How We Turn SEC Filings Into Free Stock Analysis

#webdev #programming #datascience #productivity

Every number on Intrinsiqq comes from one place: companies' own filings with the U.S. Securities and Exchange Commission. That sounds simple, but turning raw SEC XBRL data into clean quality scores, DCF valuations, and 10 years of financials for 8,000+ stocks is genuinely messy work. Here is how the pipeline works, and the parts that are harder than they look.

Why SEC EDGAR

Most finance sites buy aggregated data from third-party vendors. SEC EDGAR is the primary source underneath much of that data: it is free, public, and legally usable, and it contains exactly what companies reported to regulators, not consensus estimates or vendor adjustments. Building directly on it means the numbers you see match the 10-K and 10-Q, and it is why the analysis can stay free.

The pipeline, end to end

Pull the filings. We read each company's structured financial data from the EDGAR XBRL CompanyFacts API, focused on the us-gaap taxonomy.
Map the concepts. Each metric (revenue, free cash flow, debt, shares) is resolved through a chain of candidate XBRL tags, because companies do not all tag the same concept the same way.
Assemble trailing-twelve-month figures. Quarterly filings are stitched into a rolling 12-month window so the numbers are current, not a year stale.
Compute the analysis. From clean financials we derive the quality score, the DCF valuation, dividend safety, and the ratios.

Why XBRL is messier than it looks

The reason "just read the SEC data" is harder than it sounds:

Challenge	Why it is hard	How we handle it
Tag switching	The same concept gets a different XBRL tag over time	A prioritized fallback chain of tags per metric
Deriving Q4	Q4 only appears inside the annual 10-K, not as a quarter	Subtract Q1 + Q2 + Q3 from the annual total
Year-to-date cash flow	10-Q cash flow is cumulative YTD, not the quarter	YTD math so quarters are not double-counted
Duration vs instant	Flows vs point-in-time balance-sheet values	Sum flows over 4 quarters; take the latest for balances
EPS	Summing quarterly EPS mixes different share counts	TTM net income divided by the latest share count
Foreign + fiscal years	20-F filers, non-USD currencies, non-calendar years	Currency fallbacks and filing-type handling

None of this is glamorous. But it is the difference between numbers that match the filings and numbers that quietly drift from them. We chose to do the unglamorous part so the output is trustworthy.

From clean data to a verdict

Once the financials are clean, the analysis is deterministic and fully documented. The quality score is a weighted composite of eight fundamental checks; the DCF is a two-stage discounted cash flow you can adjust yourself; the dividend score weighs safety and growth. The same clean, TTM-assembled data also drives a fundamental charting tool, so any metric (or its historical valuation multiple) can be plotted over a decade and broken down by business segment. Every figure traces back to a specific filing, and the full methodology is public.

See the output on a real stock: analyze AAPL free →
Quality score, DCF fair value, and 10 years of SEC-sourced financials. Free, no account.

Why build it this way

Building on the primary source is more work than licensing a feed, but it is what lets Intrinsiqq be free, transparent, and auditable. You can check any number against the original filing, and we can show our work on the methodology page. For a tool meant to help people make real decisions, that traceability is the whole point.

Sources

Originally published at intrinsiqq.com.