Pre-IPO Research: How to Read EDGAR Filings and Build Industry Comparables

Most of what you need to evaluate a company approaching the public markets is already public — it is just buried in dense regulatory filings. Pre-IPO research is the discipline of getting to those filings early, reading them the way an investor does, and putting the numbers in context against comparable companies. This is a practical workflow for doing exactly that, starting from the free primary source and ending with AI tooling that makes it far faster.

Start at the source: SEC EDGAR

Every company registering to sell securities in the US files with the SEC, and those filings are published — for free — on EDGAR, the Electronic Data Gathering, Analysis, and Retrieval system. The fastest way in is EDGAR full-text search, which lets you search the full text of filings by keyword, company, and form type.

For a company heading toward an IPO, these are the forms that matter:

• S-1 (domestic) or F-1 (foreign private issuer) — the registration statement filed to offer securities to the public. This is the centerpiece of pre-IPO research.
• S-1/A — amendments. The terms, financials, and risk factors evolve across amendments; the latest amendment is what reflects the current deal.
• 424B4 — the final prospectus filed after pricing, with the actual offer price and share count.
• DRS — a draft registration statement, submitted confidentially under the JOBS Act by emerging growth companies; it becomes public (as an S-1) before the roadshow.
• 10-K / 10-Q / 8-K — annual, quarterly, and current reports. You will rely on these for the comparable public companies, even when researching a private issuer.

How to read an S-1 for pre-IPO research

An S-1 is long, but the value is concentrated in a handful of sections. Read them in this order:

1. Prospectus summary & business — what the company actually does and how it makes money.
2. Risk factors — the company's own enumeration of what could go wrong. Read these closely; the specific, non-boilerplate risks are the ones that matter.
3. Management's Discussion & Analysis (MD&A) — management's narrative of the financials: growth drivers, margin trends, and the metrics they want you to focus on (and the ones they bury).
4. Financial statements — the audited numbers. Rebuild the revenue, gross margin, and cash-burn trajectory yourself rather than trusting the summary.
5. Use of proceeds, dilution, and principal stockholders — where the money goes, how much new investors are diluted, and who is selling.
6. Underwriting — the banks, the structure, and any lock-up arrangements.

The goal is not to read every page; it is to extract a structured picture — growth rate, unit economics, burn, ownership, and the two or three risks that could break the thesis.

Build an industry comparable set

A company's numbers mean little in isolation. Comparables — "comps" — put them in context by benchmarking against similar public companies. The workflow:

1. Define the peer set. Pick public companies with a similar business model, growth profile, and end market — not just the same broad sector.
2. Pull their filings. Each peer's latest 10-K and 10-Qs are on EDGAR. Extract revenue, growth, gross margin, and operating margin.
3. Compute the multiples. For high-growth issuers, EV/Revenue (often growth-adjusted) is the common lens; for profitable businesses, EV/EBITDA and P/E. Compare the IPO candidate's implied multiple at its proposed price against where the peers trade.
4. Adjust for differences. A company growing 60% with software margins should not trade at the same multiple as a 15%-grower — normalize for growth and profitability before drawing conclusions.

Comparables are where pre-IPO research most often goes wrong: a tidy-looking comp table built from a sloppy peer set produces confident, wrong answers. Spend the time on the peer selection.

Do it faster with AI: IPOGrid and the IPOGrid MCP

Reading filings by hand is the right way to learn the craft, but it does not scale across a pipeline of candidates. IPOGrid is an IPO-intelligence platform that tracks IPO filings, pricings, terms, and listing signals sourced from SEC filings and public-market data — active candidates, priced deals, large transactions, and SPACs — and crucially, it surfaces comparable deals and post-listing outcomes alongside each issuer.

What makes it useful for an AI-assisted workflow is the IPOGrid MCP server. MCP (Model Context Protocol) is an open standard that lets an AI assistant connect directly to external data sources and tools. IPOGrid exposes its pre-IPO dataset over MCP, so an assistant can query it natively instead of you copy-pasting from a dozen filings.

The server lives at https://ipogrid.com/mcp (Streamable HTTP transport) and exposes tools including:

• list_companies — active deals, filterable by market, issuer type, and gross proceeds.
• get_company — full issuer detail by CIK (the SEC's company identifier), including terms, financial snapshots, and comparable deals.
• list_filings — SEC filing events with AI enrichment.
• list_outcomes — post-listing performance.
• list_news and get_chart — coverage and visualizations.

Anonymous access gives limited read-only use; an OAuth or API-key connection unlocks the filing-level and outcome tools. Point your AI client at the endpoint and you can ask, in plain language:

List active IPO candidates raising over $500M, then pull the terms,
financial snapshot, and comparable deals for the largest one.

The assistant calls list_companies, then get_company by CIK, and returns a structured brief — the kind of first pass that used to take an afternoon in EDGAR.

A practical workflow

1. Scan the pipeline with IPOGrid (or its MCP) — what is filed, what is pricing, what is large.
2. Shortlist candidates that fit your thesis by stage, sector, and size.
3. Pull each candidate's latest S-1/A from EDGAR and read the six sections above.
4. Comp it against a hand-picked peer set built from those peers' 10-Ks.
5. Verify every number that matters against the source filing before it goes in a memo.

The filings have always been public. The difference now is how quickly you can get from "a company filed" to "here is the structured, comparable, verified picture" — and that is where pairing EDGAR with AI tooling pays off.

Sources & further reading

• EDGAR full-text search — SEC
• Form S-1 (glossary) — SEC / Investor.gov
• Registration under the Securities Act of 1933 — SEC / Investor.gov
• IPOGrid — IPO filings, pricings, terms, and listing signals
• IPOGrid MCP server & API docs — IPOGrid
• Model Context Protocol — the open standard behind the MCP server
• Related: The IPO path · SPAC vs. IPO · The deal flow lifecycle