Stock Valuation 101: Which Multiple Should You Use?

Too Many Multiples, Not Enough Guidance

Open any stock screener and you'll find a dozen valuation ratios. P/E. Forward P/E. P/S. P/B. PEG. EV/Revenue. EV/EBITDA. EV/EBIT. Price/Cash Flow. Price/Free Cash Flow. It's overwhelming, and most finance content just lists definitions without telling you which one to actually use.

Here's the practical framework. For any given stock, one or two multiples will be the most informative. The rest are either redundant or misleading. The trick is knowing which is which.

The Decision Framework

Start With Profitability

The first question is simple: Does the company make money?

If yes (positive earnings): Start with P/E ratio. It's the most widely used, most intuitive, and most comparable metric. Earnings are what ultimately belong to shareholders, so price-to-earnings is the most direct measure of what you're paying for ownership. If no (negative earnings): P/E is undefined or meaningless. Move to P/S (price-to-sales) or EV/Revenue. These work for any company with revenue, regardless of profitability. If pre-revenue: You're in venture territory. Multiples-based valuation doesn't really apply. Look at comparable transactions, TAM-based models, or just accept that valuation is speculative.

Consider the Capital Structure

Equity multiples (P/E, P/S, P/B) use market cap. They tell you what equity holders are paying. Enterprise value multiples (EV/Revenue, EV/EBITDA) use enterprise value, which is market cap + debt - cash. They tell you what it would cost to buy the whole business.

For companies with significant debt or large cash positions, EV-based multiples are more apples-to-apples. Two companies with identical operations but different leverage will have different P/E ratios but similar EV/EBITDA ratios.

Rule of thumb: When comparing companies across different capital structures (which is most of the time), prefer EV-based multiples. When analyzing a single company's historical valuation, equity multiples work fine since the capital structure is relatively constant over time.

The Multiples, One by One

P/E (Price-to-Earnings)

Best for: Profitable, mature companies. Comparing companies in the same industry with similar growth rates. Watch out for: One-time charges or gains that distort earnings. Companies with heavy stock-based compensation (SBC isn't in GAAP EPS but dilutes shareholders). Cyclical companies at peak or trough earnings. Typical ranges: 15-25x for stable large caps. 25-40x for growth companies. Under 10x for deep value or declining businesses.

P/S (Price-to-Sales)

Best for: Unprofitable growth companies. Comparing companies with different margin profiles (but only within the same industry). Cyclical companies where earnings are temporarily depressed. Watch out for: Ignores profitability completely. A company with 80% gross margins and one with 20% gross margins shouldn't be compared on P/S alone. Also ignores debt. Typical ranges: Varies wildly by industry. 0.5-2x for low-margin businesses. 5-20x for high-growth SaaS.

P/B (Price-to-Book)

Best for: Financial companies (banks, insurance, REITs) where book value reflects tangible assets on the balance sheet. Also useful for asset-heavy industrials. Watch out for: Nearly useless for technology and services companies where the value is in intellectual property, brand, and human capital, none of which appear on the balance sheet. A software company's P/B might be 30x, which tells you nothing useful. Typical ranges: 1-2x for banks. Under 1x might signal distress (or a bargain). Over 3x for financials is expensive.

EV/Revenue

Best for: Same use cases as P/S, but accounts for debt and cash. Preferred when comparing companies with different balance sheets. Standard in M&A analysis. Watch out for: Same limitation as P/S regarding margin differences. Also more complex to calculate since you need the full enterprise value.

EV/EBITDA

Best for: Comparing operating performance across companies with different tax situations, capital structures, and depreciation policies. This is Wall Street's workhorse multiple for a reason: it strips out the stuff that varies for non-operational reasons. Watch out for: EBITDA ignores capital expenditure requirements. A company that needs to spend $1B per year on maintenance capex looks the same on EV/EBITDA as one that needs zero. Also, adding back depreciation and amortization can mask deteriorating assets. Not meaningful for financial companies. Typical ranges: 8-15x for most industries. 15-25x for high-growth companies. Under 6x for mature, slow-growth businesses.

A Quick-Reference Cheat Sheet

Company Type	Primary Multiple	Secondary Multiple
Profitable large cap	P/E	EV/EBITDA
High-growth SaaS (unprofitable)	EV/Revenue	P/S
Bank or insurance company	P/B	P/E
Cyclical industrial	EV/EBITDA	P/E (normalized)
Biotech (pre-profit)	EV/Revenue	P/S
REIT	P/FFO (or P/B)	Dividend yield
Mature dividend payer	P/E	Dividend yield
Turnaround / restructuring	EV/Revenue	EV/EBITDA

Using Multiples Over Time

A single snapshot of a P/E ratio tells you very little. Is 25x expensive for this stock? Maybe. Maybe not.

What's much more useful is seeing how a stock's valuation has changed over time. If a stock historically trades between 15x and 25x P/E, and it's currently at 30x, you know the market is pricing in higher expectations than usual. That's a flag worth investigating.

Historical multiples also reveal how the market has re-rated a company. A stock that traded at 10x P/E five years ago and now trades at 25x has been re-rated by the market, likely because the growth profile or competitive position improved.

Explore the Multiples Tab

StockResearch charts all five of these multiples (P/E, P/S, P/B, EV/Revenue, EV/EBITDA) with daily resolution. Pick any stock, switch to the Multiples tab, and see its full valuation history. Overlay a peer to compare how the market values two companies differently.

Open the Multiples Tab on StockResearch

---

This post is for informational purposes only and does not constitute financial advice. Valuation multiples are one tool among many and should not be used in isolation. Past valuation levels do not predict future returns.