Building a Company Analysis Framework from Zero Using Public Data

You don't need a Bloomberg terminal or a Wall Street pedigree to analyze a company properly. Nearly everything that matters is free. SEC EDGAR, the Federal Reserve's FRED database, Census Bureau tables, patent databases, court records, and the company's own website hold more information than any one analyst could get through in a career. The hard part is knowing what to pull, in what order, and how to turn a folder of PDFs into an actual opinion.

I've watched smart people burn a whole weekend downloading filings and come away with nothing usable. What separates a useful analysis from a pile of data is a repeatable framework. Here's the one I'd build if I were starting from zero, step by step, using only public sources.

Step 1: Decide What Question You're Answering

This sounds obvious, and almost everyone skips it. Analyzing a company in the abstract produces a report nobody uses, including you. Are you evaluating it as an investment? Sizing it up as a competitor? Doing diligence ahead of an acquisition? Deciding whether to take a job there?

Each question weights the work differently. An investor cares most about valuation, earnings quality, and growth. A competitor wants pricing, positioning, and customer base. An acquirer worries about integration risk and whether the promised synergies survive contact with reality. A job candidate mostly wants to know whether the company will still be growing in three years. Write your question at the top of the page before you open a single filing, and let it decide where your hours go.

Step 2: Start with the Company's Own Paper Trail

For public companies, SEC filings are the backbone. The 10-K gives you annual financial statements, the business description, risk factors, and management's discussion of results. The 10-Q updates that picture quarterly. The DEF 14A proxy statement covers executive pay and governance, and it's where boards disclose things they'd rather you not read too closely. Form 4 filings track insider buying and selling, and 8-Ks flag material events between reports. All of it is free on EDGAR, and EDGAR's full-text search will surface every filing that mentions a product, a customer, or a lawsuit by name.

One habit worth stealing from professional analysts is reading the footnotes before the highlights. Revenue recognition policies, lease obligations, pending litigation, and related-party transactions all live in the fine print, and the fine print is usually where the story starts diverging from the press release.

Private companies leave a thinner trail, but not an empty one. Their websites reveal team, products, and strategic direction. LinkedIn shows headcount and hiring patterns. Crunchbase and PitchBook track funding rounds. State incorporation records, UCC filings, and court dockets fill in the rest. You won't get audited financials, but you can assemble a surprisingly clear silhouette.

Step 3: Build the Financial Picture

Pull five years of data if you can get it: revenue, gross profit, operating income, net income, operating cash flow, free cash flow, total assets, total liabilities, shareholders' equity, and total debt. Five years lets you see trends instead of snapshots, and the trends are where the insight lives.

Then calculate the standard ratios: revenue growth, gross margin, operating margin, net margin, return on equity, return on assets, debt-to-equity, current ratio, and free cash flow conversion, which is free cash flow divided by net income. Plot them over time and ask simple questions. Are margins expanding or compressing? Is debt growing faster than the business? Is cash keeping up with earnings?

That last one deserves a worked example. Say a company reports net income of roughly $400 million a year, flat for three straight years, while free cash flow slides from $380 million to $150 million over the same stretch. The income statement says nothing changed. The cash flow statement says something did, and the footnotes will usually tell you what: receivables ballooning because customers are paying late, costs being capitalized instead of expensed, or revenue being recognized well ahead of the cash arriving. When earnings and cash disagree for more than a year or two, trust the cash.

If you want a second opinion on financial strength, two classic academic tools are computable straight from the statements you just pulled. The Altman Z-Score flags distress risk, with scores below 1.8 traditionally read as the distress zone. The Piotroski F-Score, from his 2000 paper, is a nine-point checklist of fundamental health covering profitability, leverage, and operating efficiency. Neither one is a verdict on its own, but both are quick, free, and force you to look at the right line items.

For private companies with no published financials, lean on proxy metrics. Headcount trends on LinkedIn, web traffic estimates, app download rankings, job posting volume, and review counts on platforms like G2 or Trustpilot won't hand you an income statement, but together they sketch the growth trajectory.

Step 4: Map the Competitive Landscape

No company exists in a vacuum, and the 10-K conveniently names the competitors management actually worries about, usually in the business description or the competition section. Start there, then branch out through trade publications, industry association reports, and the market research summaries you can often reach for free with a public library card.

Build a simple matrix comparing the top five to ten competitors on revenue, growth, product breadth, geographic reach, and whatever operational metric matters in that industry. Don't chase precision here. You're after relative position, and knowing a company is clearly number two and gaining ground is more useful than a market share estimate carried to two decimal places.

Step 5: Judge Management by What They Did with the Money

Management quality is probably the biggest driver of long-term performance, and it's also the hardest thing to measure. The best proxy I know is capital allocation history, because all of it is on the record.

Look back five to ten years. Did acquisitions create value, or did they end in goodwill impairments? An impairment a few years after a splashy deal is management grading its own homework and failing it. Were buybacks executed when the stock was cheap, or at the top? Has the dividend been grown, held, or quietly cut? These decisions are documented in the filings, and they don't flatter anyone the way an interview does.

Then look at the people. Heavy C-suite turnover is worth investigating, and a sudden CFO departure deserves particular attention. The proxy statement shows whether the board holds independent directors with real industry experience or is padded with friends of the founder. Form 4 filings help here too, since insiders sell shares for plenty of innocent reasons but tend to buy on the open market for only one. Glassdoor rounds out the picture; discount any single review, but when hundreds of employees independently describe the same leadership problems, treat the pattern as evidence.

Step 6: Check the Industry Backdrop

Even a well-run company struggles in a shrinking market. Bureau of Labor Statistics data shows employment trends by sector. FRED, the Federal Reserve Bank of St. Louis database, covers interest rates, credit conditions, and hundreds of series relevant to specific industries. Census Bureau data helps with anything consumer-facing, since demographics and geography drive demand.

Industry trade associations publish annual reports with market sizing and trend commentary, and public libraries often subscribe to research databases that individuals can't justify buying. You're trying to answer three questions. Is the industry growing or shrinking? Which structural forces are reshaping it, whether regulation, technology, or demographics? And where is it in its cycle, because buying a great cyclical business at the top of the cycle is a classic way to be right about the company and still lose money.

Step 7: Look for What Could Break It

Every 10-K has a risk factors section, and most of it is boilerplate the lawyers insisted on. The useful move is comparing this year's risk factors against last year's. Companies add new risks when something real has changed, so the year-over-year diff tells you more than the section itself ever will.

Beyond the filing, search PACER for federal litigation and state court databases for the rest. Check the relevant regulator's site for enforcement actions, then run a plain news search for recent controversies. For private companies, look at Better Business Bureau complaints, consumer review patterns, and public records for UCC filings, tax liens, and judgments. UCC filings in particular can reveal lending relationships and pledged collateral that the company would never volunteer.

Step 8: Force It All onto One Page

By now you have a lot of material, and the framework only pays off if you compress it. Write a one-page summary that covers:

• Business model: how the company actually makes money
• Financial health: whether the balance sheet and cash flows are sound
• Trajectory: growing, stable, or declining, and why
• Competitive position: where it stands and which direction it's moving
• Management: competent and aligned with shareholders, or not
• Key risks: the two or three things that could genuinely hurt it
• Your call: a conclusion stated plainly enough that you could defend it out loud

The one-page constraint forces you to prioritize, and if the thesis won't fit, you probably don't understand the company well enough yet. Writing it down is also the fastest way to find the gaps you didn't know you had.

Where Tools Fit In

Plenty of tools compress the mechanical parts of this. We built FirmAdapt to pull from these same public sources and organize them into structured company analyses, and even without a platform, EDGAR full-text search, FRED, and Google Scholar each save hours on their own. Use whatever removes the drudgery.

Just don't confuse tooling with analysis. A disciplined analyst with a spreadsheet and free public data will beat a careless one with an expensive terminal subscription, because judgment is the one layer no tool supplies.

Start with one company this week and run all eight steps, even the ones that feel unnecessary for your question. Keep the one-pagers somewhere you can find them again. After ten companies you'll have real intuition for what matters in your corner of the market, and after fifty you'll spot in minutes the problems that used to take an evening, because the framework itself gets sharper every time you run it.