The Most Overused Stat in Indian Investing
"The stock is trading at 30x PE โ it's too expensive."
You'll hear this at every chai stall discussion about markets, every CNBC panel, every WhatsApp forward from your uncle who's been in markets since 1992. And sometimes it's right. Often it's used without the context that makes it meaningful.
P/E ratio = Price per share รท Earnings per share.
If Infosys's share price is โน1,500 and it earned โน60 per share in the last 12 months, the P/E is 25. That means you're paying โน25 for every โน1 of annual profit.
Simple. But the interpretation is where most investors go wrong.
What P/E Actually Measures
A high P/E isn't automatically expensive. It means the market expects the company to grow โ fast. Investors are willing to pay a premium today because they believe the "E" (earnings) will be much higher in 2โ3 years.
Think of it this way: a stock with P/E 40 that grows earnings at 35% per year will have a P/E of 22 in just two years (assuming the price stays flat). That's actually cheaper than a "cheap" P/E 15 stock that isn't growing at all.
This is the PEG ratio concept โ P/E divided by growth rate. A P/E of 30 with 30% growth is a PEG of 1.0, which many value investors consider fair. A P/E of 15 with 5% growth is a PEG of 3.0 โ that's actually more expensive on a growth-adjusted basis.
P/E Differs Wildly by Sector on NSE
This is critical for Indian stocks, and where a lot of casual analysis breaks down.
Banking and NBFCs: Typically trade at P/E of 8โ18x, with PSU banks even lower. Banks use leverage structurally, so price-to-book is often more relevant.
IT services (TCS, Infosys, Wipro): Premium P/E of 20โ35x, justified by consistent free cash flow, high ROE, and dollar-denominated revenues.
FMCG (HUL, Nestle, Britannia): Routinely above 40โ60x. These are consumer staples with pricing power and near-zero business risk. The premium is structural, not irrational.
Metals and commodities: Low P/E (5โ12x), but cyclical. The "cheap" P/E at the top of a commodity cycle is a trap โ earnings collapse in the next cycle.
Comparing a PSU bank's P/E of 8x to HUL's 50x and concluding the bank is "6x cheaper" is like comparing the fuel efficiency of a truck to a sports car โ they're different vehicles.
Trailing P/E vs Forward P/E
The P/E you see quoted on NSE's website is typically the trailing P/E โ based on actual earnings from the last 12 months. This is backward-looking.
Forward P/E uses analyst estimates for the next 12 months. It's more useful but less reliable, since estimates can be wildly wrong. When a company is going through a transition (new product cycle, margin recovery, regulatory tailwind), the forward P/E can look very different from the trailing.
A stock where trailing P/E is 35 but forward P/E (based on expected earnings growth of 40%) is 25 is not expensive โ it's transitioning to a higher earnings base.
When P/E Is Meaningless
P/E breaks down completely in three situations:
1. Negative earnings: If a company is loss-making, P/E is negative or undefined. This is common for new-age companies (Paytm, Nykaa in early years) or turnaround stories. Don't compare them using P/E.
2. One-time items: If a company had a huge one-time gain or write-off, the reported EPS is distorted. Always check if the earnings number underlying the P/E is "normalised" or includes one-offs.
3. Cyclical peaks: A steel company reporting record profits during a commodity supercycle will show a low P/E โ but those earnings aren't sustainable. The P/E is cheap for a reason: the market knows.
How WealthLenseAI Uses P/E
In our AI Forecast for every Nifty 50 and NSE stock, P/E is one input into a fuller picture:
- Trailing P/E is cross-referenced with the sector median
- Forward P/E is estimated based on projected EPS growth
- DCF intrinsic value gives an independent cross-check that doesn't rely on P/E multiples at all
- The combination of all three helps flag whether a stock's valuation is stretched or compelling
No single number tells you whether to buy. But P/E โ used correctly, in context, sector-adjusted โ is one of the most efficient first-pass filters you have.