Valuation methods: intrinsic, relative and asset-based compared

Every valuation method tries to answer the same question: what is a share of this business worth, given what it is likely to earn and how risky those earnings are? The methods differ in where they look for the answer. Some build the value up from the company's own future cash flows. Some read it off the prices the market pays for similar companies. Some add up what the assets would fetch if sold. Each has a natural home, and each fails in a predictable way.

This guide maps the territory before the deeper pages on Discounted cash flow (DCF): a step-by-step valuation on Apple, Comparable company analysis: how to build and read a comps table and the Dividend discount model (DDM): the Gordon growth formula explained. You will learn the four families of methods, which kinds of business each suits, why careful analysts run more than one, and why no method can hand you "the" value of a stock. The running example is Apple Inc. at the end of its fiscal 2025 (the year ended September 27, 2025), when the stock closed at $255.46.

What "value" means here

Price is what a share trades for. Value is an estimate of what it is worth to an owner. The gap between them is the reason anyone does valuation work. A fair value or intrinsic value estimate is always conditional: "if the business does X and investors require Y, a share is worth Z." Change X or Y and Z moves.

That conditionality is easy to forget once a number is on the screen. A target price from a broker, a DCF output in a spreadsheet and a P/E-based estimate are all opinions dressed as arithmetic. The arithmetic is checkable. The opinion is where the work and the risk lie.

Intrinsic (absolute) valuation

An intrinsic method values the cash a business will hand to its owners over its life, discounted back to today at a rate that reflects risk and the time value of money. The core identity is the present value of future cash flows:

The family has three common members.

• Discounted cash flow (DCF). Forecasts free cash flow for five to ten years, adds a terminal value for everything after, and discounts it all. It is the most general method and works for any business with forecastable cash flow. See Discounted cash flow (DCF): a step-by-step valuation on Apple.
• Dividend discount model (DDM). Discounts the dividends themselves. It fits companies that pay out most of what they earn and grow steadily, and breaks down for low payers. See Dividend discount model (DDM): the Gordon growth formula explained.
• Residual income and economic-profit models. Start from book value and add the present value of future profits above a required return on equity. They are popular for banks and insurers, whose book values are meaningful.

The strength of intrinsic methods is that they force you to state your beliefs about growth, margins, reinvestment and risk. The weakness is that small changes in those beliefs, especially in the discount rate and the long-run growth rate, move the answer a lot. For mature companies most of a DCF's value usually sits in the terminal value, which is the least knowable part.

Relative valuation

A relative method asks what the market pays for similar businesses and applies that price to the company in question. The tools are multiples: price to earnings (P/E), price to sales, price to book, and enterprise-value multiples such as EV/EBITDA and EV/EBIT.

Relative methods are fast, grounded in real prices and easy to communicate. Their blind spot is that they only tell you whether a stock is priced in line with others, not whether the group is priced sensibly. In a bubble, every peer looks reasonable next to every other. They also assume the companies share growth, risk and accounting, which is rarely exact. The Comparable company analysis: how to build and read a comps table page walks through building a comps table properly. A cousin is precedent transactions: multiples paid in past takeovers of similar companies, which include a control premium and so run higher than trading multiples.

Asset-based valuation

Asset-based methods value what the company owns rather than what it earns.

• Book value takes shareholders' equity from the Balance sheet: assets, liabilities and equity explained with Apple. It is a starting point for banks and insurers, whose assets are largely financial and marked close to market.
• Adjusted net asset value restates assets and liabilities at estimated market values: real estate at appraisal, investments at market, hidden liabilities added. It suits holding companies, real estate firms and closed-end funds.
• Liquidation value asks what owners would receive if the assets were sold and the debts paid. It is a floor for distressed companies, and assets sold in a hurry usually fetch less than their book amount.

For most operating companies, especially those built on brands, software and people, asset methods understate value badly. Apple is the extreme case. Its book equity on September 27, 2025 was $73.7 billion, about $4.99 a share, because years of buybacks returned more cash than the company retained. The market valued it at roughly $3.77 trillion, a price-to-book of 51.2. Neither number is "wrong"; book value simply does not measure what Apple's brand, installed base and ecosystem are worth.

Sum-of-the-parts

A sum-of-the-parts (SOTP) valuation values each business line separately, with the method that suits it, then adds them up and subtracts net debt and any corporate costs not assigned to a segment. It is the natural tool for conglomerates, companies with a listed subsidiary, or firms whose divisions have very different economics.

SOTP analyses often reveal a "conglomerate discount": the parts appear worth more than the whole. Sometimes that points to a spinoff that unlocks value. Often it reflects real costs the parts would face alone, or just optimistic multiples applied to each piece.

When each method fits

Two quick rules of thumb help. First, match the method to the cash: value dividends when dividends are the cash owners get, free cash flow when the business could pay out much more than it does, and assets when assets are what an owner would realistically end up with. Second, match the method to the evidence: a company with no profits has no meaningful P/E, and a company with no close peers has no meaningful comps table.

Triangulation: why analysts run more than one

Triangulation means valuing the same company several ways and studying the gaps. The goal is not to average the answers. It is to find out which assumptions are driving each one.

Apple makes the point well. The illustrative DCF above lands near $133 a share. The market price was $255.46. A reverse DCF, which solves for the growth rate that would justify the price, says the market was implicitly assuming about 22.4% a year of free-cash-flow growth for five years at a 9% discount rate, or about 13.4% a year for ten years. And in the comps example, priced at March 31, 2026, Apple's P/E of 34.0 sat above Microsoft's 27.1 and Alphabet's 26.6. Three methods, three framings of the same question: either the market expects much faster growth than the DCF assumed, or it is applying a lower discount rate to Apple's cash flows, or both. That is a far more useful conclusion than any single number.

When methods disagree, ask:

• Is one method a poor fit? A DDM for a company that pays out 14% of earnings, for instance, will always look low.
• Are the inputs consistent? A DCF assuming 3% long-run growth and a peer multiple implying 5% are answering different questions.
• Is the market pricing something your model leaves out? New products, a buyback that shrinks the share count, or a lower perceived risk.

Why no method gives "the" value

Every method rests on forecasts of an uncertain future and on a judgment about required returns. Even the inputs that look like facts are choices: which year's cash flow to start from, whether to count stock-based compensation as a cost, whether to treat long-term marketable securities as cash. Two careful analysts with the same 10-K can reach values tens of percent apart without either making an error.

That is why practitioners talk about ranges and scenarios, and why the idea of a Margin of safety: Benjamin Graham’s rule for valuation errors exists: if your estimate could easily be 25% too high, you want the price to sit well below it before the estimate tells you anything. It is also why markets keep working. Prices move as investors revise exactly the assumptions these methods make explicit.

Common mistakes

• Treating the output as precise. A DCF that prints $132.87 is no more accurate than one that prints "about $130"; the second is more honest.
• Mixing definitions. Using equity-value multiples (P/E) on one company and enterprise-value multiples on another, or adding back cash twice.
• Ignoring the share count. Apple's shares outstanding fell 2.3% in fiscal 2025. A per-share value must use the right, current count.
• Anchoring on the market price. Tweaking inputs until the model matches the price defeats the purpose; a reverse DCF does that openly and is more useful.
• Using one method everywhere. The best method depends on the business, the industry and the question being asked.

How Finabase does it

Finabase does not compute an intrinsic value or run a DCF for you, so the methods on this page are things you do with its data rather than buttons you press. The Valuation tab shows ten fiscal years of a company's own multiples from the data provider, including P/E, P/S, P/B, P/FCF, EV/EBITDA, EV/Sales and FCF yield, alongside margins, returns and growth. That history is the raw material for a relative view of the company against its own past. The Financials tab gives the statement lines you need for a DCF: operating cash flow, capital expenditure, cash, investments and debt. The Peers tab lists up to 12 same-industry companies with their quotes, market caps and year-to-date returns; it is a starting list for comparables, not a side-by-side valuation table. The summary page's key statistics include enterprise value, trailing and forward P/E and the PEG ratio as reported by the data provider.

Frequently asked questions

What is the most accurate valuation method?

None is accurate in general; the best method is the one that fits the business. DCF suits companies with forecastable cash flow, dividend models suit steady high payers, multiples suit companies with close peers, and asset methods suit holding companies and distressed firms.

What is the difference between intrinsic and relative valuation?

Intrinsic valuation estimates value from the company's own future cash flows and a required return; relative valuation estimates it from what the market pays for similar companies. The first can disagree with the whole market; the second can only tell you how a stock compares with its peers.

When should I use a sum-of-the-parts valuation?

Use it when a company runs businesses with clearly different economics, such as a conglomerate or a firm with a listed subsidiary. Each part is valued with its own method or multiple, then the parts are added and net debt and unallocated costs are subtracted.

Why do different methods give different values?

They rest on different assumptions and different evidence. A DCF reflects your growth and discount-rate inputs, a multiple reflects the market's current pricing of peers, and an asset method reflects the balance sheet. The disagreement tells you which assumption to examine.

Summary

Valuation methods fall into intrinsic approaches that discount a company's own cash flows, relative approaches that borrow multiples from similar companies, asset-based approaches that value what the company owns, and sum-of-the-parts approaches that value each segment on its own terms. Each suits particular businesses and fails in known ways: DCF on its terminal value, multiples on the market's mood, asset methods on intangibles. Running more than one and asking why they disagree, as Apple's fiscal 2025 numbers show, produces a range and a set of questions rather than a single figure, and that range is the honest output of the work.