Small-Cap Metrics & Screening: The SCN Methodology

Most stock screeners break below $500 million in market cap. Data vendors mis-map micro-cap financials, "P/E" fields mix adjusted and GAAP earnings, and stale share counts make half the market-cap figures wrong. Screening small caps well means knowing which metrics survive contact with bad data — and pulling the numbers from the source. This page is both: the metrics SCN trusts, and the exact methodology behind the Weekly Screen. It's the standing reference every screen issue links to.

The metrics that matter below $2B

Enterprise value over EBIT (or over free cash flow), never P/E alone. Small caps carry wildly different cash and debt loads, and P/E ignores both. A "cheap" 8x P/E with a debt wall is more expensive than a 15x with net cash.

Free-cash-flow conversion. Operating cash flow minus capex, divided by net income, over three years. Below ~70% persistently, the earnings are accrual-flavored; investigate before trusting any multiple built on them.

Share-count trajectory. The five-year diluted share count is the most under-used number in small-cap investing. It's the difference between a company financing growth from operations and one running a dilution treadmill. Growth per share is the only growth that pays you.

Insider ownership and insider activity. Ownership above ~10% aligns the board with you. Open-market cluster buying is the strongest positive signal; see the methodology below for how we define it strictly.

Liquidity. Average daily dollar volume determines whether you can ever exit at the price on the screen. A rule of thumb: if your intended position is more than a few percent of a day's volume, your real exit price isn't the quoted one.

Sector-specific vitals. For the community banks that dominate the small-cap universe — net interest margin, efficiency ratio, and tangible book value growth do the work that EV/EBIT does elsewhere. For pre-revenue biotech, months of runway and the trial calendar are the whole model.

Metrics that mislead down here

Adjusted EPS taken at face value — always rebuild it. Revenue growth without unit economics — growth bought with negative gross margin is a liability. Self-reported "cash runway" — recompute it at the current burn, not management's projected burn. And any screener field you didn't verify against the actual filing: at this end of the market, the data vendor is wrong often enough to change conclusions.

How the SCN Weekly Screen is built

Every screen starts from primary sources — SEC EDGAR filings and XBRL financial data — not from a commercial screener's database. The universe is U.S.-listed companies, roughly $50M to $2B in market cap, with market caps computed from current share counts and verified market prices. Four franchises rotate:

Insider Buying Clusters. Every Form 4 accepted by EDGAR in the trailing window; open-market purchases only (transaction code P, and we drop stale transactions reported late); related filers — funds filing through multiple vehicles, joint filings by a fund and its principals — collapsed to one economic buyer; a cluster requires two or more insiders filing separately with their own money. Uniform-price, same-day "purchases" get checked against offering documents before we call them open-market conviction. Non-traded BDC and fund seedings are stripped entirely.

Uplisting Watch. OTC-listed names showing the paper trail that precedes a national-exchange listing: Form 8-A registrations, uplisting language in 10-Qs and press releases, reverse splits engineered to meet listing price minimums, and governance build-outs.

Factor-Score Movers. A composite of value, quality, and momentum factors computed from XBRL fundamentals; the screen surfaces the biggest score changes, not the biggest scores — the point is inflection, not level.

New 13D/13G Activity. Fresh activist and passive-stake filings in the band, with attention to first-time 13Ds, amendments showing accumulation, and filers with real track records.

How to read a screen

A screen is a candidate list, not a buy list. Every issue publishes the full result set, including the names that make the signal look bad — a screen that only shows winners is marketing, not research. From there, the work in the analysis guide takes over: filings, earnings quality, balance sheet, catalyst. We also publish a monthly lookback on how prior screen names performed, because keeping score is the difference between a methodology and a highlight reel.

---

Disclosure: The author has no position in any security mentioned. This article is for informational purposes only and is not investment advice. See our Disclaimer and Disclosures.