Every flag on this desk is earned.
A stock scanner that hunts overextension setups across 75 liquid names — and refuses to flag anything it can't prove. Two independent engines test conviction, stretch, exhaustion, liquidity, and risk before anything earns attention.
- stocks in the scored universe
- 75
- stocks in the scored universe
- independent strategy engines
- 2
- independent strategy engines
- standard conviction checks
- 4
- standard conviction checks
- buy/sell instructions — research only
- 0
- buy/sell instructions — research only
The gate
Most names never make it.
A flag you can trust is mostly made of refusals. Every scheduled pass walks the 75-stock universe through three evidence stages, followed by four conviction checks. Fail a required check and the name stays research-only.
Gate I — Stretch
Is the move genuinely extended against its own history — or just loud? Distance from the mean, RSI, the slope of the run.
Gate II — Evidence
Is there real exhaustion — momentum divergence, climactic volume, a candle that fails? A tall move alone proves nothing.
Gate III — Price
Can the risk actually be priced — liquid options, a defined invalidation level, honest risk-to-reward?
- illustrative sweep
- 75
- illustrative sweep
- illustrative rejects
- 68
- illustrative rejects
- flagged
- 7
- flagged
Illustrative funnel only; live pass counts vary with the tape and user thresholds.
Representative pass — illustrative, not live data
What a flag is made of
Anatomy of an earned flag
The pennant isn't a logo. It's a diagram of the method — four evidence layers and a gate, drawn into the mark.
Fig. A — The cloth · Overextension
Method
Three entries, every session
A research workflow that does the watching for you, then hands back something you can actually evaluate.
Entry01
Sweep the tape
Every session, the desk walks the whole tracked universe — pricing stretch, momentum, and volume against each name's own history. No chart-watching required.
Entry02
Score the survivors
Whatever clears the gate is graded across the four evidence layers and combined into one transparent score — with the work shown. Nothing is a black box.
Entry03
You make the call
You get the setup, the modeled play, and the invalidation level — then the decision, and the account, stay entirely yours. Anything still building stays on the watch ledger — the moment a name clears the gate, it rings the in-app alert desk.
Signal example
Three names walked in. One earned a flag.
Flick between the three shapes of output — a flag, a watch, a refusal. No hype, no price targets: the evidence, the risk, and the verdict.
Potential pullback after a sharp upside extension
Stretched, exhausted, and priceable — passed all three gates. The modeled play and invalidation are laid out below.
RSI (14)
78.4
Overbought
Dist. from 20-EMA
+14.2%
Stretched
Volume vs 30d avg
2.4×
Climactic
Options liquidity
Tight
Tradable
Risk / reward
1 : 2.6
Favorable
Invalidation
$46.20
Defined
Risk note: an extended name can stay extended. These are research signals, not recommendations — size for the downside and confirm against your own plan.
Sample data (2024 figures) · illustrative — the desk pulls live quotes inside the dashboard
A serious tool for serious research
SignalForge is a market-intelligence platform designed to help you find and evaluate trading opportunities faster. It is not financial advice, a brokerage, or a guarantee of any result. We don't place trades and we don't hold your funds — we hand you sharper, faster research so you can make better-informed decisions.
Trading involves substantial risk of loss and is not suitable for every investor. Past behavior of any signal does not guarantee future results. Always do your own research and consider consulting a licensed professional.
Research, not advice
SignalForge surfaces and scores setups. It does not tell you what to buy or sell — every decision stays yours.
Transparent scoring
Each score shows its work: the inputs, the layers, and the invalidation level. Nothing is a black box.
Risk shown up front
Downside and the level that voids a setup are part of every signal — because risk awareness is the point.
No outcome promises
No guaranteed profits, no “beat the market.” Markets are uncertain; we help you research faster, not predict.
No testimonials and no invented numbers — the desk keeps its own receipts. Each claim below links to the live page that proves it.
Replayed point-in-time
The backtesting explorer replays the real production engine over historical sessions — same code, same gates, no hindsight.
Open the backtesting explorerDated, published editions
The momentum book publishes dated editions on a fixed decision schedule — each one a public page, with the schedule as a subscribable calendar feed.
Open the edition recordThe whole universe, every sweep
Scheduled sweeps walk the full tracked universe — and the desk shows only what survived the gate.
Open the flag ledgerPricing
Get in early
SignalForge is in active development. Start free today — creating an account records your early-access interest for when Pro launches.
Free
Explore the scanner and dashboard with sample and delayed data.
- Flag + pullback scanners
- Personalized signal scoring
- Account watchlist
- Clearly labeled data provenance
Pro
The full research console for active traders — built for speed and depth.
- Options quality + modeled plays
- News, catalysts, and backtesting
- Operator-managed data refresh
- Multi-watchlist workflows
- Notification delivery
Creates a free account — your early-access interest is recorded with it.
No credit card required during beta. Pricing and feature tiers are subject to change before launch.
FAQ
Questions, answered
Stop watching charts. Start reading the tape.
Open the dashboard and see how SignalForge scores today's setups — built for traders who value time, clarity, and disciplined risk.