Beer Score Methodology
JustTheBeer uses a transparent, rules-based approach to estimate the fair value (intrinsic value) of every stock. This page explains exactly how Beer Scores are calculated — what data we use, what assumptions we make, where the model is structurally weak, and how to interpret the results.
Data Sources
All financial data comes directly from SEC EDGAR filings — specifically 10-K (annual) and 10-Q (quarterly) reports, pulled from the SEC Company Facts API. We extract revenue, net income, operating cash flow, capital expenditures, total debt, cash and equivalents, and shares outstanding from the most recent filing available, plus the prior five fiscal-year 10-K snapshots for historical context.
Trailing twelve-month (TTM) values are computed as the sum of the four most recent non-overlapping quarters. When TTM is unavailable (e.g., a company has filed only an annual recently), the model falls back to the latest 10-K annual value and flags the substitution internally.
Market prices are indicative reference prices sourced from public financial data providers. Prices update daily after market close. These are not official exchange feeds.
The Intrinsic Anchor: A Composite of Four Methods
The Beer Score is not driven by a single valuation model. We compute four independent estimates for every stock and combine the two that survive structural critique:
| Method | What It Measures | Weight in Score |
|---|---|---|
| DCF | 5-year projected free cash flows discounted at a sector-default WACC, plus terminal value | 50% |
| Owner Earnings | Buffett-style cash earnings (net income + non-cash charges − maintenance capex), capitalized with a quality-score modifier | 50% |
| Graham Number | Classical book-value × earnings formula (Benjamin Graham, 1949) | 0% — shown for context only |
| Earnings Power Value (EPV) | Bruce Greenwald perpetuity of normalized earnings, no growth assumed | 0% — shown for context only |
Why DCF and Owner Earnings only? Both methods are anchored to free cash flow, which is the cleanest measure of what a business actually produces for owners. On diverse businesses they agree to within 3–6%, which gives the composite built-in cross-validation. Graham is excluded from the score because book value is a structurally wrong yardstick for asset-light modern companies (e.g., Apple's book value per share of roughly $7 implies a Graham number under $40, which measures the wrong quantity rather than being "conservative"). EPV is excluded because its zero-growth perpetuity assumption mechanically underprices growth names and is unstable on noisy inputs. Both are still computed and shown on individual stock pages for transparency.
DCF Default Assumptions
| Sector | Default WACC | Terminal Growth |
|---|---|---|
| Technology | 10.0% – 11.0% | 3.0% |
| Healthcare | 9.0% – 10.0% | 3.0% |
| Financials | 9.0% | 2.5% |
| Consumer | 8.5% – 10.0% | 2.5% – 3.0% |
| Industrials | 9.0% – 9.5% | 2.5% |
| Energy | 9.5% – 10.0% | 2.0% |
| Utilities | 7.5% | 2.0% |
| Real Estate | 7.5% – 8.0% | 2.0% – 2.5% |
Beer Score Formula
Once the composite anchor produces an estimated intrinsic value (fair value per share), the Beer Score is calculated as:
Beer Score = (Intrinsic Value / Market Price) × 100
The score is capped at 100. A Beer Score of 100 means the stock's estimated fair value meets or exceeds its current market price — the glass is "full of beer." A score of 50 means only half the price is supported by fundamentals — the rest is "foam" (speculation premium).
Score Zones
| Zone | Score Range | Interpretation |
|---|---|---|
| Pure Beer | 80 – 100 | Price well-supported by fundamentals. Low speculation premium. |
| Good Pour | 60 – 79 | Solid fundamental backing with moderate growth premium. |
| Getting Foamy | 40 – 59 | Significant speculation in price. Less than 60% supported by fundamentals. |
| All Foam | 0 – 39 | Price driven primarily by narrative, momentum, or future expectations. |
Where the Model Is Structurally Weak
An FCF-based intrinsic anchor is well-suited to mature, cash-generative businesses with stable margins. It is less well-suited to several common situations:
Hyper-CapEx phases: When a company is reinvesting aggressively, current-period free cash flow can collapse or turn negative even when the business is healthy. The model treats sustained negative FCF as "no intrinsic anchor" (Honest-None) rather than guessing — for example, Oracle's most recent fiscal year showed roughly $21B of AI-infrastructure capex against $20.8B of operating cash flow, producing negative free cash flow and triggering the Honest-None path. This is the model working as intended, not a bug.
Asset-light businesses: Graham and EPV measure book-value-anchored quantities that have minimal economic meaning for software, platforms, and IP-heavy businesses. We surface them for context but weight them at zero in the score.
Stable compounders at premium prices: Conservative DCF inputs (single-digit terminal growth, mid-cycle margins) can produce wide downside gaps on businesses the market is pricing for unusual durability — e.g., Coca-Cola or Costco often print large foam percentages under our defaults. Whether the market or the model is right is a judgment call we don't make for you.
Sensitivity to assumptions: A 1% change in WACC can shift the DCF estimate by 15–25%. The 50/50 composite damps this somewhat by averaging with Owner Earnings, which is structurally less sensitive to discount-rate choice.
Per-Ticker Notes
A handful of tickers in our universe have characteristics worth flagging directly. We list them here rather than burying the caveat in a per-stock page.
| Ticker | Note |
|---|---|
| AMZN | Amazon's reported free cash flow is dragged down by AWS-related capex and long-duration lease obligations. The 5-year historical FCF series our DCF projects from understates the steady-state economics of the business; readers should interpret the anchor as a lower bound rather than a center estimate. |
| TSLA | Tesla's free-cash-flow profile swings sharply with capacity build-outs and working-capital effects (deposits, inventory). The composite anchor will be unusually volatile from quarter to quarter and shouldn't be read as a steady fair-value signal. |
| PLTR | Palantir has a relatively short history of GAAP positive earnings and free cash flow, and stock-based compensation is large relative to net income. Owner Earnings adjusts for this only partially. Treat the anchor as informational rather than load-bearing. |
| GOOGL / GOOG | Alphabet has a dual-class share structure: GOOGL is the Class A (voting) share and GOOG is the Class C (non-voting) share. Their economic claim on the business is identical, so both pages use the same intrinsic-value calculation. Persistent price gaps between GOOGL and GOOG reflect voting-premium and float dynamics, not a difference in fundamentals. |
| BRK-B | Berkshire Hathaway is classified internally as Financials, but its operating businesses span insurance, energy, and consumer goods. The standard sector default WACC is a poor fit; we apply a conservative override. |
Important Limitations
Beer Score is a model-based educational indicator, not investment advice. Beyond the structural notes above:
Backward-looking data: SEC filings reflect past performance. The model does not account for management guidance, pending acquisitions, regulatory changes, or other forward-looking factors.
Not a price target: A low Beer Score does not mean "sell" and a high score does not mean "buy." The score measures the gap between current price and a model estimate of fundamental value — one signal among many.
Update frequency: Scores update daily after market close. New earnings data from SEC filings is incorporated within 24 hours of publication.
How to Use Beer Score
Beer Score is best used as a screening tool — a quick way to gauge how much of a stock's price is backed by current fundamentals versus speculation. Use it alongside your own research, not as a replacement for it.
For the full ranking of all stocks, visit the Beer Score Heatmap. For daily highlights, check Today's Foam Report.