You now have every internal ratio. The last family attaches the market's numbers to the company's. Here you learn to compute them correctly, and what they imply about cheapness or quality is the business of Relative Valuation. Mechanics come first on purpose. Most multiple abuse in the wild is mechanical error.
Per-share foundations. EPS = profit attributable to equity holders of the parent ÷ weighted average shares outstanding. Make diluted EPS your default. It treats options and RSUs as exercised, along with convertibles, so it's the fair denominator. BVPS = common shareholders' equity ÷ shares outstanding. Then:
P/E = price per share ÷ diluted EPS: equivalently market cap ÷ profit attributable to parent.P/B = price per share ÷ BVPS: equivalently market cap ÷ common equity.
How do you calculate enterprise value?
Enterprise Value follows takeover logic. Market cap prices only the equity claim. To own the business outright you'd buy the equity and repay or assume the debt. You'd buy out the preferred holders and the minority shareholders of consolidated subsidiaries, then pocket the cash you find in the till:
Advertisement
EV = market cap + total debt (incl. leases — state your convention) + preferred stock + minority (non-controlling) interest − cash & liquid investments
Every term is there for a reason. Debt is added because the buyer inherits it. Preferred is a fixed claim senior to equity. Minority interest is subtler, and it's the term people forget most. Consolidation puts 100% of a subsidiary's revenue and EBITDA into the group numbers even when the parent owns 70%. If the denominator of your multiple contains 100% of the EBITDA, the numerator must contain 100% of the claims on it, including the 30% owned by outsiders. Take NCI at book value as a first pass. Use market value when the subsidiary is listed, as with Bharti Airtel's stake in listed Airtel Africa, the classic Indian case, which Worked example 4 computes. Cash is subtracted because it comes with the keys. Every caution from the solvency ratios still applies, though. Cash that's restricted or pledged isn't really "in the till", and neither is cash the operations need, as IndiGo showed.
What does a worked enterprise value build look like?
Now a synthetic build, Trishul Industries (₹ cr). It has 50 cr shares at ₹240, so market cap is 12,000. Borrowings are 3,000 plus lease liabilities of 500. NCI is 400 and preference capital 100. Cash and liquid investments come to 1,600.
EV = 12,000 + 3,500 + 400 + 100 − 1,600 = ₹14,400 cr
With EBITDA of 1,800, EV/EBITDA = 14,400 ÷ 1,800 = 8.0×. PAT (parent) of 900 gives EPS of ₹18, so P/E = 240 ÷ 18 = 13.3×. Common equity of 6,000 gives BVPS of ₹120, so P/B = 2.0×.
EV can also come in below market cap. Infosys (illustrative, ~FY25 scale) carries roughly ₹55,000 cr of cash and investments against ~₹8,000 cr of lease liabilities and no borrowings. Its EV is about ₹47,000 cr less than its market cap. A leveraged telecom is the mirror image. Its EV is far above market cap, because lenders have a claim on most of the business before shareholders get anything.
Which multiples should you pair with enterprise value?
The one commandment of multiples is the consistency rule. Match the claim in the numerator to the income in the denominator:
| Numerator (claim) | Belongs with (income available to that claim) | Valid pairings |
|---|---|---|
| Enterprise value (all capital) | Income before interest: sales, EBITDA, EBIT, NOPAT, FCFF | EV/Sales, EV/EBITDA, EV/EBIT |
| Market cap / price (equity only) | Income after interest: net profit, book equity, FCFE | P/E, P/B, P/FCFE |
P/EBITDA is gibberish. It compares the equity holders' price to a cash stream that partly belongs to lenders, so a leveraged firm looks artificially cheap on it. EV/net income fails the same test in reverse. The lease rule from the solvency ratios follows you here too. Lease liabilities inside EV demand a lease-consistent, IFRS 16/Ind AS 116-style EBITDA in the denominator. Put an ASC 842 rent-deducted EBITDA under a lease-inclusive EV and US companies look cheaper than Indian ones for no real reason. Why EV/EBITDA exists at all is a valuation question. It compares engines while ignoring how each one is financed, and it has blind spots, capex above all. That's where the valuation phase opens. For now, your job is to become unable to compute it wrong.