The Flaws of Relative Valuation (Multiples)

Wall Street defaults to relative valuation because it's fast and defensible. "Stock A trades at 12x earnings; peers trade at 15x; it's undervalued." But this reasoning is circular. If all peers are overvalued, the fact that your stock is cheaper doesn't mean it's cheap. Relative valuation is a useful cross-check—a sanity test—but it's not a path to intrinsic value. A stock can be undervalued relative to peers while still overpriced relative to its own earnings power.

Quick definition: Relative valuation compares a company's multiples (P/E, EV/EBITDA, Price-to-Book) to those of comparable companies, assuming similar valuations for similar businesses. The critical flaw: it assumes the market prices the peer group correctly.

Key Takeaways

• Relative valuation assumes the peer group is correctly priced, which is often false
• Multiples are context-dependent: interest rates, sentiment, and growth expectations vary
• Comparing multiples across different business models (organic vs. acquired growth) is misleading
• Capital structure distortions (debt, preferred stock, options) make multiples incomparable
• Earnings quality and accounting differences make P/E ratios untrustworthy
• Absolute valuation (DCF, EPV) is more honest than relative valuation

Why Relative Valuation is Flawed

The Peer Group Assumption

Relative valuation assumes that if peers trade at 15x earnings, a 12x multiple is a bargain. But what if the entire industry is overvalued? In 2007, all banks traded at 12–15x earnings, suggesting they were fairly valued. In 2008, banks collapsed, revealing that the entire peer group was overvalued at any multiple.

Conversely, in 2008–2009, after the crash, entire industries traded at 3–5x earnings. Value investors who bought at 5x based on historical 12x multiples made fortunes. The peer group multiple said "cheap," but intrinsic value said "insanely cheap."

Context Dependency

A P/E of 15 is "expensive" when risk-free rates are 5% and "cheap" when they are 1%. A business earning 7% return on capital (barely above cost of capital) at a 15x P/E is worth nothing. The same business earning 15% ROIC at a 15x P/E is undervalued. The multiple alone tells you nothing.

Earnings Quality Variation

Two companies with the same P/E might have wildly different intrinsic values if their earnings are different quality:

• Company A: 10% growth from recurring customers, stable margins
• Company B: 10% growth from aggressive accounting, one-time gains, seasonal swings

Same P/E; different values. Relative valuation flattens these distinctions.

How Accounting Tricks Distort Multiples

Depreciation and Capitalization

A capital-intensive company expensing maintenance capex has lower reported earnings (lower P/E) than a company capitalizing the same expenses. Same underlying economics, different multiples.

Revenue Recognition

A SaaS company recognizing annual contracts upfront shows high revenue growth; one recognizing ratably shows slower growth. Same cash flows, different P/E ratios based on reported earnings.

Intangible Assets

A company that developed a valuable brand internally doesn't have it on the balance sheet. A company that bought the same brand records goodwill. When calculating enterprise value and EV/EBITDA, one appears cheaper.

Stock-Based Compensation

A company paying employees in stock might report artificially high earnings if not adjusting for dilution. Its P/E looks better than a company paying cash, despite lower total shareholder returns.

The Problem with Sector Multiples

Industries with different risk profiles trade at different average multiples:

• Utilities: 12–15x (stable, regulated, low growth)
• Software: 25–40x (high growth, high ROIC potential)
• Manufacturing: 10–14x (cyclical, capital-intensive)
• Financial Services: 12–18x (leverage-dependent, cyclical)

This variation makes sense. But within sectors, multiples often diverge from fundamentals. A software company with higher growth, stronger margins, and better customer retention should trade at a higher multiple than one with low growth and churn. Sometimes it does; sometimes market sentiment inverts this logic.

When Relative Valuation Fails Most Spectacularly

Comparing across business models: Comparing the P/E of an organically growing business to one growing through acquisitions is comparing apples to oranges. The acquirer might show lower reported earnings due to integration costs, making its multiple look worse despite better underlying quality.

During sentiment shifts: When sentiment toward an industry turns, multiples collapse faster than fundamentals. A "safe" business at 18x earnings with 5% growth becomes a value trap. Relative valuation lags behind intrinsic value during these shifts.

In market bubbles: At peak euphoria, the entire tech sector trades at 40x–50x earnings. Buying the "cheapest" in the group at 35x is still overpaying. Relative valuation provided no protection in 2000 or 2021.

With small-cap vs. large-cap: Small caps trade at lower multiples than large caps due to liquidity risk, but liquidity premiums vary. Buying a small cap at a discount to large-cap multiples without assessing liquidity risk is dangerous.

Capital Structure Distortions

Enterprise Value (EV) = Market Cap + Net Debt + Preferred Stock + Minority Interest

Two companies with identical EBITDA but different capital structures:

• Company A: $100M market cap, $20M net debt, $120M EV
• Company B: $90M market cap, $40M net debt, $130M EV

EV/EBITDA says B is more expensive, but A has higher leverage risk. The multiples hide the different risk profiles.

Worse, when a company raises debt to buyback stock, it can maintain earnings per share (and P/E) while destroying equity value. EV/EBITDA might look unchanged, but the risk has shifted to debt holders.

Specific Multiple Pitfalls

P/E Ratio:

• Ignores capital structure, growth, and ROIC
• Sensitive to working capital timing and one-time items
• Varies widely based on leverage

EV/EBITDA:

• Assumes comparable capex needs (often false)
• Ignores cash tax rates
• Hides debt burden

Price-to-Book:

• Assumes book value reflects real asset value (often false for intangible-heavy businesses)
• Doesn't account for ROIC on assets
• Distorted by intangible accounting

Price-to-Sales:

• Ignores profitability entirely
• Two revenue sources with different margins look the same
• Can justify any unprofitable company as "cheap"

When Relative Valuation Makes Sense

Market dislocations: When a specific stock diverges sharply from peers due to temporary news or forced selling, relative valuation can highlight opportunities. If the fundamentals haven't changed, buying at a discount to peers makes sense.

Screening: Relative valuation is useful for screening thousands of stocks down to a handful to analyze. "Buy the bottom decile of P/E ratios by sector" is a rough filter, not a final valuation.

Sanity checking: After calculating intrinsic value using DCF or EPV, check if the implied multiple is reasonable relative to peers. If your $100/share valuation implies a 25x P/E and peers trade at 12x, your assumptions might be too optimistic.

Simple businesses with stable characteristics: A mature utility or packaged goods company with stable growth and ROIC might trade fairly consistently against peers. Relative valuation has more predictive power here.

The Right Use of Multiples: Adjusted Valuation

Rather than using raw multiples, adjust them for key variables:

PEG Ratio (P/E divided by growth rate): A stock trading at 15x P/E with 15% growth has a PEG of 1.0. One trading at 15x with 5% growth has a PEG of 3.0. This adjusts for growth, though imperfectly.

EV/EBITDA adjusted for capex: Calculate Unlevered Free Cash Flow (EBITDA - taxes - capex) and divide by that instead. This accounts for capital intensity.

P/E adjusted for ROIC: Don't just compare raw P/E; compare P/E relative to ROIC. A 20x P/E is cheap if ROIC is 20%; it's expensive if ROIC is 8%.

These adjustments move closer to absolute valuation, making comparisons more meaningful.

Real-World Example: The Trap

2015: Telecom Stocks

AT&T traded at 13x earnings; Verizon at 13x; competitors at similar multiples. By relative valuation, they looked undervalued vs. tech stocks (20–30x) and offer yields. But normalized FCF growth was ~2%, and capital intensity was high. An absolute valuation (DCF or EPV) would have shown that 13x was fair, not cheap, for 2% growth. Relative valuation lured investors into "cheap" telecom that underperformed for a decade.

2010: Bank Stocks

After the crisis, banks traded at 0.8–1.0x book value vs. historical 1.5x. Relative valuation screamed "buy." But ROIC had compressed, capital requirements had increased, and return stability was suspect. Absolute valuation would have shown that even at 1.0x book, many banks were fairly valued for their newfound risk profile. Relative valuation provided no safety margin.

Common Mistakes

Using the wrong comp group. Comparing a high-growth SaaS company to mature software vendors distorts the analysis. Comparable means similar business model, growth, ROIC, and risk.

Ignoring leverage differences. Two companies with the same EBITDA but different debt levels should be compared on equity value, not enterprise value, without adjusting for risk.

Anchoring to historical multiples. "This stock traded at 20x earnings for the last decade, so 15x is cheap" is backward reasoning. The historical multiple might have been wrong.

Treating multiples as prices. A low multiple doesn't mean low risk. A cyclical stock at 8x earnings might be at peak cycle with earnings about to collapse.

FAQ

Q: Is relative valuation ever more reliable than absolute valuation? A: No, but it's faster. Use it to screen; use absolute valuation to decide. Relative valuation tells you "this is cheap vs. peers"; absolute valuation tells you "this is cheap vs. what it's worth."

Q: What if I can't calculate intrinsic value? A: Then you can't invest with confidence. Relative valuation without an intrinsic value anchor is gambling. Better to admit ignorance than guess based on multiples.

Q: Are historical multiples useful? A: Only to understand what the market has paid before, not what it should pay. A 20x historical multiple doesn't justify paying 20x today if fundamentals have deteriorated.

Q: Should I ignore multiples entirely? A: No. After calculating intrinsic value, multiples are a useful sanity check. If your valuation implies an unreasonable multiple relative to peers and history, your assumptions might be wrong.

Q: How do I adjust for accounting differences? A: Calculate normalized earnings (removing one-time items, adjusting for capitalization vs. expensing). Compare normalized multiples, not reported ones. This is tedious but essential for fair comparison.

Related Concepts

• Intrinsic Value: The true worth of a business, independent of multiples
• Normalized Earnings: Earnings adjusted for cycles, one-time items, and accounting choices
• Price Anchoring: The psychological tendency to rely too heavily on historical prices
• Market Sentiment: How emotion drives multiples away from fundamentals
• Accounting Quality: The reliability of reported earnings

Summary

Relative valuation is a useful check, not a path to intrinsic value. Comparing multiples assumes the peer group is correctly priced, which is often false during market dislocations, bubbles, and crashes. Accounting differences, capital structure variations, and business model divergences make multiples incomparable across companies. The most dangerous use of relative valuation is buying "the cheapest stock in an expensive sector" and assuming that discount provides safety. It doesn't; the entire sector might be overvalued. Always anchor on intrinsic value using DCF, EPV, or Greenwald's framework. Use multiples to cross-check that your intrinsic value implies reasonable multiples, and to identify screening opportunities. But never let multiples be your final answer on value.

Next

Proceed to the next article: When to Use P/E and EV/EBITDA.