Financial Health Score Methodology: StockTitan Complete Guide

Educational Guides beginner fundamental-analysis financial-health

What this score measures and what it doesn't

When investors look at a company, two separate questions need answering: is the business good, and is the stock a good price? The Financial Health Score answers the first question only.

Business quality is about durability and operational health. A high-quality business generates profits consistently, grows its revenue year after year, doesn't drown in debt, keeps enough cash on hand to handle a rough quarter, converts earnings into free cash flow, and produces strong returns on the capital it uses. These are structural properties that change slowly. They're what makes Warren Buffett want to own a business for twenty years.

Stock valuation is about price relative to value. You can have a beautiful business that's been bid up to an absurd multiple, and a mediocre business trading at a steep discount. Valuation changes every second the market is open. It's a different question entirely.

Our score measures the first question rigorously. It says nothing about the second. When you see NVIDIA score 95 on Financial Health and Apple score 78, that reflects business-quality metrics only. Whether either stock is attractive at its current price is a valuation question, one you need to answer separately with multiples, DCF analysis, or your own judgment.

The six radar dimensions

1. Profitability

Measures how much money the company keeps from each dollar of sales after operating costs. We use an adjusted operating margin, operating income minus one-time gains on asset sales, weighted over the last three years (50% latest, 30% prior, 20% two years back). The adjustment matters: a company that sold a subsidiary for a one-time $500M gain isn't becoming more profitable, and the adjusted margin excludes that distortion. Tech companies typically land in the 20 to 60% range; industrials more commonly in 5 to 15%; grocers and distributors in 2 to 5%. Because we rank within sector peers, tech and industrial companies are each compared to their own distributions, a 50% operating margin in software is average; a 20% operating margin for a grocery chain is exceptional.

2. Growth

Measures revenue expansion over the past three years. We compute the compound annual growth rate (CAGR): if revenue went from $10B to $15B over three years, that's roughly 14.5% annual growth. Over short periods this can be volatile, a company that lost a major customer and then won it back will show a growth blip. The three-year window smooths that. For very young companies with under three years of public filings, we fall back to year-over-year growth.

3. Leverage

Measures how much debt the company carries relative to its ability to service that debt. Our primary metric is Net Debt / EBITDA: long-term debt minus cash on hand, divided by earnings before interest, taxes, depreciation, and amortization. This is the single most widely used leverage ratio in credit analysis. A Net Debt/EBITDA below 1x is very conservative; 3x is moderate; above 6x is aggressive. Lower is better. Some industries (REITs, utilities, telecoms) run higher leverage by design, which is why each family is ranked within its own peer group rather than against the universe.

For companies where book equity has been compressed by aggressive share buybacks (Apple, for example, with its huge repurchase program), traditional Debt-to-Equity ratios become misleading. Net Debt/EBITDA captures actual debt-service capacity without being distorted by accounting quirks of the equity side.

4. Liquidity

Measures whether the company can cover its short-term obligations. The current ratio, current assets divided by current liabilities, is the standard. Above 1.0 means the company has more near-term assets than near-term bills. Above 2.0 is very comfortable. Mature companies with efficient working-capital management (Apple, Costco) often run ratios near 1.0 because they collect from customers faster than they pay suppliers, which is a feature, not a risk. Our scoring recognizes this by ranking within sector peers.

5. Cash Flow

Measures how much actual cash the business generates after funding its own capital needs. The free cash flow margin is (operating cash flow - capital expenditures) divided by revenue. Real FCF, not an operating-cash-flow proxy, if the business has heavy capex requirements, that shows up in the numerator. 20%+ is exceptional, 10 to 20% is strong, 5 to 10% is healthy for most mature businesses. Negative FCF can be fine for a fast-growing company investing in expansion, or a warning sign for a mature company losing its pricing power. Context (and the Growth dimension) disambiguates.

6. Returns

Measures how efficiently the company uses invested capital to generate profit. We use Return on Capital Employed (ROCE), operating income divided by (total assets minus current liabilities). ROCE is the metric Warren Buffett and Charlie Munger talk about in Berkshire Hathaway's annual letters. It tells you whether the business is earning more than the cost of its capital. Companies with ROCE consistently above 15% are creating substantial economic value; below 8% and they're barely covering their cost of capital, or destroying value outright.

Stability matters here almost as much as the level. A company with ROCE jumping from 5% to 25% and back to 10% is less attractive than one with consistent 15% year after year. We embed a stability factor into the Profitability and Returns dimensions: 70% of the score is the level, 30% is how stable that level has been over five years.

Sector families, why one template doesn't fit all

Applying an operating-company scoring template to a bank produces nonsense. Banks don't have "cost of goods sold"; their "inventory" is loans; their "debt" is customer deposits; their "efficiency" is measured by a ratio that has nothing to do with gross margin. REITs take on huge debt loads by design, Debt-to-Equity ratios that would be alarming in a tech company are completely normal for a warehouse-owning trust. Clinical-stage biotechs are supposed to be losing money; that's the operating plan.

StockTitan scores each company under rules tailored to its kind of business. Six scoring families cover the US-listed universe:

Operating (default)

Technology, Industrials, Healthcare (excluding clinical-stage biotech), Consumer Cyclical, Consumer Defensive, Basic Materials, Communication Services. The six dimensions described above apply directly.

Banks

Banks are scored on: ROA (earnings on the asset base), net-revenue growth, equity-to-assets (a proxy for regulatory capital adequacy), efficiency ratio (noninterest expense as a percentage of net revenue, lower is better, best-in-class ~50%), provision-to-loans (credit-quality proxy, higher provisions mean more trouble in the loan book), and earnings stability (five-year coefficient of variation on net income). These are the metrics every bank analyst uses; they're also what regulators look at under the CAMELS framework.

Insurers

Insurers are scored on: loss ratio (claims as a percentage of premiums earned, lower is better, best insurers run 60 to 70%), premium growth (written-premium CAGR), capital (equity-to-assets), investment yield (return on invested reserves), combined ratio (claims plus underwriting expenses, divided by premiums, below 100% means underwriting profit), and earnings stability. Progressive, Travelers, and Chubb are examples of insurers that score well on these metrics, low loss ratios, combined ratios under 100%, stable earnings.

Real Estate (REITs)

REITs are scored on Funds From Operations (FFO)-based metrics rather than net income, because GAAP net income is distorted by depreciation charges on real estate that is typically appreciating. FFO ≈ Net Income + Depreciation & Amortization - gains on property sales + impairments. We compute it from the underlying XBRL components extracted from the 10-K. Dimensions: FFO margin, FFO growth, leverage (Net Debt / EBITDA, REITs run higher than operating companies but 4 to 6x is healthy, 8x+ is stretched), interest coverage, AFFO margin (FFO minus recurring capex), and dividend coverage (FFO divided by dividends paid, above 1.3x means the dividend is well-covered).

Energy

Operating-template dimensions, but smoothed over a five-year window instead of three to dampen commodity-price cycles. Exxon scored against oil cycles looks different than Exxon scored against a single high-oil-price year.

Utilities

Operating-template dimensions with five-year windows. Percentile ranking within the utilities cohort handles the debt question automatically: utilities run structural leverage that would be alarming elsewhere, so a 4x Net Debt/EBITDA is around median for the group.

Emerging (clinical-stage biotech and similar pre-revenue companies)

When a company has had negative net income in three or more of its last five years, has revenue under $200M (or R&D spending over 30% of revenue), and revenue under $1B total, we classify it as Emerging. These are typically clinical-stage biotechs, gene-editing startups, or other pre-commercial development-stage companies. Scoring them on Profitability and Returns is meaningless, they're supposed to be burning money. Instead we score:

• Cash Runway: how many months the company can operate at its current burn rate before needing more cash.
• Dilution Discipline: the three-year CAGR of shares outstanding. Companies that issue more and more stock to fund operations are diluting existing shareholders; companies that hold share count steady are more disciplined.
• R&D Intensity: how much of operating expenses go to R&D. For a clinical-stage biotech, a high R&D intensity is a feature, that's the whole point of the company.
• Revenue Progress: CAGR of revenue (milestone payments, collaboration income, early commercial revenue).
• Burn Trend: is the operating cash-flow burn getting better or worse over three years?
• Balance Sheet: cash on hand divided by total liabilities.

A healthy Emerging-family company isn't one that's profitable, it's one with enough runway, contained dilution, real R&D spend, improving burn, and a strong balance sheet. Emerging scores should be interpreted within the family and not compared directly to a mature Operating-family company. A clinical biotech scoring 60 under Emerging rules is a healthy clinical biotech; it doesn't make it more attractive than Apple at 78.

Percentile-within-family ranking

For each dimension, we collect the raw metric values from every company in the family (all Operating companies, or all Banks, or all REITs, etc.), sort them, and assign percentile ranks from 1 to 99. That percentile is the dimension score. A 95 on Profitability means the company ranks in the top 5% of its family on that specific metric.

This approach has several advantages over hardcoded thresholds. It's self-calibrating: as the universe evolves over years, the percentile ranking stays meaningful, a 90 still means "top 10%" even if the distribution shifts. It handles sector differences automatically: tech companies have structurally higher margins than industrials, so a 30% operating margin is merely average in tech but exceptional in industrials, and percentile-within-family captures that distinction without requiring per-sector calibration. And it produces honest rankings: a tech company at 85 and a bank at 85 are each in the top 15% of their peers, which is genuinely comparable even if the underlying numbers look very different.

Data-quality flags

The numeric score says how the company compares. The flags tell you when to dig deeper before trusting the score. Flags don't move the number, they show up as pill-shaped badges next to the row.

• One-time Gain: recent net income was significantly inflated by non-operating items (asset sales, discontinued-operations gains, tax reversals). The ranking score adjusts for this automatically via the adjusted operating margin, but the flag alerts you to look at the reported figures carefully.
• One-time Loss: reported net income was depressed by non-operating charges (impairments, write-downs, litigation settlements). Operating performance may be stronger than net income suggests.
• Discontinued Ops: a significant portion of reported net income came from discontinued operations. These won't recur in the future.
• Buyback Skew: the company's book equity has been heavily compressed by share repurchases, making Debt-to-Equity ratios misleading. We use Net Debt/EBITDA for Leverage anyway, but the flag is informational.
• Limited History: fewer than three years of annual filings available. The 3-year averages compress to whatever data exists; confidence is lower.
• Emerging Stage: the company is scored under Emerging-family rules. Compare within the family rather than cross-family.
• Sector-Specific: scored under Banks, Insurers, Real Estate, Energy, or Utilities rules rather than the default Operating template. Cross-family comparisons are informative but not apples-to-apples.
• Data Suspect: some raw figures in the SEC filings exceeded plausible bounds (operating margin above 200%, return on equity above 1000%, etc.) and were excluded from the scoring. This is typically an XBRL tagging error by the filer and affects about 1% of tickers.
• Earnings Quality: the Beneish M-Score, an eight-ratio model that detects patterns sometimes associated with earnings manipulation, flagged this ticker. False-positive rates are around 25%, so this is a prompt to review carefully, not a definitive conclusion.
• Missing Data: two or more of the six radar dimensions couldn't be computed due to unavailable inputs. The missing dimensions count as zero in the overall score, which is the conservative treatment.

Worked examples

NVIDIA (Operating family, target score ~95)

NVIDIA has 60%+ operating margins, >100% revenue CAGR over three years, near-zero net debt (cash exceeds long-term debt), current ratio near 4, FCF margin in the mid-40s, and ROCE approaching 80%. Every dimension is top-decile among Operating companies. The resulting score approaches the theoretical maximum. This says nothing about whether NVIDIA at $1,200 per share is a good investment, that's a valuation question.

JPMorgan Chase (Banks family, target score ~70)

JPM has a strong ROA of 1.5%, is well-capitalized, maintains an efficiency ratio in the mid-50s (excellent for a money-center bank), and shows stable earnings over five years. Its ranking score under Banks rules reflects that it's a top-tier bank, but banks as a group are structurally lower-returning than tech companies, so even the best bank doesn't score 95. That's honest.

Prologis (Real Estate family, target score ~85)

Prologis runs FFO margins near 80%, has Net Debt/EBITDA around 4x (comfortable for a logistics REIT), strong interest coverage, growing FFO per share, and well-covered dividends. Under Real Estate family rules it ranks near the top of REITs.

Apple (Operating family with Buyback-Skew flag, target score ~80)

Apple has 30%+ operating margins, extremely low Net Debt/EBITDA, and one of the best ROCE numbers in the S&P 500. But book equity has been compressed by massive buybacks, and Apple's revenue growth has slowed to low single digits in recent years. The score reflects that: elite profitability and returns, but mature growth, tight liquidity, and some compression on leverage ratios when you look at D/E instead of Net Debt/EBITDA. The Buyback-Skew flag alerts you that the equity side of the balance sheet is distorted.

Beam Therapeutics (Emerging family, target score ~30)

Beam has about 10 months of cash runway, 12% share dilution per year, 78% R&D intensity, declining milestone revenue, worsening burn, and cash equal to 1.2x total liabilities. Under Emerging rules the score reflects a clinical-stage company with reasonable runway but elevated dilution and deteriorating burn, an honest low-to-moderate score that isn't comparing Beam to Apple.

What the score cannot tell you

The Financial Health Score is a rigorous read on current fundamentals. It has important limitations.

It is not forward-looking. Scores reflect recent historical data. If a company just won a major contract or lost a patent case, the score won't reflect that until the next annual filing.

It does not measure management quality or competitive moat. Those are qualitative assessments that require human judgment. A company can have excellent metrics and terrible management, or mediocre metrics and an unassailable competitive position. We don't claim to score those things.

It does not measure valuation. Repeating this because it's the single most common misunderstanding: a 95 is not a buy signal, a 40 is not a sell signal. Quality and valuation are separate questions. Combine health analysis with valuation analysis for investment decisions.

It is not a substitute for your own research. The score synthesizes a lot of information in one number, which is useful for comparison but not for conviction. Before buying a stock, read the 10-K, understand the business, form a thesis, and stress-test it.

Frequently asked questions

Why does my favorite stock score low?

Usually one of three reasons. The company is in a structurally-lower-returning sector (utilities, commodity producers) and is ranked within that lower-returning peer group. Or recent financials reflect a challenging period, a failed product, a contract loss, a cycle downturn. Or the XBRL data for that company has extraction issues (check for the Data Suspect or Missing Data flags). The score measures the current picture; a low score is sometimes a great buying opportunity when you have a thesis that says the picture will improve.

How often does the score update?

Scores are recalculated after each annual filing (10-K). The weekly data pipeline processes new filings and refreshes the percentile rankings across all families. Quarterly filings (10-Q) update the underlying data but don't change the annual-based score until a new fiscal year closes.

Is a 90 a buy signal?

No. A 90 says the business is in the top 10% of its peer group for quality. Whether the stock at today's price is a good purchase depends on the valuation, what you're paying relative to the business's earnings, cash flow, or book value, and on your expected time horizon. A 90-quality business at a 40x earnings multiple may be a worse investment than a 65-quality business at 15x. The two questions are independent.

Can I use this for day trading or short-term trading?

No. Financial Health is a long-horizon signal based on annual financial statements. Daily price movements are driven by news, positioning, flow, and sentiment, factors this score doesn't attempt to capture. For short-term trading, you want technical indicators, momentum signals, and news flow, not fundamental quality scores.

Why are banks, REITs, and biotechs scored separately?

Because applying the same formula to all of them produces garbage. Banks don't have a current ratio in any meaningful sense; their "current liabilities" are customer deposits. REITs run leverage ratios that would be alarming in a tech company but are normal for a warehouse-owning trust. Clinical-stage biotechs are supposed to be losing money. Each of these groups needs a scoring template that matches how the business actually works. The cost is that cross-family comparisons are not perfectly apples-to-apples, but they're closer to honest than a one-size-fits-all score would be.

What's the difference between this and the Piotroski F-Score or Altman Z-Score?

Those are classic, narrow models. The Piotroski F-Score is a nine-item binary checklist for fundamental-trend strength (published 2000). The Altman Z-Score is a five-ratio linear model designed to predict bankruptcy (originally 1968). Both are useful, both are shown on StockTitan alongside the Financial Health Score. The Health Score is broader: sector-aware, percentile-ranked, multi-dimensional, and updated with modern quality factors (ROCE, gross profitability, margin stability). The three signals together give you a multi-angle view.

The information provided in this article is for educational and informational purposes only. It does not constitute financial advice, investment recommendation, or an endorsement of any particular investment strategy. Past performance does not guarantee future results. Investors should conduct their own research and consult with a qualified financial advisor before making investment decisions.