← Back to Defense Invest

How we score defense stocks

Methodology — Defense Invest

Defense Invest is an information and research platform. Nothing here is investment advice, a recommendation, or a personalized suggestion to buy or sell any security.

What this site is

Defense Invest tracks 116 publicly listed companies in the defense, aerospace and adjacent technology sectors. Every trading day, an automated pipeline collects news, regulatory filings, government contract awards, public tenders, market prices and analyst estimates, then recomputes a composite score for each company in the universe.

The score is a sorting tool: it helps decide which of 116 names deserve a closer look first. It is not a rating, not an analyst opinion, and not investment advice.

How the composite score works

Each company receives six pillar scores from 0 to 100, recomputed daily. The composite is their weighted average:

  • Financial quality — ×1.5: profitability, net loss relative to revenue, cash runway, revenue growth and margins from reported financials.
  • Valuation — ×1.2: price relative to growth — native PEG when available, otherwise a PEG computed from forward P/E and expected growth, otherwise an EV/Sales ladder adjusted for growth.
  • Fundamentals vs price (divergence) — ×1.0: whether the fundamental picture is improving or deteriorating faster than the share price suggests.
  • Analyst revisions — ×1.0: direction and magnitude of recent changes in analyst estimates and price targets.
  • Momentum — ×0.5: weighted flow of genuinely new information over recent weeks (see News weighting below), plus insider buying.
  • Dated catalysts — ×0.5: presence and proximity of confirmed dated events — earnings, program decisions, contract deadlines — within the coming months.

Why timing signals weigh less

Momentum and dated catalysts describe when something may happen, not whether the underlying business is sound. A burst of news flow or an upcoming earnings date says nothing about balance-sheet quality, so both pillars are deliberately capped at half the weight of the quality core (financial quality, valuation, divergence, revisions).

Internal review notes are excluded from the score entirely. The scores are fully deterministic; AI-generated notes only document why a score moved. Feeding a news-derived rating back into a score that is already driven by that same news would be circular reasoning, so we don't do it.

Handling missing data

Coverage is uneven across a global universe. When a pillar has no usable data for a company, it is excluded from the weighted average rather than filled with a neutral placeholder value — averaging a fabricated 50 with real measurements would silently distort the result. The remaining weights are renormalized, and a pillarsUsed field records exactly which pillars fed each published score.

The financial gate

Independently of the pillar average, hard caps apply based on reported financials:

  • Net loss greater than 5× revenue → composite capped at 35
  • Net loss greater than 2× revenue → composite capped at 55
  • No financial data available → composite capped at 40

This is a guardrail against speculative stories that would otherwise score well on momentum and catalysts alone: heavy news flow and an upcoming milestone cannot lift a company that is losing several times its revenue.

Score bands

  • 65+ — strongest combination of quality and timing signals in the tracked universe
  • 58-64 — solid signal profile
  • 50-57 — mixed
  • below 50 — weak, or capped by the financial gate

These thresholds are calibrated on the actual score distribution of the tracked universe, not on an abstract scale. If the distribution shifts materially, the bands are recalibrated.

News weighting

Not all news is equal. Each item carries a signal weight:

  • 1.0 — primary sources: material SEC filings, government contract awards, public tender notices, specialized defense press.
  • 0.25 — aggregators and syndications: reprints and wire pickups of information already counted elsewhere.
  • 0 — administrative exhibits: routine filings and attachments with no informational content.

The aggregate contribution of low-weight sources is also capped, so a syndication spike cannot dominate the momentum pillar.

Data sources

  • Yahoo Finance — prices, reported fundamentals, analyst estimates and price targets
  • SEC EDGAR — filings and insider transactions
  • USAspending.gov — US federal contract awards
  • SAM.gov and TED (EU public procurement) — open tenders
  • Specialized defense and aerospace press — sector reporting and program news

Collection and recomputation run automatically every day.

Limitations

We would rather state these plainly than overstate precision.

  • Coverage is uneven. Companies listed outside English-language markets often have little or no accessible press coverage, which weakens their momentum and catalyst signals relative to US-listed peers.
  • Activity descriptions are generated automatically from public company profiles and may contain inaccuracies or be out of date.
  • The score is a triage tool, not a recommendation, and it says nothing about position sizing, portfolio fit or risk tolerance.
  • The data history is recent. Trend and delta signals should therefore be read with caution until longer series accumulate.

Not investment advice

This site is provided for informational and research purposes only. It does not constitute investment advice, a recommendation, or a personalized offer or solicitation of any kind. Investing carries a risk of capital loss, including total loss. Past performance and displayed scores are not indicative of future results. Consult a licensed financial professional before making any investment decision.