How to Compare Two Companies Properly: A Checklist

To compare two companies properly, first establish that they are genuinely comparable: same core business, same geography and currency exposure, same accounting basis. Only then compare operations over multiple years, normalise the accounting traps, and leave valuation for last.

Most head-to-head comparisons skip that first step, which is why most of them are broken before a single ratio gets computed. The internet is full of “X vs Y: which stock is better?” pieces that put two tickers in a table, colour the higher number green, and declare a winner. That format is fast, and it is usually wrong, because the two columns are not measuring the same thing.

Here is the method, as a reusable checklist.

Step 1: Check that the comparison is even valid

Before any numbers, answer three questions.

Same business? Two companies in the same index bucket can be very different machines. A pure-India passenger carmaker and a global luxury-vehicle conglomerate both sit under “automobiles,” but one comparison is really three: a domestic mass-market business against a domestic commercial and passenger business, a zero-export book against a large overseas one, and a single-brand operation against a multi-brand portfolio. If you cannot describe both companies’ revenue in one sentence each and have the sentences rhyme, the head-to-head is already suspect.

Same geography and currency? A company earning largely in rupees from Indian consumers and a company earning in dollars, euros, and pounds face different demand cycles, different input costs, and different translation effects on every line of the P&L. Their margins can move in opposite directions in the same quarter for reasons that have nothing to do with execution.

Same accounting basis? Consolidated versus standalone is the classic silent break. If one company has large subsidiaries and you compare its standalone statements against a peer’s consolidated ones, revenue, margins, and debt are all incomparable. Segment mix matters too: a company that is 70 percent one segment should not be casually benchmarked against a peer where that segment is 20 percent of the mix.

If the answers diverge badly, you have two options: narrow the comparison to the overlapping segment, or accept that you are comparing capital allocators, not operators, and change the questions accordingly.

Step 2: Compare operations before you touch valuation

Once comparability is established, resist the urge to jump to P/E. Valuation is the last chapter, not the first. Work through the operating engine of each business, and do it over a multi-year window, ideally a full cycle, because a single year flatters whoever had the better year.

The core dimensions:

• Growth. Revenue and profit growth over five or more years, and whether growth came from volume, price, or acquisitions. Acquired growth and organic growth deserve different respect.
• Margins. Level and stability. A business with steady margins through a downcycle is telling you something a peak-year margin cannot.
• Returns on capital. ROCE and ROE over time. This is where the quality difference between two similar-looking companies usually lives.
• Cash conversion. Does reported profit turn into operating cash flow? Persistent gaps between the two are a question, not a footnote.
• Balance sheet. Leverage, working-capital intensity, and how each company funds its growth. Debt-fuelled growth and internally funded growth are different products.

One period proves nothing in any of these. Trends and ranges over years are the unit of comparison.

Step 3: Normalise the traps

Even between two genuinely comparable companies, the raw numbers hide landmines. A proper comparison neutralises them explicitly.

Fiscal and reporting differences. Confirm both companies are on the same year-ends and that you are comparing the same periods. Trailing-twelve-month figures help when reporting calendars or recent quarters diverge.

One-offs. Exceptional gains, asset sales, write-offs, and one-time tax effects can swing a single year’s profit enough to flip a comparison. Strip them, or at least flag them, before reading any profit-based ratio.

Restatements. If either company has restated past figures, decide which series you are using and use it consistently for both.

Share-count changes. Buybacks, QIPs, bonus issues, and splits all move per-share numbers without moving the business. EPS growth that outpaces profit growth, or lags it, is usually a share-count story. Compare absolute profits alongside per-share figures.

Lenders are a different species. For banks and NBFCs, ROCE and EV/EBITDA are meaningless, because debt is the raw material, not the financing. Use the bank toolkit instead: return on assets, net interest margin, cost-to-income, and asset quality (gross and net NPAs, provisioning coverage). Comparing a lender to a non-lender with a common ratio set is a category error, not a shortcut.

Step 4: Now, and only now, compare valuation

With the operating picture and the normalisations done, valuation becomes interpretable. Two rules.

Match the multiple to the business. Price-to-book for lenders, read against ROE, because book value is what a lender’s earnings are built on. EV/EBITDA for capital-heavy businesses, because it looks through differing debt loads and depreciation policies. P/E for most others, but always with growth context: a higher P/E on a faster, more durable grower can be the cheaper stock in any sense that matters.

A gap in multiples is a question, not an answer. If Company A trades at twice Company B’s multiple, the comparison is not finished, it has just become interesting. The gap is the market’s implied claim about relative growth, returns, and risk. Your job is to test that claim against the operating record you just assembled.

The checklist, in one table

Step	Question to answer	Common failure mode
1. Comparability	Same business, geography, currency, accounting basis?	Comparing standalone vs consolidated; mixing domestic and global earners
2. Growth	Multi-year revenue and profit growth; organic vs acquired?	Judging on one strong year
3. Margins	Level and stability across a cycle?	Anchoring on peak-year margins
4. Returns	ROCE/ROE trend over 5+ years (ROA/NIM for lenders)?	Using ROCE on a bank
5. Cash	Does profit convert to operating cash flow?	Ignoring a persistent profit-cash gap
6. Balance sheet	Leverage and how growth is funded?	Treating debt-fuelled and internal growth as equal
7. Normalisation	One-offs, restatements, share-count changes handled?	EPS “growth” that is really a buyback
8. Valuation	Right multiple for the model, gap explained?	Declaring the lower P/E the “winner”

The honest final question

Every comparison should end the same way: what, specifically, would make one of these companies worth more than the other, and is that thing visible in the numbers or does it live only in the story?

Sometimes the answer is in the numbers: consistently higher returns on capital, better cash conversion, a cleaner balance sheet. Sometimes it is a story about a turnaround, a new segment, or a management change, which may be true but is not yet evidence. Knowing which kind of answer you are holding is the entire point of the exercise. A comparison that ends with “A scores 6, B scores 4” has told you less than one that ends with “A is worth more if, and only if, its margin recovery is real.”

The takeaway checklist:

1. Confirm comparability before computing anything: business, geography, currency, accounting basis.
2. Compare operations over multiple years: growth, margins, returns, cash conversion, balance sheet.
3. Normalise one-offs, restatements, fiscal differences, and share-count changes.
4. Use the lender toolkit (ROA, NIM, asset quality) for banks and NBFCs, never ROCE.
5. Compare valuation last, with the multiple matched to the business model.
6. Treat any valuation gap as a claim to test, not a verdict to accept.
7. Finish by asking what would make one worth more, and whether that is in the numbers or the story.

This is education, not advice. The method tells you how to read two companies side by side; what you do with the reading is your own decision.