Intrinsic Value Is Not a Number —
It's a Range: Thinking in Probabilities
The most common mistake in equity valuation is not getting the discount rate wrong, or misforecasting growth, or even using the wrong model. It is treating the output as a precise number. "The stock is worth ₹842." No, it isn't. Under one set of assumptions it might be worth ₹842. Under another, ₹620. Under a third, ₹1,100. Intrinsic value is a range, not a point.
This distinction matters enormously for capital allocation. If your single-point estimate says a stock is worth ₹842 and it trades at ₹780, you might call it "undervalued" and buy. But if your honest range is ₹620 to ₹1,100, the current price sits comfortably within that range and offers no clear margin of safety. The decision changes.
Benjamin Graham understood this intuitively. His concept of "margin of safety" was not a fixed percentage discount. It was an acknowledgment that our estimates of value are imprecise, and that the gap between price and the lower bound of estimated value is the investor's true protection against error.
Aswath Damodaran, perhaps the most influential valuation teacher alive, uses Monte Carlo simulations on his DCF models — running thousands of scenarios with varying inputs to produce a probability distribution of value. The median of that distribution is informative. The shape of the distribution — its width, its skew — is even more informative.
"It is better to be approximately right than precisely wrong." — Warren Buffett, paraphrasing John Maynard Keynes
Bruce Greenwald's approach at Columbia Business School takes this further. He layers three independent valuation methods — asset value, earnings power value, and growth value — and only considers a stock attractive when all three converge below the market price. The probability of error drops dramatically when multiple independent lenses agree.
The practical framework we recommend: build your DCF. Then stress-test the three inputs that matter most — typically revenue growth, operating margins, and terminal value assumptions. Create a bull case, a base case, and a bear case. Weight them by probability. Buy only when the probability-weighted value exceeds the current price by at least 25%. That is your margin of safety.
This approach is slower. It requires more work. It will cause you to pass on many stocks that look "cheap" on a single metric. That is the point. The goal is not to be invested in as many ideas as possible. It is to be invested in ideas where the odds are overwhelmingly in your favour.