← Concepts

Concepts · fundamentals

Margin of safety

Pay enough below what something is worth that being wrong still leaves you whole. The single most useful idea in investing — and the one most people skip.

May 18, 2026 #risk#valuation#graham

In 2008, a lot of people who were right about good companies still lost money. They bought HDFC, Infosys, L&T — genuinely excellent businesses — and watched them fall 40, 50, 60%. The businesses were fine. The prices they paid were not. They had no margin of safety.

The idea

Margin of safety is the gap between what you pay and what something is worth. The wider that gap, the more room you have to be wrong — about the business, the timing, the economy — and still not lose money.

Benjamin Graham, who taught Warren Buffett, called it the three most important words in investing. The logic is almost embarrassingly simple: every estimate of value is a guess about the future, and the future humiliates guesses. So you don’t let your guess be the only thing standing between you and a loss. You demand a discount.

Why it matters

Here’s the trap. You build a careful case for a stock. You decide it’s worth ₹100. The market is offering it at ₹98. “Close enough,” you think, and you buy.

But your ₹100 was never precise. It was a range — maybe ₹80 to ₹120, depending on assumptions you can’t fully verify. Pay ₹98 and you’ve left yourself almost no room. If earnings come in 15% light, or the sector de-rates, or you simply mis-judged, you’re underwater — not because the business failed, but because you paid for perfection and the world delivered ordinary.

Now suppose you’d waited and paid ₹65. You can be 20% too optimistic about the whole thing and still not lose money. That cushion — the distance between ₹65 and your worst-case ₹80 — is the margin of safety doing its job.

ESTIMATED VALUE RANGE₹80 (worst case)₹120 (best)pay ₹98 — no cushionpay ₹65MARGIN OF SAFETYroom to be wrong and still be fine

The same stock, two entry prices. Buy near the bottom of your value range and the gap absorbs your mistakes. Buy near fair value and there’s nothing to absorb them.

How to use it

  1. Estimate a range, not a number. If your answer is a single figure, you’re pretending to a precision you don’t have. Think “₹80–120,” not “₹100.”
  2. Buy near the bottom of the range. Your purchase price, not your target price, is the thing you actually control.
  3. Widen the margin when you’re less sure. A stable, boring business needs a smaller discount than a fast-growing, hard-to-predict one. Uncertainty is paid for in price.
  4. Let it make you patient. No stock in your range? That’s a complete, acceptable answer. The discount comes to those who wait for it.

The takeaway

Most permanent losses don’t come from picking bad businesses. They come from paying full price for a good story and having the story arrive merely on time instead of early. Margin of safety is the unglamorous discipline that turns “I was wrong” into “I was wrong, and it didn’t matter.”