Why debt makes P/E a bad yardstick.
EV/EBITDA measures the all-in price — debt, cash, and earnings together
Two big companies, same analyst coverage, completely different metrics
In 2021, a CA in Chennai was comparing Airtel and TCS on his screener. Both large, both analyst-covered. But Airtel's P/E showed nothing useful — the company was reporting losses. A colleague said, 'That's why we use EV/EBITDA for telecom.' He'd been investing six years and had never stopped to ask why.
P/E is market price divided by earnings per share. When earnings are zero or negative — as happened with Airtel through years of spectrum debt payments — P/E becomes meaningless. Even when earnings are positive, P/E ignores how a company was financed. Two businesses can earn the same profit, one debt-free and one carrying heavy loans. P/E treats them identically.
Enterprise Value fixes the numerator. EV is market capitalisation plus total debt minus cash and equivalents. If you wanted to buy a company outright, you'd pay the market cap to existing shareholders, inherit all its debt, and keep whatever cash sits on the balance sheet. EV captures that real, total acquisition cost.
EBITDA fixes the denominator. It strips out interest, tax, depreciation, and amortisation — the financing choices and the accounting treatments. What remains is raw operating earning power: what this business generates from running its operations, regardless of how it borrowed money to get here.
EV/EBITDA then answers one clean question: how many years of operating profit would it take to pay for this business at today's price? Ten years means a ratio of 10. Twenty years means 20. Lower is cheaper — but only when you're comparing companies inside the same sector.
Two shops, same profit, different real price
Two traders in Crawford Market each earn ₹10 lakh a year. Ramesh owns his shop outright — no debt. Suresh borrowed ₹50 lakh to set his up. If you price both at a P/E of 10, each appears to cost ₹1 crore. But to truly own Suresh's shop, you inherit his ₹50 lakh loan. The real cost is ₹1.5 crore. EV/EBITDA catches this immediately. P/E does not. Debt is part of what you're buying — the metric you use had better count it.
Why this matters
Every screener puts P/E in the first column. For companies carrying real debt — telecom, cement, steel, infrastructure — that number quietly misleads you. EV/EBITDA forces you to count the debt you'd be inheriting and credit the company for cash it holds. Indian cement companies historically trade at 10–18x EV/EBITDA depending on cycle position; analysts rarely cite P/E as their primary lens there. When Airtel's P/E turned negative after AGR spectrum payments, EV/EBITDA became the consensus metric across brokerages. Knowing which number fits the business separates a screener skim from real analysis.
Move the sliders. Watch EV/EBITDA reveal the real price.
Two companies sit in front of you with identical P/E ratios. Adjust their debt and cash — watch EV/EBITDA diverge and show which business you're really paying more to own.
Same P/E, different price to own
Both companies trade at an identical P/E of 20×. Company A carries ₹3,000 in debt — a buyer must absorb that debt, so its true acquisition cost (EV) rises and its EV/EBITDA reaches 12.8x. Company B holds ₹3,000 in cash — a buyer effectively gets that cash back, so its EV falls and its EV/EBITDA drops to 7.2x. P/E saw them as equals; EV/EBITDA reveals a 5.6× gap in what you actually pay per rupee of operating profit. Assumptions: Market cap ₹10,000 Cr and EBITDA ₹1,000 Cr held fixed for both — illustrative only.
The all-in cost is what the screener hides
Where people go wrong
- Comparing EV/EBITDA across very different sectorsCement at 12x and software at 25x can both be fair value. Capital intensity, growth rates, and reinvestment needs differ so sharply across sectors that cross-sector comparisons produce no useful signal.
- Trusting EBITDA without checking capex needsEBITDA doesn't deduct capital expenditure. A steel plant or telecom company spending heavily on physical assets looks more profitable on EBITDA than it really is. Only compare companies with similar capex intensity.
- Using P/E for debt-heavy companiesHeavy debt means heavy interest charges, which suppress reported earnings. P/E then looks high — or meaningless — even when the underlying business is reasonably priced. EV/EBITDA removes the financing noise.
- Treating any low multiple as a buy signalA single low EV/EBITDA means nothing without the sector baseline. Always ask: what does this sector normally trade at, and why does this company trade differently?
EV adds debt and subtracts cash — it's the true all-in cost to acquire a business, not just market cap.
EBITDA strips out interest, tax, and depreciation — leaving raw operating earning power before financing choices.
Cement, telecom, and steel trade primarily on EV/EBITDA because P/E breaks down when debt is large.
P/E appears on every screener by default, so investors anchor to it — even when the business they're studying has borrowed heavily to operate. Debt is real. The metric that ignores it is not.