Skip to content

Stock Valuation Methods: A Practical Overview

By Alexander Chepakovich · Updated

The map of methods

Every valuation method answers the same question — what is this business worth? — from a different pile of evidence. Broadly there are four families: discounted cash flow, which values the future cash stream; earnings formulas, which capitalize profits with Graham-style shorthand; asset-basedapproaches, which value what the company owns; and relative valuation, which prices it off what the market pays for similar businesses. No family is universally best — each fails in characteristic ways.

Graham and Dodd's Security Analysis established the intellectual frame: value independently, from evidence, then compare with price. Every method below is a different way of doing that.

Discounted cash flow

DCF is the most fundamental approach: project the cash the business will return, discount it at a required rate of return, add a terminal value. It is the backbone of most serious valuation — including the InvestorsCraft intrinsic value — because it prices the thing itself rather than the market's mood. Its weakness is sensitivity: discount rate and growth assumptions move the answer enormously, so honest DCF work is a range with a sensitivity grid, not a point.

Conventional DCF also assumes today's margins project forward — the failure mode for growth and loss-making companies covered in the growth-company guide.

Earnings formulas

The Graham Formula (V = EPS × (8.5 + 2g)) and its descendants compress a valuation into earnings and a growth estimate — fast, transparent, and easy to sanity-check. The Graham-Dodd tradition adds the balance sheet, weighing normalized earnings power and asset coverage; the Graham Number (see the Graham Number guide) caps price at 15 times earnings and 1.5 times book simultaneously.

These formulas shine on stable, profitable businesses where the inputs are meaningful — and break exactly where the inputs stop meaning anything: losses, cyclical peaks, unsustainable growth assumptions.

Asset-based approaches

Asset-based valuation asks what the company's net assets are worth — book value, adjusted book, or in the strictest Graham form, net current assets minus all liabilities. It is the most conservative method: it values the evidence on the balance sheet and gives no credit for future earning power.

That makes it a floor rather than a value for a going concern — indispensable for financials, asset-heavy industrials and liquidation analysis; nearly useless for asset-light businesses whose worth lives in intangibles the balance sheet does not record.

Relative valuation

Multiples — P/E, EV/EBITDA, price-to-book, price-to-sales — price a stock off what the market pays for comparable businesses or off its own history. Fast, comparable, and the market's native language. The catch is philosophical: relative valuation contains no intrinsic anchor. If the whole sector is expensive, "cheap versus peers" can still mean expensive — a warning Security Analysis issued decades before multiples became standard practice.

Used honestly, multiples are corroboration: when a stock screens cheap on its own P/E range and cheap against a DCF, two independent measures point the same way.

How to choose

Match the method to the business. A stable, profitable company with projectable cash flows earns a DCF and a Graham check; an asset-heavy or financial business warrants asset-based weight; a growth or loss-making company needs a model that derives margins from cost structure rather than assuming them — the fixed/variable approach described on the model page.

Whatever you use, run at least two genuinely different methods. Agreement between a DCF and a deep-value estimate is meaningful evidence; agreement between two variations of the same calculation is not.

Four methods on every stock page

InvestorsCraft publishes four independent valuations per covered stock: the intrinsic value (DCF-based, on the Chepakovich model), an AI fair value, a Graham-Dodd deep value, and a Graham Formula value. Keeping them separate is deliberate — the disagreement between methods is itself information about how uncertain a valuation is.

Every stock page compares price with the estimates and shows free signal bands; exact values are part of Pro. The stock screener combines all four into a composite signal — the median of whichever methods publish a value — so the ranking only exists where at least two independent estimates do.

Try it

See it on real stocks

Frequently asked questions

Which valuation method should I use?

Match the method to the business: DCF for going concerns with projectable cash, asset-based for asset-heavy or financial companies, earnings formulas for stable profitable firms, multiples only as corroboration. Comparing several estimates is better than perfecting one.

Why do different methods give different values?

Each method weighs different evidence — cash flow, earnings, assets, or market prices — and encodes different assumptions. The spread between methods is honest information about how uncertain the valuation is.

Is a higher valuation method count better?

Independent methods matter more than the count. Four estimates built on the same assumptions agree trivially; two built on genuinely different evidence are a real cross-check.

This guide is educational content, not investment advice. Worked examples use hypothetical numbers unless labelled otherwise.