable on whose money?

Because lending is funded by capital, and capital always has a cost.

We haven't even touched that yet.

But something important is already clear

Even before introducing debt or leverage, one thing is obvious:

A high lending rate does not guarantee a high-quality business.

What matters is:

how slowly or quickly the principal comes back
how heavy your operating structure is
how much loss the product naturally carries

This is why two lenders charging the same rate can have completely different outcomes.

Why we are not adding debt yet

Many discussions rush straight to:

cost of funds
leverage
ROE

That is a mistake.

If a loan does not make sense at this stage, leverage will not save it.
It will only magnify the mistake.

So before touching debt, we must answer one more question.

The right question at this point

Given:

an 18% yield
this EMI structure
realistic opex
realistic credit loss

Is there enough economic surplus left to even think about borrowing and scaling?

That is where we go next.

Before debt, understand your true ROE

So far, we have analysed one loan without talking about borrowing.

That was intentional.

Before asking whether debt helps, we must first answer a simpler and more fundamental question:

If I fund this loan entirely with my own money, what return does it generate?

This is your base ROE.
Everything else in lending builds on this.

The surplus the loan generates

Let's return to same example.

From the same ₹100 unsecured MSME loan:

Yield: 18%
Average outstanding: ~₹55.8

Annual economics (approx):

Interest income: ₹10.1
Opex (3%): ₹1.7
Credit loss (~1.3%): ₹0.7

What remains is:

₹7.7 per year

This ₹7.7 is the entire economic surplus generated by the loan before anyone is paid for capital.

No debt yet.
No leverage yet.

Just the business itself.

ROE without debt (all-equity lending)

If you fund the loan fully with equity:

Equity deployed (average): ~₹55.8
Annual surplus: ₹7.7

Your return on equity is:

₹7.7 ÷ ₹55.8 ≈ 13.8% per year

Pause here.

This number is extremely important.

What this number really means

This 13.8% is the maximum sustainable return this loan can produce before leverage.

It already accounts for:

EMI structure
Time
Opex
Credit loss

No financial engineering can push the underlying economics beyond this.

So one rule becomes clear:

If your cost of debt (pre-tax) is higher than ~13.8%, debt will destroy value.

This is not a theory. It's arithmetic.

Now ask the right question about debt

Most people ask:

"My borrowing cost is X%. Will leverage improve my ROE?"

That's the wrong starting point.

The correct question is:

Is my cost of debt lower than my unlevered ROE?

In our case:

Unlevered ROE ≈ 13.8%

So:

Debt at 10% → potentially helpful
Debt at 13% → marginal, fragile
Debt at 14% → value destructive

Debt does not create returns.
It only reshapes returns that already exist.

Why this framing matters so much

This is where many lending businesses quietly go wrong.

They think:

Yield > cost of debt → leverage works

But the correct comparison is:

Unlevered ROE vs cost of debt

Because:

Yield is earned on shrinking balances
Losses hit principal
Opex is continuous
Equity absorbs volatility

Once you see this, many aggressive lending strategies suddenly look reckless.

A subtle but powerful insight

Notice something interesting here.

Even though:

Loan yield is 18%

The business itself only generates:

~13.8% ROE without debt

The remaining gap is eaten by:

EMI structure
Opex
Credit loss

This is why high-yield products are not automatically high-return businesses.

Now we are ready for leverage

Only now does it make sense to introduce debt.

Because now we know:

What the loan can generate on its own
What the ceiling return looks like
What borrowing costs are even eligible to be used

Leverage is an amplifier, not a saviour

At this point, we know two critical things about our loan:

The loan, on its own, generates ~13.8% ROE if funded fully with equity
Any debt we introduce must cost less than this, or it will destroy value

Now let's bring leverage into the picture — slowly and honestly.

What leverage actually does

Leverage does not improve the quality of a lending business.

It does only one thing:

It reallocates the same economic surplus between debt and equity.

If the surplus is weak, leverage magnifies the weakness.
If the surplus is healthy, leverage can improve equity returns — up to a point.

Introduce debt, gently

Assume:

Cost of debt: 10%
This is comfortably below our unlevered ROE of 13.8%

Now let's try different levels of leverage on the same loan.

Remember:

Average outstanding ≈ ₹55.8
Annual economic surplus before capital costs ≈ ₹7.7

Case 1: No leverage (all equity)

We already did this:

Equity: ₹55.8
Surplus: ₹7.7

ROE ≈ 13.8%

This is our baseline.

Case 2: Moderate leverage (50% debt, 50% equity)

Debt: ₹27.9
Equity: ₹27.9

Debt cost:

10% × 27.9 = ₹2.8

Surplus left for equity:

₹7.7 − ₹2.8 = ₹4.9

ROE:

₹4.9 ÷ ₹27.9 ≈ 17.6%

ROE improves from 13.8% to 17.6%.

This is leverage doing its job.

Case 3: Higher leverage (75% debt, 25% equity)

Debt: ₹40
Equity: ₹14

Debt cost:

10% × 40 = ₹4.0

Surplus left for equity:

₹7.7 − ₹4.0 = ₹3.7

ROE:

₹3.7 ÷ ₹14 ≈ 26.4%

ROE improves again.

At this point, many people conclude:

"More debt is always better."

That conclusion is wrong.

Why ROE does not rise forever

Look closely at what just happened.

Total surplus stayed the same: ₹7.7
Debt took a bigger share of it
Equity got a smaller base to divide the remainder

This works only as long as:

debt is cheap
surplus is stable
losses behave as expected

Now introduce reality.

The fragility problem

At 75% leverage:

Equity is only ₹14
A small shock hurts a lot

If:

credit loss increases slightly
collections slip
opex creeps up
borrowing cost rises

The ₹3.7 buffer collapses quickly.

This is why highly leveraged lenders look great in good times and struggle suddenly in bad times.

Leverage compresses margin of safety.

The peak ROE insight

There is always a point where:

additional debt adds risk faster than it adds return

That is the peak ROE point.

Beyond this:

volatility rises
equity becomes fragile
the business becomes dependent on perfect execution

This peak is different for:

secured vs unsecured products
short vs long tenure
stable vs volatile credit behaviour

But it always exists.

This is why leverage is strategic, not tactical

Good lenders don't ask:

"How much debt can I raise?"

They ask:

"At what leverage does my product remain robust?"

That answer depends on:

unit-level economics
loss behaviour
opex discipline
funding stability

Not ambition.

Don't trust the conclusions. Stress them.

If you've read this far, you already know more about lending than most people who talk about it.

But there's one last step that matters.

You should not take any of the numbers in this article as truth.

Because lending is not about finding the right answer.
It's about understanding what breaks first.

Why this cannot stay theoretical

Everything we've done so far used:

one loan
one tenure
one yield
one opex
one credit loss
one cost of debt

Real lending never behaves so politely.

Small changes compound:

a slightly longer tenure
a small delay in collections
one bad underwriting cohort
a sudden rise in funding cost
a temporary spike in opex

On paper, nothing looks dramatic.
In reality, equity feels everything.

That's why intuition alone is dangerous in lending.

Use the model like a stress lab, not a calculator

I'm sharing this simple Excel model for you to play around with your assumptions.

Not as a "solution".
Not as a forecasting tool.
Not as a template to copy blindly.

Use it as a sandbox.

Here's how to use it properly:

First, recreate the exact numbers we discussed
Then change one thing at a time
Observe what moves the most
Pay attention to where ROE peaks and then collapses

The goal is not to maximise ROE.
The goal is to understand fragility.

Things I strongly encourage you to try

If you want to really understand lending, try this:

Increase tenure while keeping yield constant
Increase credit loss slightly and see how fast equity returns vanish
Raise cost of debt by 1% and observe how leverage turns hostile
Reduce opex and notice how powerful operational discipline really is
Compare a high-yield, high-loss product with a low-yield, low-loss one

Each experiment teaches something no article can.

What this model is — and what it is not

Let's be clear.

This model is:

a unit-level thinking tool
a way to build first-principles intuition
a foundation for understanding ROA, ROE, and leverage

It is not:

an ECL model
a regulatory model
a portfolio simulator
a substitute for judgement

Those come later.

Without this foundation, they remain abstract.

The real takeaway

Lending is not mysterious.

It is just unforgiving.

It rewards clarity and discipline.
It punishes sloppy thinking, especially when leverage is involved.

If you understand one loan deeply, you can understand a thousand.

Everything else is scaling.