Two ways to answer “what is this worth”

There are broadly two families of valuation. The first works out what a business is worth from its own future cash flows, in isolation — that’s discounted cash flow (DCF), and it takes the next few posts to build properly. The second is quicker and much more widely used in practice: figure out what the market is paying for similar businesses, and apply that to this one.

That’s relative valuation, or valuation by comparables — “comps.” It’s what people are doing whenever they say a stock looks expensive at 60 times earnings. The logic is borrowed wholesale from property: nobody builds a discounted cash flow model to price a two-bedroom flat in Powai. They look at what the last three similar flats in the building sold for, adjust for the floor and the view, and land on a number.

The whole method turns on one word in that sentence — similar. Get the comparison set wrong and every number after it is decoration.

The formula

There’s no single formula, because relative valuation is a procedure rather than a calculation:

1. Pick a peer set — companies genuinely comparable to the target
2. Pick a multiple  — P/E, P/B, EV/EBITDA, or a sector-specific one
3. Compute that multiple for every peer, on a consistent basis
4. Take the median (not the mean — one outlier ruins a mean)
5. Apply it to the target's own metric
6. Explain every gap between the target and that median

Step 6 is the actual work. Steps 1 to 5 are arithmetic.

The multiples you already know

This blog has a post on each of the three workhorse multiples, so this is a reference table rather than a re-explanation:

Multiple What it compares Best for Breaks down when
P/E Price to earnings per share Profitable, stable companies Earnings are negative, tiny, or distorted by one-offs
P/B Price to book value per share Banks, financials, asset-heavy businesses Most value is intangible (brands, software)
EV/EBITDA Enterprise value to operating profit Comparing across different debt levels Capex needs differ wildly between peers

EV/EBITDA deserves a note here, because it’s the one that most often belongs in a comps table. P/E is computed on the equity value alone, so two identical businesses with different borrowings will show different P/Es purely because of how they’re financed. Enterprise value adds debt back and strips cash out, which puts companies with different capital structures onto a common footing. When your peer set has a mix of debt-laden and net-cash companies, EV/EBITDA is usually the fairer comparison.

Worked example: Desi Bites Foods Ltd at its IPO

Desi Bites listed at ₹640 a share on 15 June 2025. Here’s the full multiple set at that price, using the FY25 figures from the case study:

Multiple Calculation Value
P/E ₹640 / diluted EPS ₹16.72 38.3x
P/B ₹640 / BVPS ₹182.24 3.51x
EV/EBITDA EV ₹6,520 lakh / EBITDA ₹441 lakh 14.8x

Notice how differently the same company looks depending on which lens you pick. The P/E of 38.3x reads as a fairly demanding growth valuation. The EV/EBITDA of 14.8x looks far more modest — because enterprise value subtracts the ₹1,880 lakh of cash sitting on the post-IPO balance sheet, most of it the IPO proceeds themselves. Same company, same day, same price; two defensible-looking answers.

That’s not a flaw to be resolved. It’s the method telling you something real: a large chunk of what an investor pays at ₹640 is cash, not operating business, and any multiple that ignores the balance sheet will miss that.

Worked example: Britannia Industries

Same price and financials used across this blog’s valuation posts — the NSE (National Stock Exchange) close on 30 June 2025, against the audited consolidated FY25 results. Historical, for illustration only.

Multiple Calculation Value
P/E ₹5,851 / EPS ₹90.45 64.7x
P/B ₹5,851 / BVPS ₹180.81 32.4x
EV/EBITDA EV ₹1,40,751.24 Cr / EBITDA ₹3,187.15 Cr 44.2x

Why you cannot simply compare these two

Put the two tables side by side and the temptation is immediate:

Multiple Desi Bites Britannia
P/E 38.3x 64.7x
P/B 3.51x 32.4x
EV/EBITDA 14.8x 44.2x

Britannia trades at nearly ten times the P/B and three times the EV/EBITDA. The naive conclusion — one is cheap, one is expensive — is exactly the mistake this post exists to prevent. Both companies make packaged food in India. That is roughly where the similarity ends, and the gaps explain almost the entire spread:

  • IPO cash sitting in the book. Most of the P/B gap starts here. Desi Bites’ post-IPO book of ₹2,278 lakh includes the ₹1,600 lakh the IPO just raised — about 70% of the book is fresh cash that hasn’t earned anything yet. Cash is worth roughly its book value, so a book stuffed with it drags P/B towards 1x.
  • Return on equity. From the last post, Britannia earns 52.5% on shareholders’ equity against Desi Bites’ 34.0%. A business that compounds equity faster is worth a higher multiple of that equity. But that 34.0% was earned on the pre-IPO equity. P/B = P/E × ROE only holds when the ROE is measured on the same book as the P/B. On the post-IPO book, Desi Bites’ ROE is ₹16.72 / ₹182.24 = about 9.2%, and 38.3x × 9.2% gets you back to roughly the 3.51x above. The ROE gap on its own (roughly 1.5 times) explains only a small slice of a nearly tenfold P/B gap.
  • Scale and track record. Britannia has a century of history, national distribution, and brands people ask for by name. Desi Bites is a fictional mid-sized manufacturer with three years of audited accounts.
  • Size and liquidity. Desi Bites is a ₹8,000 lakh company on the main board, with most of its shares still held by the promoters, so its free float is tiny and the shares trade thinly. Illiquidity means a buyer can’t easily get out, and the market prices that in with a discount.
  • The cash distortion, again. Desi Bites’ EV/EBITDA is held down by IPO cash that hasn’t yet been put to work. Britannia’s balance sheet has no equivalent lump.

Every one of those is a reason the multiples should differ. Relative valuation done properly isn’t the act of noticing a gap — it’s the work of accounting for it, item by item, until you either understand the gap or conclude that you can’t.

About the peer set

Two companies do not make a comps table. A real one needs four to six genuine peers, each with their multiples computed on the same basis, from their own filings — same fiscal year, same treatment of exceptional items, same definition of EBITDA. That last point matters more than it sounds: “EBITDA” is not a defined term under Indian accounting standards, and two data providers will happily give you two different numbers for the same company.

This post deliberately doesn’t invent a peer table. Sourcing one properly means opening five annual reports, and a fabricated table would teach the method badly. If you’re building one yourself, the peer test is worth stating plainly: a genuine peer sells to similar customers, faces similar input costs, needs a similar asset base, and grows at a broadly similar rate. “Also listed under FMCG” — fast-moving consumer goods, a label broad enough to cover both a biscuit maker and a shampoo maker — is not a peer test.

Common mistakes

  • Comparing multiples across sectors. A 60x P/E means something entirely different in fast-moving consumer goods than in cement. Sector norms exist for real reasons — growth rates, capital intensity, and the stability of earnings all differ.
  • Using the mean instead of the median. One peer with a collapsed earnings figure and a 400x P/E will drag a mean somewhere useless. The median shrugs it off.
  • Mixing trailing and forward multiples in one table. A trailing P/E on one company and a forward P/E on the next is not a comparison. Pick one basis and hold it across every row.
  • Forgetting that the whole sector can be mispriced. Relative valuation tells you what a company is worth relative to its peers. If the entire sector is priced for perfection, the cheapest name in it is still priced against that same optimism. This is the structural limitation of the method, and it’s precisely why the DCF posts that follow exist.
  • Treating a low multiple as a finding. Companies usually trade cheaply for a reason — weaker growth, worse returns, governance concerns, or a business in structural decline. The cheap multiple is the beginning of the question, not the answer to it.

Takeaway: Relative valuation prices a company by asking what the market pays for similar businesses — fast, intuitive, and only as good as the word “similar.” The number that comes out is never the finding; the explanation for why your company differs from its peers is. And because the method assumes the peers themselves are sensibly priced, it can never tell you whether an entire sector has lost its mind.