Skip to content

The same return on equity, arrived at two completely different ways

A vehicle financier and a housing financier both report a return on equity of about 21%. One earns a spread three times the other’s and borrows half as much. The identity that separates them takes two lines, and it decides which of them survives a bad credit year.

Fundamental AnalysisAdvanced14 min read
Browse Fundamental Analysis(116)

You are comparing two lenders because a screen put them next to each other. Both report a return on equity a shade over 21%, both have grown their books at a similar rate for three years, and both are described in the same paragraph of the same brokerage note as high-quality compounders. One of them finances used commercial vehicles at yields near 17% and carries about six rupees of assets for every rupee of its own net worth. The other writes home loans at 9.6% and carries twelve. The identical headline return is not a coincidence and it is not evidence that the two businesses are alike. It is the arithmetic of a lender doing what it always does, which is to multiply a small number by a large one — and the two companies have chosen opposite ends of both.

Think of it like this
Two transporters, the same annual profit

One man owns a single lorry, runs it on a specialised route and makes ₹6 lakh a year on it. Another owns twelve lorries on thin, competitive freight and also makes ₹6 lakh. Same profit, same year. Now suppose diesel rises ten per cent. The first man’s route can absorb it. The second man is running twelve vehicles on margins that cannot, and the same shock that dents one business closes the other.

In the market

Return on equity says how much was earned per rupee of owners’ money. It does not say whether that came from a fat margin on a small balance sheet or a sliver of margin on a very large one. The number is identical and the exposure to a bad year is not.

Spread and margin are not the same number

These two get used interchangeably and they are different measurements with a precise gap between them. The lending spread is what the company earns on its assets minus what it pays on its borrowings — two rates, subtracted. The net interest margin is net interest income divided by average interest-earning assets. The margin is always the larger of the two for a solvent lender, and the reason is worth understanding rather than memorising: part of the asset base is funded by the shareholders’ own money, and that part carries no interest cost at all.

Net interest margin − Lending spread = Cost of funds × (share of assets funded by net worth)
Yield on assets
Interest income ÷ average interest-earning assets
Cost of funds
Finance cost ÷ average borrowings
Lending spread
Yield on assets − cost of funds
Net interest margin
Net interest income ÷ average interest-earning assets
What the identity assumes
That the interest-earning assets are funded by borrowings plus net worth and nothing else. Real balance sheets carry other liabilities too, so treat the identity as close rather than exact — it explains the direction and roughly the size of the gap, which is all it is needed for

Example: Yield 17.06%, cost of funds 8.25%, so the spread is 8.81 points. Average assets ₹17,000 crore of which ₹14,300 crore is funded by borrowings and ₹2,700 crore by net worth. The margin is 10.12% — and the gap of 1.31 points is exactly 8.25% × (2,700 ÷ 17,000). The equity is doing free work, and the margin counts it while the spread does not.

The two-line identity that separates the two companies

Return on equity = Return on assets × Equity multiplier
Return on assets
(Net interest margin + fee income − operating cost − credit cost) × (1 − tax rate), all as a percentage of average assets
Equity multiplier
Average total assets ÷ average net worth — how many rupees of assets each rupee of owners’ money carries
Why it matters here
Everything a lender does to its business changes the first term. Everything a lender does to its balance sheet changes the second. The two are then multiplied, so leverage magnifies whatever the business did — including the bad years
As a percentage of average assetsThe vehicle financierThe housing financier
Net interest margin10.123.20
Fee income1.120.35
Operating costs(4.24)(1.00)
Credit cost(2.53)(0.25)
Pre-tax return on assets4.472.30
Less tax at 25%(1.12)(0.58)
Return on assets3.351.72
Equity multiplier (times)6.312.0
Return on equity21.1%20.7%
Illustrative figures for two lending models, not two particular companies — the first column is the same company worked through in the previous lesson. The point is the route to the answer rather than the answer. A branch-heavy business financing used commercial vehicles for small operators carries an enormous operating cost ratio and a real credit cost, and pays for both out of a very wide margin. A home loan business runs on a margin a third as wide, spends almost nothing to originate and almost nothing on losses, and gets to the same place by borrowing twice as hard.

Why the identical answers behave differently

Worked example
The same one-point deterioration in credit cost
Both lenders above, credit cost rising by one percentage point of assets
Vehicle financier — pre-tax return on assetsThe one point comes straight off4.47% → 3.47%
Vehicle financier — after taxTaxed at 25%3.35% → 2.60%
Vehicle financier — return on equityMultiplied by 6.321.1% → 16.4%
Housing financier — pre-tax return on assetsThe same one point, off a much thinner base2.30% → 1.30%
Housing financier — after taxTaxed at 25%1.72% → 0.98%
Housing financier — return on equityMultiplied by 12.020.7% → 11.7%
Damage to the vehicle financierPainful and survivable4.7 points of return on equity
Damage to the housing financierThe same shock, almost exactly twice the effect9.0 points of return on equity
The leverage that produced the matching headline number also multiplies the bad news, and it multiplies it by the same factor in both directions. This is not an argument that high leverage is wrong — a home loan against a self-occupied property genuinely does lose less than a loan against a used truck driven by the man repaying it, and lending against low-loss collateral without leverage would earn nobody anything. It is an argument that the two identical numbers were never comparable, and that the sensible follow-up question is not “which return on equity is higher” but “how much has to go wrong before this return on equity halves”. That question has an arithmetic answer for any lender, from a table you can build out of one annual report in about twenty minutes.

The trap inside a fast-growing book

  • Look at credit cost against the book as it was, not as it is. Provisions this year divided by the average book of two years ago is crude and far more honest than the published ratio in a fast-growing lender.
  • Ask what proportion of the book was written in the last twelve months. Where that proportion is very high, no ratio drawn from loss experience means much yet, because most of the book has not had time to fail.
  • Fee income deserves its own look, and it does not all behave the same way. A processing fee that is integral to what the loan yields is folded into the effective interest rate and recognised across the life of the loan, so it never appears as a lump. Commission for distributing somebody else’s insurance, and charges for services rendered, are earned when the service is rendered — which is at or near origination. A lender whose fee income is growing much faster than its book is earning a rising share of its return at the point of sale, which flatters a growth year and thins out in a slow one.
  • Operating cost per rupee lent is a business-model fact, not an efficiency score. A branch network that lends ₹80,000 at a time cannot have the cost ratio of a company writing ₹40 lakh mortgages, and comparing the two on cost-to-income says nothing about how well either is run.
Check yourself

A lender reports net interest margin up from 9.4% to 10.1%, while its lending spread is unchanged. It raised fresh equity during the year. What is the most likely explanation?

Simple bhasha mein
Do bhai, ek jaisa munafa, alag himmat

Dono ka return on equity lagbhag 21% hai. Pehla purane truck kharidne walon ko loan deta hai: margin mota, par har rupya apne paise ka 6.3 guna hi kaam karta hai. Doosra ghar ke loan deta hai: margin patla, par 12 guna. Ab dono ke doobne wale loan sirf 1 percentage point badh jaaye. Pehle ka return 21.1% se 16.4% — 4.7 point ka jhatka. Doosre ka 20.7% se 11.7% — 9 point, yaani dugna nuksaan, wahi ek point se. Screener pe dono ka number ek jaisa dikhta hai. Sawaal "kiska zyada hai" nahi hai — sawaal yeh hai ki kitna bigadne pe yeh aadha ho jaata hai.

What to remember
  • Lending spread subtracts two rates; net interest margin counts the equity-funded assets as free — the gap between them is arithmetic, not performance.
  • Return on equity is return on assets multiplied by the equity multiplier, and the second term magnifies bad years exactly as much as good ones.
  • The same headline return on equity can come from a wide margin with low leverage or a thin one with high leverage; ask how much must go wrong to halve it.
  • Credit cost is understated in a fast-growing book because the losses come from a book that was smaller.
  • A capital raise lifts net interest margin without touching lending spread — check the spread when margin expansion is being celebrated.
Finished this lesson?

Mark it done to track your progress through the curriculum.