Intrinsic value is the estimated "true" economic worth of an asset derived from its underlying fundamentals — for a business, the present value of the cash it can generate over its remaining life — as opposed to its quoted market price. It is the anchor of value investing: Benjamin Graham's entire framework rests on the premise that a stock has a worth independent of, and frequently divergent from, what Mr. Market will quote on any given day, and that buying when price sits well below that worth is the path to satisfactory long-run returns. The central tension is that intrinsic value is unobservable and must be estimated — every method is a model fed by forecasts, so the number carries the analyst's assumptions and biases with it. Warren Buffett calls it "the only logical approach" to valuation while conceding it is "an estimate rather than a precise figure" that two honest analysts will compute differently.

How it's calculated / formed

There is no single formula; intrinsic value is whatever a chosen model outputs. The main families:

• Discounted Cash Flow (DCF): Value = sum of projected free cash flows discounted to present value, plus a discounted terminal value. The discount rate is typically WACC (firm-level) or cost of equity (equity-level). Most of the value in a standard 5–10 year DCF sits in the terminal value (often the majority of the total). Terminal value is usually computed either by perpetuity growth (Gordon: TV = FCF × (1+g) / (r − g)) or an exit multiple.
• Dividend Discount Model (DDM): A DCF variant valuing the stock as the present value of expected future dividends; the Gordon Growth version is V = D₁ / (r − g). Best for stable dividend payers.
• Residual Income Model (RIM): Value = current book value + present value of future "residual income" (earnings above the equity charge, Net income − r×Equity). An advantage cited by the CFA curriculum is that the terminal value is a smaller share of total value than in DCF, reducing reliance on far-future guesses.
• Graham Number (a quick floor, not a true intrinsic value): √(22.5 × EPS × Book Value per Share). The 22.5 is Graham's product of a maximum 15× P/E and 1.5× P/B.
• The "Graham formula": V = EPS × (8.5 + 2g), with g the expected 7–10 year growth rate; Graham's 1974 revision multiplied by 4.4 / Y (Y = current AAA corporate bond yield). Note: Graham presented this as a critique of growth-rate extrapolation, not a recommended stock-picker — a widely propagated misreading.

How it's used in practice

Intrinsic value is rarely used as a precise target. The disciplined practice is to pair the estimate with a margin of safety — only buying when price sits materially below the estimate, so that modeling error is absorbed by the discount. Graham's logic was explicitly to protect against "unknown risks." A commonly cited rule of thumb is a 20–50% discount to the intrinsic estimate before buying, though the figure is a convention, not a derived constant. Practitioners typically:

1. Triangulate, don't pinpoint. Compute a range of intrinsic values across methods and across a bull/base/bear set of assumptions, rather than trusting one point estimate. 2. Run sensitivity tables. Because output is dominated by the discount rate and terminal growth, analysts vary those inputs explicitly to see how fragile the conclusion is. 3. Compare to price, then wait. Value emerges only when the market offers a wide enough gap; much of the work is patience.

Two philosophical camps share the concept. Graham's original "cigar-butt" / deep-value approach buys statistically cheap assets below liquidation or book value. Buffett (under Charlie Munger's influence) shifted toward paying a fair price for high-quality businesses with durable competitive advantages — "wonderful company at a fair price" rather than "fair company at a wonderful price" — where the intrinsic value reflects compounding earnings power, not a balance-sheet bargain.

Adoption, debate & evidence

Intrinsic value is foundational to fundamental/value investing and is institutionally mainstream (every equity-research DCF computes one). But it is genuinely contested on two fronts.

Estimation reliability. Aswath Damodaran (NYU Stern), among the most authoritative voices on valuation, stresses that DCF is the most theoretically rigorous method and the most easily abused: inputs are "noisy and difficult to estimate" and "can be manipulated by the analyst to provide the conclusion he or she wants." Because small changes in terminal growth or discount rate swing the answer dramatically, the same business can be "valued" almost anywhere — the classic garbage-in, garbage-out problem. Damodaran's own caveat: the danger in terminal value is not the growth rate per se but the excess returns assumed alongside it.

Does buying below intrinsic value pay? The academic proxy for this is the value premium (Fama-French HML — high book-to-market beating low). It was robust historically, but U.S. value materially underperformed growth through the 2010s; the Fama-French value (HML) factor returned roughly −2.6% annualized over 2010–2019 (Advisor Perspectives, "A Lost Decade for the Fama-French Factors," 2020), prompting "value is dead" claims. Notably, Eugene Fama and Ken French themselves concluded there is "too much volatility in monthly returns to decisively determine" whether the premium has disappeared. Value rebounded in 2022–2023 as rates rose. Honest read: the concept of intrinsic value is sound accounting/finance; the empirical edge of mechanically buying cheap is regime-dependent and currently unsettled.

Strengths & limitations

Strengths: Forces explicit assumptions about growth, margins, reinvestment and risk; provides a price-independent anchor that resists market mania and panic; the margin-of-safety discipline structurally limits permanent capital loss.

Limitations: The output is only as good as the forecasts — and terminal value, the dominant component, is the least knowable. It performs poorly for early-stage, high-growth, or asset-light firms whose value lives in uncertain future options; the Graham Number in particular is unsuitable for growth, tech, financials and REITs where book value is misleading. The single most common misuse is false precision — treating a model's $87.42 output as fact and skipping both the sensitivity analysis and the margin of safety. A wide intrinsic-value range honestly drawn beats a confident point estimate. A secondary trap is reverse-engineering inputs to justify a stock already wanted (Damodaran's "valuation as sales pitch").

Sources

• Aswath Damodaran (NYU Stern), "Ten Myths About Discounted Cash Flow Valuation" and "Myth 5.5: The Terminal Value ate my DCF!" — terminal-value dominance, input manipulation, excess-returns caveat.
• Warren Buffett, Berkshire Hathaway owner's manual / shareholder commentary (via CNBC, Yahoo Finance) — intrinsic value as estimate, "wonderful company at fair price."
• CFA Institute — Discounted Dividend Valuation and Residual Income refresher readings (model mechanics, RIM terminal-value advantage).
• Wikipedia, "Graham number" and "Benjamin Graham formula"; StableBread — √(22.5×EPS×BVPS) derivation; EPS×(8.5+2g) and 1974 ×4.4/Y revision, plus the "warning not recommendation" caveat.
• Advisor Perspectives, "A Lost Decade for the Fama-French Factors" (2020) — value factor ~−2.6% annualized 2010–2019. Morningstar ("It's Too Soon to Say the Value Premium Is Dead"), Chicago Booth Review, Institutional Investor (Ken French), and Fama & French, "The Value Premium" (2020) — the "too much volatility to decide" finding and the unresolved debate.
• Investopedia / WallStreetPrep — value investing and margin-of-safety (20–50% discount convention).