P/E Ratio Explained for Indian Investors

In March 2023, HEG Ltd — India's largest maker of graphite electrodes — traded around ₹184 a share against trailing earnings of ₹27.58 per share. That is a price-to-earnings ratio of about 6.7 (Inve data, FY23). On any screen, on any WhatsApp tip, that is "cheap." Two years later the same stock changed hands near ₹483 — higher — on a trailing P/E of roughly 81 (Inve data, FY25). The stock went up and got vastly more "expensive" at the same time. Anyone who bought the 6.7x because it looked cheap had bought the single most dangerous number in cyclical investing: a low P/E sitting right at the top of the cycle. This is an illustration of how the multiple behaves, not a view on the stock.

That is the thing nobody puts on the screen. A P/E is not a verdict on price. It is a forecast — the market's compressed bet that current earnings are real, will grow, and will last long enough to repay the premium. And for a huge share of companies, that forecast is seeded almost entirely by management's own guidance. Across 15,726 management commitments tracked on Inve over 1,547 listed Indian companies, barely half — about 54.6% — are delivered as stated (Inve data, as of 2026-06-12). A P/E pays full price today for guidance that, on average, comes good about half the time.

This article unpacks what a P/E is actually pricing, why the same number means opposite things in different sectors, and how to test the forecast hiding inside it.

What is the P/E ratio, really?

The mechanical definition is the easy part. The price-to-earnings ratio is the share price divided by earnings per share (EPS) — how many rupees you pay for one rupee of annual profit.

You will meet two versions. Trailing P/E uses the last four reported quarters of EPS — real, audited, backward-looking. Forward P/E uses analysts' estimate of next year's EPS. The gap between them is the whole game. A stock on a trailing P/E of 40 but a forward P/E of 25 is one the market expects to grow earnings ~60% over the next year. That expectation did not appear from nowhere. It was seeded by what management guided on the last concall and propagated through every analyst model that followed.

So the useful way to read a P/E is not "how expensive is this?" It is: "what does this number assume about the future, and who made that assumption?" The answer, more often than retail investors realise, is management.

Why is a P/E really an embedded forecast?

Strip a P/E down and it contains three beliefs, stacked on top of each other.

First, that current earnings are real — not inflated by a one-off land sale, a tax write-back, or a quarter where the company under-provisioned. Second, that earnings will grow at some assumed rate. Third, that the growth will last long enough to repay the premium. A P/E of 15 quietly assumes modest, durable growth. A P/E of 60 assumes a steep growth runway that holds for many years. The number is a single figure standing in for a multi-year earnings story.

Here is the uncomfortable consequence. When you buy a high-P/E stock, you are not betting on the business as it is. You are betting the business hits the trajectory implied by the multiple — and that trajectory was largely set by management's revenue and margin guidance. If you have never checked whether this management actually delivers what it guides, you have outsourced the most important assumption in your valuation to the most interested party in the room. Don't ask the barber whether you need a haircut.

That is why two stocks at the same P/E are not equally priced. One management has guided conservatively for years and quietly beaten; the other has guided big and revised down twice. The first 30x is cheaper than the second 30x, even though the screener shows them identical.

Why does the same P/E mean opposite things across sectors?

The single most common P/E mistake in Indian markets is comparing across sectors as if the number were currency. It isn't.

A consumer-staples company with steady 12–15% earnings growth and a long demand runway deserves a higher P/E than a commodity producer, because its earnings are predictable and durable. Predictability compresses the risk in the embedded forecast. The market pays up for forecasts it trusts.

Now the trap that catches the most people: cyclicals. Go back to HEG. Think of a graphite-electrode maker as a deckhand whose pay is the spread between electrode prices and needle-coke costs — when the steel cycle runs hot, the spread is fat and the wage looks permanent; when it cools, the wage doesn't shrink, it nearly vanishes. The numbers show exactly that. HEG's FY23 earnings — the ₹27.58 EPS that produced the seductive 6.7x — were peak-cycle earnings, and management said so on the call. "In comparison with the last few quarter results, performance this quarter was the strongest in terms of all parameters, volumes, revenue, EBITDA and operating margins," the management opened its Q1 FY23 concall, noting "best ever capacity utilization" of 92% (HEG Q1 FY23 concall, August 2022). That is the sound of a peak.

Then the spread cooled. Operating margins fell from about 25% in FY23 to roughly 7% in FY25 (Inve data), and EPS followed the spread down a cliff.

Source: Inve data (financials FY23–FY25); price at each fiscal year-end. Illustration of how a cyclical P/E moves, not a recommendation on the stock.

Read that table the way the screener never lets you. The price rose across the whole stretch — ₹184 to ₹483 — yet the P/E multiplied twelve-fold, from 6.7x to 81.2x. The multiple didn't balloon because the stock got pricey. It ballooned because the denominator imploded: EPS fell 78% as the cycle turned. The "cheap" 6.7x was an arithmetic illusion produced by dividing a still-low price by peak earnings that were about to evaporate. For a cyclical, a low P/E is most often a warning that you are buying at the top, not a bargain. The number lies because it assumes the current earnings level persists, and persistence is the one thing a cyclical cannot offer.

The general rule, with the cases behind it:

The lesson links two ideas: a P/E is only as trustworthy as the durability of the earnings in its denominator. Where earnings are volatile, the multiple is noise dressed as signal.

One more distortion to flag before you trust any P/E: leverage. Debt amplifies EPS, so a heavily indebted firm can show a flattering P/E that says nothing about how risky the underlying business is — which is why the financials row demands extra care. For debt-heavy companies, check the debt levels first and cross-check with a capital-structure-neutral multiple like EV/EBITDA before reading anything into the P/E.

How do you test the forecast inside a P/E?

If a P/E is a bet on delivered guidance, then the way to price the bet is to check the guidance record. This is the step almost no retail investor takes, because doing it by hand across a portfolio is brutal.

Two real records, two opposite habits.

First, Infosys's FY26 guidance, as parsed by Inve's Concall AI. Illustration of how guidance evolves across a results cycle, not a recommendation on the stock.

A management team that guides low and revises up every quarter is behaving very differently from one that guides high and lets the number slide. The first earns the right to a richer multiple; the second should make you discount its forward P/E, because the EPS estimate feeding that ratio rests on guidance that has historically drifted the wrong way.

Now the other habit — the one that quietly hollows out a multiple. On its Q1 FY26 call, HEG's management put a number on its new green-energy arm: "for the FY '26, you can see consider a revenue of, say, INR500 crores to INR600 crores plus EBITDA of around INR200 crores to INR225 crores" (HEG Q1 FY26 concall). A quarter later, on the Q2 FY26 call, management reaffirmed it — saying they expected the EBITDA to at least double in FY27 versus the FY26 figures (Inve's Promise Tracker logs this Q2 update as on track). Then, on the Q4 FY26 call, the FY26 Greentech EBITDA figure simply wasn't mentioned again. Inve's Promise Tracker marks it ghosted (Inve data, FY26). Not missed with an explanation — just gone. Illustration, not a view on the stock.

That is the difference a screen cannot show you. A guidance number that gets walked up and one that quietly disappears can sit behind the very same P/E. The multiple looks identical; the quality of the forecast inside it is not. And across Inve's tracked universe this is not rare: 1,337 commitments were simply ghosted — never mentioned again on any later call — and 47% of companies (734 of them) have at least one piece of guidance that went silent (Inve data, as of 2026-06-12). When a growth commitment that propped up a forward P/E quietly vanishes, the multiple has nothing left to stand on, and the re-rating down can be violent.

Where this lens could be wrong

The honest counter-case deserves stating better than a sceptic would. A low P/E on a cyclical at the peak is not always a trap — sometimes the market has correctly judged that the cycle has structurally lengthened, and the "peak" earnings are closer to a new normal. HEG itself argues exactly this: management points to "unprecedented" growth in electric-arc-furnace steel capacity outside China as a multi-year demand driver for electrodes (HEG Q4 FY26 concall). If that thesis holds, FY23's earnings were not a peak at all, and the 6.7x really was cheap. The point of this article is not that low multiples on cyclicals are always wrong — it is that the multiple alone cannot tell you which it is. The only way to know is to form a view on where you are in the cycle and whether normalised earnings, not last year's, justify the price. The number outsources that judgement; you cannot.

Equally, plenty of high-multiple compounders have earned and kept their premium by delivering exactly what they guided, year after year — for them the high P/E is a fair price, not a warning. The lens here is a question to ask, not an answer to assume.

What should a five-year owner actually check?

The owner's question is not "is the P/E high or low?" It is: "what must come true for this multiple to be reasonable, and is management on record delivering that kind of thing?"

A practical sequence:

1. Decompose the P/E into its assumed growth — and the ROE behind it. A P/E of 40 in a market that pays ~20x for average growth is implicitly pricing roughly double the growth and/or durability. Write that implied number down. If management's own guidance can't reach it, the multiple is doing the dreaming. Growth is only half the story: a defensible premium rests on high, durable return on equity too — a compounder earns its richer P/E by reinvesting capital at high returns, not by growth alone.
2. Check whether current earnings are clean. Strip one-offs. A P/E built on a tax write-back or asset sale is built on sand — this is where a quality-of-earnings check on profit versus cash flow earns its keep.
3. For cyclicals, normalise. Use mid-cycle earnings, not peak. HEG's 6.7x in FY23 became 81x in FY25 on the same business — the only thing that changed was which point of the cycle the denominator was caught at. A low P/E at the top of the cycle is the market's most reliable bear trap; how to value a cyclical stock walks through the mid-cycle maths.
4. Audit the guidance record. Has this management delivered the growth it guided, or serially revised down and ghosted it — the way the Greentech FY26 EBITDA target quietly went silent? A high P/E paired with a strong delivery record is a different animal from the same P/E paired with dropped commitments.
5. Compare within the sector, never across. A bank's P/E and an FMCG firm's P/E are denominated in different units of risk.

The first three are arithmetic you can do tonight. The fourth — auditing delivery across a 10–15 stock portfolio, quarter after quarter — is the one nobody manages by hand, which is exactly the gap Inve's Promise Tracker is built to close: it holds the multi-quarter memory of what each management guided and whether it actually showed up.

Frequently asked questions

A P/E ratio is the market's compressed opinion about a company's future, and that opinion leans heavily on what management said it would do. The mistake is treating the number as a fact about price. It is a forecast — and like every forecast, it deserves to be checked against the track record of the people who made it. The cheap stock whose earnings are peaking, the expensive stock whose guidance keeps coming good: in both cases the multiple alone misleads, and the delivery record settles the argument. So before you call any number cheap or dear, ask the only question that matters to a five-year owner — what earnings is this multiple really resting on, and has this management ever shown it can deliver them?

Related reading

• DCF Valuation in India: Skip the Spreadsheet Theatre
• How to Value a Bank: Book Value, P/B, ROE, NIM