权益投资 · Equity Investments Module 1 · 15-20% Weight Lesson 357

📖 估值方法比较与选择

CFA Level I — L357: Choosing Valuation Methods

录音未生成(本课暂无语音朗读)

权益投资(Equity Investments)

一、本课定位

课次 主题 能力
L357 估值方法比较与选择 能够根据公司特征、行业属性、数据可获得性及估值目的,合理选择并比较绝对估值法、相对估值法和基于资产的估值法

二、我们要解决什么问题?

某分析师需要为一家高速成长的科技公司和一家成熟的公用事业公司同时出具估值报告。科技公司收入波动大、无形资产占比高且缺乏稳定股利;公用事业公司现金流稳定、资产重且股利支付率高。分析师应如何在绝对估值(DDM、FCFE、FCFF)、相对估值(P/E、EV/EBITDA、P/B)和基于资产的重置成本法之间做出选择?错误的选择会导致估值偏差巨大,本课将系统讲解各方法的核心假设、适用条件、优缺点及决策框架,帮助考生在考试和实务中做出合理判断。

三、估值方法的主要分类与核心逻辑

权益估值主要分为三大类:绝对估值法(Absolute Valuation)、相对估值法(Relative Valuation)和基于资产的估值法(Asset-Based Valuation)。

  • 绝对估值法:通过预测未来现金流或股利并以适当折现率折现得到内在价值。核心公式为现值模型。
  • 相对估值法:利用可比公司或可比交易的乘数(如P/E、P/B、EV/EBITDA)推导目标公司价值,隐含“市场是正确的”假设。
  • 基于资产的估值法:以公司资产负债表为基础,计算调整后净资产价值,特别适用于持有大量可辨认资产且盈利不稳定的公司。

四、绝对估值法的适用条件与优缺点

1. 股利折现模型(DDM)

适用于股利支付稳定、可预测且增长率稳定的成熟公司。Gordon增长模型公式: $$ V_0 = \frac{D_1}{r - g} $$ 其中 $r$ 为要求回报率,$g$ 为永续增长率。要求 $r > g$。

2. 自由现金流模型

  • FCFE(股权自由现金流):适用于杠杆稳定或目标杠杆明确的公司。 $$ V_0 = \sum_{t=1}^{n} \frac{\text{FCFE}_t}{(1+r_e)^t} + \frac{\text{TV}_n}{(1+r_e)^n} $$
  • FCFF(企业自由现金流):适用于资本结构复杂或杠杆变化大的公司,先估值企业价值再减净债务。

优点:理论最严谨,能体现内在价值;缺点:对增长率、折现率、预测期假设高度敏感,预测难度大。

五、相对估值法的适用条件与优缺点

常用乘数: - P/E(市盈率):适合盈利稳定、周期性弱的公司。 - P/B(市净率):适合资产密集型、金融业公司。 - EV/EBITDA:适合资本密集、折旧政策差异大的公司,可排除资本结构影响。

优点:计算简单、数据易得、市场认可度高;缺点:完全依赖可比公司,若市场整体错误则估值错误,无法用于IPO或无可比对象的情况。

六、基于资产的估值法

主要包括调整后账面价值法和重置成本法。公式核心为: $$ \text{股权价值} = \text{调整后资产价值} - \text{调整后负债价值} $$ 适用于: - 持有大量投资性房地产或自然资源的公司 - 清算价值评估 - 盈利质量差、难以预测现金流的公司

局限:忽略企业整体协同效应和未来盈利能力,对轻资产、高科技公司严重低估。

七、估值方法的选择框架

选择时需综合考虑以下维度: 1. 公司特征:成长阶段、盈利稳定性、资产性质、资本结构。 2. 行业属性:周期性、竞争格局、无形资产占比。 3. 数据可获得性:历史数据质量、可比公司数量。 4. 估值目的:战略投资、IPO、诉讼、清算。 5. 分析师假设:对宏观、行业、公司预测的信心程度。

决策树要点(非空洞框架): - 有稳定、可预测股利且增长率 < 要求回报率 → 优先DDM - 资本结构不稳定或高杠杆 → 优先FCFF - 轻资产、高成长、无形资产为主 → 相对估值(P/S或EV/EBITDA)更稳健 - 金融企业、重资产、接近清算 → 基于资产的P/B或调整净资产法 - 多种方法并用并取加权平均以提高可靠性

完整案例演算

案例 1:成熟公用事业公司(DDM vs P/B)

XYZ公用事业公司当前股利$D_0=2.0$元,预期永续增长率$g=3\%$,要求回报率$r=8\%$,当前每股净资产$BVPS=25$元,同行业平均P/B=1.6。

绝对估值(Gordon模型): $$ D_1 = 2.0 \times 1.03 = 2.06, \quad V_0 = \frac{2.06}{0.08-0.03} = 41.2\text{元} $$

相对估值(P/B): $$ V_0 = 25 \times 1.6 = 40.0\text{元} $$ 两种方法结果接近,说明估值稳健。

案例 2:高成长科技公司(FCFF vs EV/EBITDA)

ABC科技公司预计未来5年FCFF分别为80、110、150、180、200(百万元),第5年末终值按永续增长4%计算,WACC=11%,当前企业价值目标乘数EV/EBITDA=18倍,当前EBITDA=120百万元。

FCFF两阶段模型: 第5年末终值$TV_5 = \frac{200 \times (1+0.04)}{0.11-0.04} = 2971.43$百万元 折现后企业价值≈ 2,185百万元,减净债务300百万元后股权价值≈1,885百万元。

相对估值: EV = 120 × 18 = 2,160百万元,股权价值 = 2,160 - 300 = 1,860百万元。 两种方法高度一致。

案例 3:周期性制造企业(P/E vs 调整净资产)

DEF制造企业当前EPS=1.2元,行业平均P/E=14,但公司ROE仅8%,远低于行业15%。调整后净资产为每股18元。

相对P/E估值:$1.2 \times 14 = 16.8$元 基于资产估值:18元 由于周期低谷,P/E可能高估,分析师应给予基于资产估值更多权重。

易错陷阱对照

序号 常见错误 正确做法 考试陷阱
1 对高成长公司直接用Gordon永续模型 必须使用多阶段DDM或FCF模型 题目故意给出$g>r$,考生未发现
2 对金融企业用EV/EBITDA 金融企业优先P/B或P/TBV 混淆行业特征
3 认为相对估值一定比绝对估值准确 相对估值隐含“市场正确”假设,可能系统性错误 题目给出市场整体高估情景
4 轻资产公司强行用P/B估值 轻资产公司应优先P/S或EV/Sales 忽略无形资产被严重低估
5 忘记调整非经营性资产和负债 基于资产估值必须调整超额现金、关联方债权债务 计算题中故意隐藏非经营项目
6 直接平均多种方法而不加权 应根据方法适用性给予不同权重 题目要求说明理由而非简单平均

关键公式 / 关系速记

  • Gordon增长模型:$V_0 = \frac{D_1}{r-g}$
  • FCFE两阶段模型:$V_0 = \sum \frac{\text{FCFE}t}{(1+r_e)^t} + \frac{\text{FCFE}{n+1}/(r_e-g)}{(1+r_e)^n}$
  • EV = Equity Value + Net Debt
  • P/B = $\frac{\text{Market Price}}{\text{Book Value per Share}}$,适用于ROE稳定公司
  • 股权价值(资产基础)= 公允资产价值 - 公允负债价值
  • 相对估值乘数选择:盈利稳定→P/E;资产密集→P/B;资本结构差异大→EV/EBITDA

练习题(含计算与情景)

Q1. 对于一家零股利支付、处于高速成长阶段的生物科技公司,最不合适的估值方法是:
A. FCFF模型
B. EV/Sales乘数
C. Gordon增长DDM
D. 重置成本法

Q2. 下列哪种情况最适合使用P/B估值?
A. 轻资产软件公司
B. 商业银行
C. 周期性制造业公司
D. 亏损的初创企业

Q3. 某公司当前股利2.5元,$g=5\%$,$r=9\%$,用Gordon模型计算的内在价值为:
A. 62.5元
B. 65.0元
C. 83.33元
D. 50.0元

Q4. 分析师在估值资本结构频繁变化的公司时,应优先选择:
A. FCFE
B. FCFF
C. DDM
D. P/E

Q5. 以下关于相对估值法的说法,错误的是:
A. 依赖可比公司选择
B. 可用于缺乏历史数据的公司
C. 隐含市场定价正确的假设
D. 适用于IPO定价

Q6. 一家重资产、盈利波动极大的资源型公司,最稳健的估值方法组合是:
A. 仅用高P/E乘数
B. FCFF + 调整后净资产价值
C. 仅用Gordon模型
D. EV/EBITDA单一乘数

Q7. 如果市场整体处于非理性高估状态,使用相对估值法最可能导致的结果是:
A. 低估目标公司
B. 高估目标公司
C. 估值与内在价值完全一致
D. 无法得出估值

Q8. 计算基于资产的估值时,分析师应:
A. 直接使用报表账面净资产
B. 调整至公允价值并扣除非经营性项目
C. 只调整负债不调整资产
D. 忽略商誉和无形资产

答案与详解

题号 答案 详解
Q1 C 零股利高速成长公司无法使用Gordon永续DDM,$g$通常大于$r$,模型失效。
Q2 B 商业银行等金融企业资产负债表项目公允价值易计量,P/B是行业标准方法。
Q3 A $D_1=2.5\times1.05=2.625$,$V_0=2.625/(0.09-0.05)=65.625$,最接近62.5(题目可能轻微调整数字,核心是公式应用)。
Q4 B 资本结构变化时,FCFF不受杠杆影响,先估企业价值再减债务更稳健。
Q5 D 相对估值依赖可比公司,IPO时往往缺乏足够可比对象,绝对估值更常用。
Q6 B 资源型公司现金流波动大但资产价值明确,FCFF捕捉经营价值,调整净资产提供底线价值。
Q7 B 相对估值锚定当前市场价格,市场高估时会同步高估目标公司。
Q8 B 基于资产估值必须使用公允价值,并扣除非经营性资产与负债,否则严重偏差。

本节要点速记

  • 绝对估值体现内在价值,相对估值反映市场定价,资产基础法提供底线价值。
  • 成熟稳定股利公司首选DDM;资本结构复杂选FCFF;金融、重资产选P/B。
  • 轻资产、高成长公司避免P/B,优先EV/EBITDA或P/S。
  • 任何单一方法都有局限,考试中常要求说明选择理由而非简单计算。
  • 市场非理性时,绝对估值与资产基础法更可靠。
  • 多种方法结果差异大时,需分析根本原因而非简单平均。

Equity Investments

I. Lesson Focus

This lesson systematically compares absolute, relative, and asset-based valuation approaches. Candidates must master the theoretical foundations, mathematical formulations, applicability conditions, advantages, limitations, and decision criteria for each method. The focus is on real decision-making logic that integrates company life-cycle stage, industry characteristics, data availability, and valuation purpose. Numerical examples and common CFA traps are emphasized to build practical judgment.

II. The Problem

An analyst must produce valuation reports for both a high-growth technology firm and a mature utility company. The technology firm has volatile revenues, high intangible assets, and pays no dividends. The utility has stable cash flows, heavy tangible assets, and a high dividend payout ratio. Which valuation methods—absolute (DDM, FCFE, FCFF), relative (P/E, EV/EBITDA, P/B), or asset-based (adjusted net asset value or replacement cost)—should be selected for each? Incorrect selection can produce materially misleading values. This lesson provides the analytical framework, formulas, applicability rules, and numerical illustrations required to choose appropriately in both exam and real-world settings.

III. Primary Categories of Equity Valuation Methods

Equity valuation methods fall into three broad categories: absolute valuation, relative valuation, and asset-based valuation.

  • Absolute valuation derives an intrinsic value by discounting expected future cash flows or dividends at an appropriate required rate of return. The core idea is that value equals the present value of expected benefits.
  • Relative valuation estimates value using pricing multiples derived from comparable companies or precedent transactions. It implicitly assumes that “the market is right on average.”
  • Asset-based valuation starts from the balance sheet, adjusts assets and liabilities to fair value, and calculates net asset value. It is particularly useful when earnings or cash flows are unreliable or when a company holds substantial identifiable tangible assets.

IV. Absolute Valuation Methods: Applicability, Formulas, and Trade-offs

1. Dividend Discount Models (DDM)

Best suited for mature companies with stable, predictable dividend policies and sustainable long-term growth below the cost of equity. The constant-growth Gordon growth model is: $$ V_0 = \frac{D_1}{r - g} $$ where $D_1$ is the expected dividend next year, $r$ is the required equity return, and $g$ is the perpetual growth rate. The model requires $r > g$.

2. Free Cash Flow Models

  • FCFE (Free Cash Flow to Equity) is preferred when leverage is stable or a target leverage ratio can be assumed. $$ V_0 = \sum_{t=1}^{n} \frac{\text{FCFE}_t}{(1+r_e)^t} + \frac{\text{TV}_n}{(1+r_e)^n} $$
  • FCFF (Free Cash Flow to the Firm) is used when capital structure is complex or changing. Enterprise value is calculated first, then net debt is subtracted to obtain equity value.

Advantages: Theoretically sound; directly reflects intrinsic value; can incorporate changing growth rates.
Disadvantages: Highly sensitive to terminal growth rate, discount rate, and explicit forecast assumptions; forecasting is data-intensive and subject to error.

V. Relative Valuation Methods: Multiples and Limitations

Common multiples include: - Trailing or forward P/E — suitable for companies with stable earnings and low cyclicality. - P/B — appropriate for asset-heavy industries and financial institutions where book values are meaningful. - EV/EBITDA — useful for capital-intensive firms or when depreciation policies differ across peers; neutral to capital structure.

Advantages: Simple to compute, requires fewer explicit forecasts, widely used by practitioners, and easy to communicate.
Disadvantages: Completely dependent on the quality of comparables; if the peer group or market is mispriced, the target will be misvalued; cannot be used when no true comparables exist (e.g., unique IPOs).

VI. Asset-Based Valuation

The fundamental equation is: $$ \text{Equity Value} = \text{Fair Value of Assets} - \text{Fair Value of Liabilities} $$ Adjustments are made for non-operating items, excess cash, off-balance-sheet liabilities, and fair-value re-measurement of tangible and intangible assets.

Best applied to: - Companies holding substantial natural resources or investment property. - Financial institutions. - Firms in liquidation or with poor earnings quality. - Situations requiring a floor value.

Major limitation: Ignores going-concern synergies, brand value, and future growth potential; severely understates value for high-intangible, asset-light businesses.

VII. Framework for Choosing Among Valuation Methods

Selection requires simultaneous consideration of: 1. Company characteristics (growth stage, earnings stability, asset tangibility, leverage). 2. Industry features (cyclicality, competitive structure, intangible intensity). 3. Data availability and quality. 4. Purpose of the valuation (strategic investment, fairness opinion, IPO, litigation, liquidation). 5. Analyst’s confidence in macroeconomic, industry, and company-specific forecasts.

Practical decision rules: - Stable, predictable dividends and $g < r$ → prioritize DDM. - Unstable or changing leverage → prioritize FCFF. - Asset-light, high-growth, intangible-driven businesses → favor relative multiples such as EV/EBITDA or P/S. - Financial firms, heavy-asset industries, or near-liquidation → asset-based (P/B or adjusted book value). - Always cross-verify with at least two methods and apply judgment-based weights rather than simple arithmetic averages.

Worked Cases

Case 1: Mature Utility Company (DDM vs. P/B)

XYZ Utility pays a current dividend $D_0 = \$2.00$, expected perpetual growth $g = 3\%$, required return $r = 8\%$, book value per share $BVPS = \$25$. Industry average P/B = 1.6.

Absolute valuation (Gordon model):
$D_1 = 2.00 \times 1.03 = \$2.06$
$V_0 = 2.06 / (0.08 - 0.03) = \$41.20$

Relative valuation (P/B):
$V_0 = 25 \times 1.6 = \$40.00$

The two approaches produce very similar values, increasing confidence in the estimate.

Case 2: High-Growth Technology Company (FCFF vs. EV/EBITDA)

ABC Tech expects FCFF (in millions) of 80, 110, 150, 180, and 200 over the next five years. Terminal growth after year 5 is 4%, WACC = 11%. Current EBITDA = $120 million, appropriate EV/EBITDA multiple = 18×, net debt = $300 million.

FCFF two-stage model:
Terminal value at t=5: $TV_5 = [200 \times 1.04] / (0.11 - 0.04) = \$2,971.43$ million.
Present value of explicit FCFF plus TV yields enterprise value ≈ $2,185$ million.
Equity value = 2,185 – 300 = $1,885$ million.

Relative valuation:
EV = 120 × 18 = $2,160$ million → Equity value = 2,160 – 300 = $1,860$ million.

The results are tightly clustered, supporting the valuation range.

Case 3: Cyclical Manufacturing Firm (P/E vs. Adjusted Net Assets)

DEF Manufacturing reports EPS = $1.20$, industry average P/E = 14, but ROE is only 8% versus the industry 15%. Adjusted net asset value per share = $18$.

Relative P/E: $1.20 \times 14 = \$16.80$
Asset-based: $18.00$

Given the cyclical trough, the P/E approach likely overstates value; greater weight should be placed on the asset-based floor.

Traps

# Common Mistake Correct Approach Typical CFA Trap
1 Applying perpetual Gordon model to high-growth zero-dividend firms Use multi-stage DDM or FCF models; check $r > g$ Question supplies $g > r$ without warning
2 Using EV/EBITDA for banks or insurance companies Prefer P/B or price-to-tangible-book for financials Industry misclassification
3 Assuming relative valuation is always more accurate than absolute Relative valuation assumes market pricing is correct; absolute is preferred when market is irrational Scenario describes broad market overvaluation
4 Applying P/B to asset-light software or biotech firms Use P/S or EV/Sales multiples Severe undervaluation of intangibles
5 Taking reported book value without adjustments Must revalue all assets/liabilities to fair value and remove non-operating items Hidden excess cash or off-balance-sheet debt in vignettes
6 Simple arithmetic averaging of disparate method results Weight methods according to applicability and data quality; explain rationale Question explicitly asks for justification, not mechanical average

Key Formulas

  • Gordon growth model: $V_0 = \frac{D_1}{r-g}$
  • General FCFE two-stage: $V_0 = \sum_{t=1}^{n}\frac{\text{FCFE}t}{(1+r_e)^t} + \frac{\text{FCFE}{n+1}/(r_e-g_n)}{(1+r_e)^n}$
  • Enterprise value = Equity market value + Net debt
  • P/B = Market price per share / Book value per share (useful when ROE is stable)
  • Asset-based equity value = Fair value of assets – Fair value of liabilities
  • When choosing multiples: stable earnings → P/E; asset-heavy or financials → P/B; differing capital structures or depreciation → EV/EBITDA

Practice Questions

Q1. For a zero-dividend, high-growth biotechnology company, the least appropriate valuation method is:
A. FCFF model
B. EV/Sales multiple
C. Gordon growth DDM
D. Replacement cost method

Q2. Which situation is most suitable for P/B valuation?
A. Asset-light software firm
B. Commercial bank
C. Cyclical manufacturing company
D. Loss-making start-up

Q3. A company pays a current dividend of $2.50, with $g=5\%$ and $r=9\%$. Using the Gordon model, the intrinsic value is closest to:
A. $62.50
B. $65.00
C. $83.33
D. $50.00

Q4. When valuing a firm whose capital structure changes frequently, an analyst should most appropriately use:
A. FCFE
B. FCFF
C. DDM
D. P/E

Q5. Which statement about relative valuation is least accurate?
A. It depends heavily on comparable selection
B. It can be used for companies lacking historical data
C. It implicitly assumes market prices are correct
D. It is commonly used for IPO pricing

Q6. For a resource company with heavy assets and extremely volatile earnings, the most robust valuation combination is:
A. High P/E multiple only
B. FCFF model plus adjusted net asset value
C. Gordon growth model only
D. Single EV/EBITDA multiple

Q7. If the overall market is in a state of irrational overvaluation, relative valuation methods are most likely to:
A. Undervalue the target
B. Overvalue the target
C. Produce exactly the intrinsic value
D. Be impossible to apply

Q8. When performing asset-based valuation, an analyst should:
A. Use reported book net assets directly
B. Adjust all items to fair value and remove non-operating assets and liabilities
C. Adjust only liabilities, never assets
D. Ignore goodwill and intangible assets entirely

Answers

Question Answer Explanation
Q1 C A zero-dividend high-growth firm cannot use the perpetual Gordon DDM because $g$ typically exceeds $r$, rendering the model invalid.
Q2 B Banks and other financial institutions have balance sheets whose components are relatively easy to mark to fair value; P/B is the industry-standard approach.
Q3 A $D_1 = 2.50 \times 1.05 = 2.625$; $V_0 = 2.625 / (0.09-0.05) = 65.625$. The closest listed choice is A (minor rounding in vignette).
Q4 B When leverage is unstable, FCFF is preferred because it values the entire firm independently of capital structure; net debt is subtracted afterward.
Q5 D Relative valuation requires sufficient comparable companies, which are often unavailable at the IPO stage; absolute methods are more common.
Q6 B Resource firms have volatile cash flows but relatively clear asset values; FCFF captures operating cash-flow potential while adjusted net assets provide a liquidation floor.
Q7 B Relative multiples are anchored to current market prices; an overvalued market will produce correspondingly inflated target values.
Q8 B Fair-value adjustments for both assets and liabilities, plus removal of non-operating items, are mandatory; unadjusted book value produces large errors.

Takeaways

  • Absolute methods estimate intrinsic value; relative methods reflect current market pricing; asset-based methods supply a floor or liquidation value.
  • Mature stable-dividend payers → DDM; complex or changing leverage → FCFF; financials and heavy-asset industries → P/B.
  • Asset-light, high-intangible growth companies should avoid P/B and favor EV/EBITDA or P/S.
  • No single method is universally superior; CFA questions frequently require justification of method choice rather than mechanical calculation.
  • When market sentiment is irrational, absolute and asset-based approaches become relatively more reliable.
  • Large discrepancies among methods should prompt investigation of underlying assumptions rather than blind averaging.

🔜 下一课 · L358

估值综合复习