The price-to-earnings ratio is the most cited number in all of investing — and the most misused. Damodaran dedicates substantial attention in The Little Book of Valuation to dissecting the P/E: what fundamentally determines it, why comparing P/Es across different growth rates is meaningless without adjustment, and why the median multiple matters more than the mean in sector comparisons. Understanding P/E at a deep level transforms it from a simple ratio into a window on market expectations.
P/E Ratio — Full Decomposition
From the Gordon Growth Model: P/E = b / (Ke − g) — every P/E has fundamental drivers
Derivation from Gordon Growth Model
Step 1
P₀ = D₁ / (Ke − g)
Stock price = PV of dividends
Step 2
D₁ = E₁ × b
Dividend = EPS × payout ratio
Step 3
P₀ = E₁ × b / (Ke − g)
Substitute into GGM
Step 4
P/E = b / (Ke − g)
Divide both sides by E₁
The Three Fundamental Drivers of P/E
Payout Ratio (b)
D/E — dividends as % of earnings
↑ b → ↑ P/E (more cash returned)
Must hold growth constant — higher payout means less growth unless ROIC is very high
Cost of Equity (Ke)
Rf + β × ERP (CAPM)
↑ Ke → ↓ P/E (higher discount rate reduces all future values)
Rising interest rates → rising Ke → lower justified P/E for all stocks
Growth Rate (g)
Retention rate × ROIC
↑ g → ↑ P/E (future earnings worth more today)
Only value-creating when ROIC > Ke; value-destroying growth lowers P/E
PEG Ratio — P/E Adjusted for Growth
PEG = P/E ÷ EPS Growth Rate (%)
Peter Lynch's rule: PEG < 1.0 is attractive; > 1.5 is expensive; > 2.0 is overpriced
| P/E Multiple | EPS Growth | PEG Ratio | Interpretation |
|---|---|---|---|
| 8× | 4% | 2.00× | Expensive relative to growth |
| 15× | 10% | 1.50× | Moderate — needs quality check |
| 20× | 20% | 1.00× | Fairly valued on growth basis |
| 30× | 25% | 1.20× | Slight premium — monitor |
| 40× | 30% | 1.33× | Growth priced in; risk increases |
Trailing vs. Forward P/E — Which to Use
Trailing P/E (TTM)
Price / Last 12 months actual EPS
Pros
• Based on realized earnings — no estimate error
• Can't be 'managed' by forward guidance
Cons
• Backward-looking; irrelevant for fast-changing businesses
• Distorted by one-time items in prior year
Best for: Mature, stable businesses; historical comparison
Forward P/E (NTM)
Price / Next 12 months consensus EPS estimate
Pros
• Forward-looking; more relevant for investment decisions
• Reflects current business momentum
Cons
• Estimate risk — analysts systematically over-estimate EPS
• Based on consensus that may embed wrong assumptions
Best for: Growth companies; M&A analysis; sector screening
Figure 10.1 — Full P/E decomposition. P/E is not arbitrary — it is determined by three measurable fundamentals. The PEG ratio corrects for growth differences; forward P/E corrects for backward-looking distortions.
The P/E ratio is not an arbitrary market multiple — it has a precise mathematical derivation from the Gordon Growth Model. Starting from P₀ = D₁ / (Ke − g) and defining the dividend payout ratio as b = D₁/EPS₁, we get P₀ = EPS₁ × b / (Ke − g), and therefore P₀/EPS₁ = b / (Ke − g). This derivation reveals exactly what drives P/E — and what doesn't.
P/E Ratio Derivation from Gordon Growth Model
P/E = b / (Ke − g)
b = dividend payout ratio (or earnings retention complement if using EPS-based approach); Ke = cost of equity; g = sustainable earnings growth rate
| Driver | P/E Relationship | Why | Real-World Examples |
|---|---|---|---|
| Growth rate (g) | Positive — higher growth → higher P/E | Higher g shrinks the (Ke−g) denominator, increasing the multiple | High-growth tech: 30–60× P/E; no-growth utilities: 12–18× P/E |
| Cost of equity (Ke) | Negative — higher Ke → lower P/E | Higher Ke grows the denominator; investors require more earnings per dollar of price | Low interest rates (2010–2021) supported high P/Es; rising rates (2022) compressed them |
| Payout ratio (b) | Positive — higher payout → higher P/E, IF ROIC>Ke | Higher payout means more cash returned; valuable if ROIC exceeds cost of capital | Mature companies with high dividends often trade at premium P/Es to growth companies with same earnings |
| Risk (via Ke) | Negative — higher risk → higher Ke → lower P/E | Risk premium in Ke reflects systematic risk; risky earnings are worth less per dollar | Utility (Ke=7%): P/E~25×; biotech (Ke=15%): P/E~10–15× on near-term earnings |
In 2022, the Federal Reserve raised rates by 425 basis points. For a growth stock with g=8% and Ke=10% (ERP=5%, β=1.0), the original P/E denominator was (10%−8%)=2%, giving P/E=50×. After rate hikes: Rf rises 4.25% → Ke rises to ~14.25% → (Ke−g) = (14.25%−8%) = 6.25% → P/E = 1/0.0625 = 16×. The same business, with the same growth rate, went from 50× to 16× earnings. The 68% P/E compression was entirely mechanical — driven by the risk-free rate increasing, not by any change in the underlying business. This is why high-P/E stocks are called 'long duration assets' — they are extremely sensitive to discount rate changes.
Peter Lynch popularized the PEG ratio (P/E divided by earnings growth rate) as a quick heuristic for identifying whether a P/E multiple is justified by growth. The intuition: a company growing earnings at 20% per year 'deserves' a higher P/E than one growing at 5% per year. PEG normalizes the P/E for the growth rate, allowing comparisons across different growth profiles:
PEG Ratio
PEG = (P/E) / (Annual EPS Growth Rate %)
Lynch's rule of thumb: PEG < 1 = potentially undervalued; PEG = 1 = fairly valued; PEG > 2 = potentially overvalued. Growth rate typically the 5-year expected EPS CAGR.
| Company | P/E | 5-yr EPS Growth Rate | PEG | Interpretation |
|---|---|---|---|---|
| Fast-growing SaaS | 45× | 35% | 1.29 | Somewhat elevated but not extreme given growth trajectory |
| Mid-growth tech | 28× | 20% | 1.40 | Slightly above Lynch's 'fair' threshold; reasonable for quality business |
| Consumer staples | 22× | 6% | 3.67 | High PEG — 22× earnings for only 6% growth is expensive; market likely paying for stability premium |
| Cyclical industrial | 12× | 15% | 0.80 | Below 1 — potentially undervalued if growth is sustainable; warrants further analysis |
The PEG ratio has three significant limitations: (1) It treats all growth as equal quality — but 20% growth from a capital-light franchise (Visa) is worth far more than 20% growth from a capital-intensive business (steel company) because the reinvestment rate differs dramatically. (2) It uses near-term growth rates, which are often extrapolated too far — a company growing at 30% today will not grow at 30% forever; the PEG based on next-year's expected growth overstates long-run value for high-growth companies at the peak of their S-curve. (3) It ignores risk — a cyclical company and a stable franchise with the same P/E and growth rate have very different risk profiles. Use PEG as a screening tool, not a valuation tool.
The P/E ratio can be calculated on either trailing earnings (last 12 months of actual reported EPS) or forward earnings (next 12 months of consensus forecast EPS). Each has specific strengths and weaknesses, and the choice matters significantly in certain market conditions:
| Metric | Formula | Advantage | Disadvantage | Best Use Case |
|---|---|---|---|---|
| Trailing P/E (TTM) | Price / Last 12 months EPS | Backward-looking, objective — based on reported GAAP results, not estimates | Reflects past results that may not predict future; distorted by one-time items; misleading at earnings cycle troughs (cyclical stocks look expensive at cycle lows) | Comparing historical valuation levels for the same company; initial screening across sectors |
| Forward P/E (NTM) | Price / Next 12 months consensus EPS | Forward-looking — better for valuation that prices future performance; aligns with how institutional investors think | Subject to analyst estimate errors and bias; consensus estimates typically too optimistic by 10–20%; NTM creates a 'moving window' problem as time passes | Cross-sector comparisons in the current period; M&A and investment decisions; Fed model comparisons to interest rates |
| Normalized P/E | Price / Mid-cycle or normalized EPS | Removes cyclicality — values business at 'average' earnings, not trough or peak | Defining 'normal' is subjective; requires multi-year earnings history | Cyclical companies (automotive, mining, chemicals); companies recovering from one-time disruptions |
Damodaran's specific warning about sector P/E comparisons: never use the sector mean P/E as the benchmark. Use the median. The reason: P/E distributions within sectors are right-skewed — a handful of extremely high P/E companies (often loss-making companies with theoretically infinite or negative P/E) pull the mean far above the median, making the 'average sector P/E' a meaningless and misleading statistic.
Semiconductor sector: 10 companies with P/Es of 18, 22, 25, 27, 30, 33, 38, 45, 120, N/M (loss-making). Arithmetic mean = (18+22+25+27+30+33+38+45+120)/9 = 40×. Median = 30×. A semiconductor company trading at 35× compared to the 40× mean would appear 'cheap.' Compared to the 30× median, it appears 'expensive.' The mean was pulled up by the one outlier at 120× — which is not a representative valuation for a semiconductor business at normal profitability. Using the median, the analyst correctly identifies the 35× company as expensive relative to its most relevant peers.
Key Takeaways
Company A has a P/E of 40× with a 30% EPS growth rate. Company B has a P/E of 20× with a 5% EPS growth rate. Using PEG ratios, which appears more attractively valued?