Skip to content

CFA Level II Exam · Equity Valuation: Applications and Processes

How to Select a Valuation Model and Convert Forecasts

Updated 7 October 2026 · Fact-checked

Choosing a valuation model means matching the method to the company, the data available and the purpose of the valuation. Absolute models such as DCF estimate intrinsic value. Relative models such as multiples compare value with peers. You then convert forecasts into a value, compare it with price, and recommend buy, hold or sell.

Understand Selecting the Valuation Model and Converting Forecasts

A valuation model is a tool that turns forecasts of a company's future into an estimate of what its shares are worth. No single model fits every company. Your job in an item set is to pick the model the vignette supports and then apply it with the data given.

Absolute valuation models estimate intrinsic value from the company's own cash flows and risk. Examples are the dividend discount model (DDM), free cash flow to the firm (FCFF), free cash flow to equity (FCFE) and residual income. Relative valuation models estimate value by comparing multiples, such as P/E, P/B, P/S and EV/EBITDA, with those of similar companies or with the company's own history.

Three questions guide model choice. First, does the model suit the company? A stable, mature dividend payer suits a DDM. A company with no dividends but positive, forecastable cash flow suits FCFF or FCFE. A company with negative free cash flow but reliable book value and clean accounting may suit residual income. Second, are the inputs available and reliable? A model that needs a growth rate you cannot estimate is a poor choice. Third, what is the purpose? A control purchase may call for FCFF because the buyer can change the dividend and capital structure. A minority holder cares more about dividends.

Converting forecasts to value has a fixed flow. Forecast the drivers (revenue, margins, investment, financing). Build the cash flows or earnings. Choose a discount rate or a comparable multiple. Compute value. Test it with sensitivity analysis. Compare the value per share with market price. If value is above price by enough to cover estimation error and costs, the stock looks undervalued. The recommendation must follow from that gap, not from the model's complexity.

The exam also checks judgement. Using several models and reconciling them is often better than relying on one. When models disagree, find the input that drives the difference, such as the growth rate or terminal value.

Key formulas to remember

Intrinsic value versus price
Value gap = Intrinsic value − Market price
Positive gap suggests undervalued; negative suggests overvalued. Act only if the gap is large relative to estimation error.
Justified multiple from a model
Justified leading P/E = (D₁ ÷ E₁) ÷ (r − g); Justified trailing P/E = (D₀ ÷ E₀) × (1 + g) ÷ (r − g)
Leading P/E uses next year's payout, D₁/E₁, and trailing P/E uses the current payout, D₀/E₀. Valid when the Gordon growth model applies and r > g. Shows what drives a multiple: growth, risk, payout.
Gordon growth value
V₀ = D₁ ÷ (r − g)
Use for stable, mature companies with steady dividend growth and r > g.
Value from a multiple
Value per share = Benchmark multiple × Company's metric per share
The benchmark should come from truly comparable firms or the company's own history.
Enterprise value to equity value
Equity value = EV − Debt + Cash (and other non-operating assets), less preferred stock and non-controlling interest
Needed after FCFF or EV multiples, before dividing by shares.
Weighted value from several models
Final value = Σ (weight × model value), weights sum to 1
Weights reflect how reliable each model's inputs are.

How to solve Selecting the Valuation Model and Converting Forecasts questions

Use this order on any model-selection or forecast-to-recommendation question in an item set.

  1. 1Read the vignette for the company profile: dividend history, cash flow sign, leverage, earnings stability, growth stage and whether accounting is reliable.
  2. 2Note the purpose: minority stake, control transaction, or relative ranking. This changes which cash flow or multiple is suitable.
  3. 3Match the model to the facts. Stable dividends point to DDM. Positive forecastable free cash flow points to FCFF or FCFE. Weak cash flow with reliable book value points to residual income. A need for peer comparison points to multiples.
  4. 4Check input quality. Reject a model if a key input, such as growth or beta, is unstable or cannot be estimated.
  5. 5Turn forecasts into value: use the stated cash flows, discount rate and growth, or apply the benchmark multiple to the right metric. Check units and the per-share step.
  6. 6Adjust for claims: subtract debt and other claims if you valued the whole firm, and divide by shares outstanding.
  7. 7Compare value with price and apply the decision rule. Consider sensitivity and any stated margin of safety.
  8. 8Choose the answer that gives the recommendation and a reason tied to the vignette.

Quickest way: Three-check model screen

When to use it: Use when the question asks which model is most appropriate and the vignette is long.

  1. Check 1, dividends: stable and policy-driven? If yes, DDM is the candidate.
  2. Check 2, free cash flow: positive and forecastable? If yes, FCFF for capital structure changes or whole-firm value, FCFE for stable leverage.
  3. Check 3, accounting and comparables: no useful cash flow but clean book value points to residual income; need for market context points to multiples.
  4. Eliminate options that depend on an input the vignette says is unreliable.
  5. Pick the remaining option and check the direction of the recommendation.

Common mistakes in Selecting the Valuation Model and Converting Forecasts

  • Choosing DDM for a company that pays no dividend or pays an unstable one.

    DDM is the first model students learn, so they default to it.

    Fix: Check whether dividends are stable and tied to earnings. If not, move to FCFE, FCFF or residual income.

  • Treating relative valuation as if it gives intrinsic value.

    A multiple gives a number per share, so it looks like a DCF answer.

    Fix: Remember that a multiple only shows value relative to peers. If the peer group is mispriced, the result is too.

  • Forgetting to subtract debt after an FCFF or EV-based valuation.

    Students stop at the present value and divide by shares.

    Fix: Go from firm value to equity value by deducting debt, preferred stock and non-controlling interest, and adding non-operating cash.

  • Comparing value with price and recommending without considering estimate error.

    The rule 'value above price means buy' is memorised without nuance.

    Fix: Check for sensitivity or a margin of safety in the vignette. A small gap may not justify action.

  • Using comparables that are not truly similar in growth, risk or accounting.

    Same industry label is assumed to mean comparable.

    Fix: Compare growth, leverage, margins and accounting policies. Adjust or reject poor peers.

  • Applying a Gordon growth model when r is not above g.

    Students plug in numbers without checking the condition.

    Fix: Confirm r > g first. If not, the single-stage model is invalid and a multistage model is needed.

Worked examples

Example 1

Vignette: Arvind Textiles, a mature manufacturer, pays a steady dividend and has grown it at about 4% a year. The next dividend, D₁, is expected to be 6.00 per share. The analyst's required return is 10%. The share trades at 90.00. Q1: Which model is most appropriate? Q2: What is the intrinsic value? Q3: What is the recommendation?

Show the solution
  1. Q1: The company is mature, pays stable dividends and has steady growth. A dividend discount model, specifically the Gordon growth model, fits well.
  2. Q2: Check r > g: 10% > 4%, so the model is valid.
  3. V₀ = D₁ ÷ (r − g) = 6.00 ÷ (0.10 − 0.04) = 6.00 ÷ 0.06 = 100.00.
  4. Q3: Compare value with price: 100.00 − 90.00 = 10.00, so the share is undervalued by about 11% of the market price (10 ÷ 90).
  5. The gap is positive and sizeable, so the recommendation is to buy, subject to checking sensitivity to g and r.

Answer: Q1: Gordon growth DDM. Q2: 100.00 per share. Q3: Buy, because intrinsic value exceeds price.

Example 2

Vignette: Nova Software pays no dividend and reinvests heavily. Forecast FCFF next year is 400 million and is expected to grow at 5% indefinitely. WACC is 9%. Debt is 1,200 million, cash is 200 million, and there are 100 million shares. The price is 85.00. Q1: Why is FCFF suitable? Q2: What is the value per share? Q3: What is the recommendation?

Show the solution
  1. Q1: There is no dividend, so DDM is unsuitable. FCFF is positive and forecastable, and it values the whole firm before financing effects, so it is suitable.
  2. Q2: Firm value = FCFF₁ ÷ (WACC − g) = 400 ÷ (0.09 − 0.05) = 400 ÷ 0.04 = 10,000 million.
  3. Equity value = 10,000 − 1,200 + 200 = 9,000 million.
  4. Value per share = 9,000 ÷ 100 = 90.00.
  5. Q3: Value 90.00 exceeds price 85.00 by 5.00, about 5.9% of price (5 ÷ 85).
  6. The stock appears modestly undervalued. Because the gap is small and a terminal-style growth input drives the value, the analyst should test sensitivity before a strong buy.

Answer: Q1: Positive, forecastable FCFF and no dividend. Q2: 90.00 per share. Q3: Modestly undervalued; support a buy only if the value remains above price after sensitivity testing of g and WACC.

Exam tips

  • Read the vignette for signals: no dividends, negative free cash flow, high leverage, or unreliable accounting. Each signal rules a model in or out.
  • Always finish the chain: value, then compare with price, then state the recommendation. Many wrong answers stop one step early.
  • In FCFF questions, look for debt, cash, preferred stock and non-controlling interest before dividing by shares.
  • If two models give different values, look for the input difference. The answer often names growth, discount rate or peer quality.
  • There is no penalty for wrong answers, so never leave an item blank.

Selecting the Valuation Model and Converting Forecasts: frequently asked questions

What is the difference between absolute and relative valuation?

Absolute valuation estimates intrinsic value from a company's own forecast cash flows or earnings and its required return. Relative valuation estimates value by comparing multiples such as P/E or EV/EBITDA with peers. Absolute models depend on your inputs. Relative models depend on how fairly the peers are priced.

When should I use FCFF instead of FCFE?

FCFF suits whole-firm valuation, firms with changing capital structure, and firms with negative FCFE but positive FCFF. FCFE suits firms with stable leverage where the equity cash flow is clear. After FCFF you must subtract debt and other claims to reach equity value.

Which model is best for a company with no dividends?

A dividend model is a poor fit. Use FCFF or FCFE if free cash flow is positive and forecastable. If cash flows are weak but book value and accounting are reliable, residual income can work. Multiples are also an option for a quick cross-check.

How do I turn a valuation into a recommendation?

Compare intrinsic value per share with market price. If value is clearly higher, the stock looks undervalued and a buy view is supported. If it is clearly lower, it looks overvalued. Allow for estimation error and sensitivity before acting.