Frequently Asked Questions

Early Signal is an AI-powered quantitative investment platform that tracks the entire AI supply chain across 10 layers — from raw materials and semiconductor manufacturing to AI applications and secondary beneficiaries. It generates daily, weekly, and quarterly signals to help investors understand momentum, risk, and opportunity in AI-related stocks. Read the full methodology.

Free users get a 24-hour delayed dashboard snapshot, access to the full AI universe browser (all 10 layers), a personal portfolio tracker, and watchlist. You also get previews of backtests, bottlenecks, and stress tests. Just sign in with your email — no credit card required. Compare plans.

Early Signal uses passwordless magic-link authentication. Enter your email address and we'll send you a sign-in link that's valid for 20 minutes. Click the link to access your account — no password to remember or reset.

Free gives you delayed snapshots and universe browsing. Operator unlocks real-time signals, 30 run credits per month, engine configuration editing, universe management, email alerts and digests, and CSV data export. See full plan comparison.

Composite signals are built from three timeframe components: Daily (momentum, options flow, news sentiment, earnings surprise, insider trading), Weekly (analyst revisions, sector rotation, supply chain indicators, technical levels), and Quarterly (fundamental shifts, valuation resets, CapEx trends, market share). Components are blended with layer-specific weights. See the full methodology.

Conviction measures the degree of agreement between daily, weekly, and quarterly signal components. HIGH conviction means all three timeframes align on the same direction. It reflects signal consistency, not a guarantee of returns.

Analysis runs can be triggered on three timeframes: daily, weekly, and quarterly. Each run produces fresh signals, recommendations, and risk assessments. The schedule is configurable by Operator-tier users.

The free-tier dashboard shows the most recent public snapshot, which is at least 24 hours old. This gives Operator subscribers a real-time advantage while still providing free users with meaningful market intelligence.

Every stock in the universe is assigned to one of ten layers that trace the AI value chain: Raw Materials (Layer 0), Semiconductor Manufacturing (1), Chips & Accelerators (2), Infrastructure (3), Cloud & Compute (4), AI Platforms & Tools (5), Foundation Models (6), AI Applications (7), AI-Enabled Transformation (8), and Secondary Beneficiaries (9). Browse the full universe.

Strong signals in one layer propagate to connected layers with 60% attenuation per hop, up to 3 hops maximum. Propagation lags range from 7 days (foundation models to downstream applications) to 90 days (upstream commodity layers). This models how supply chain dynamics ripple through the AI ecosystem. Learn more.

The engine monitors supply chain constraints — capacity utilization, lead times, and inventory stress — across all 10 layers. When a bottleneck reaches CRITICAL or MODERATE severity, it identifies which tickers are affected and which might benefit. View current bottlenecks.

Positions are allocated across three sleeves: Core (60% target, high-conviction quarterly-driven, up to 15 names), Satellite (30%, medium-conviction weekly, up to 25 names), and Tactical (10%, daily short-term, up to 20 names). Position sizing blends Fractional Kelly, volatility-adjusted, risk-parity, and signal-proportional methods. See methodology details.

Four predefined scenarios: Rates +100bps (growth multiple compression), Taiwan Supply Shock (foundry disruption), Hyperscaler CapEx Cut (AI infrastructure spend reduction), and Valuation Compression (broad de-rating of high-multiple AI stocks). Each shows estimated portfolio impact by layer. Run a stress test.

Yes. The personal portfolio feature lets you add holdings manually or via CSV import, track cost basis and P&L, and run overlay analysis that maps the engine's recommendations to your specific positions. Go to My Portfolio.

Early Signal uses market price data, quarterly financial fundamentals, news sentiment, options flow, analyst estimates, and proprietary supply chain indicators. Data quality is graded (A/B/C) and displayed on every run.

When live data feeds are temporarily unavailable, the engine falls back to statistically modeled synthetic data. This is clearly labeled on the dashboard and in every run summary so you always know what you're looking at.

The complete methodology — covering the 10-layer taxonomy, signal architecture, cross-layer propagation, portfolio construction, risk management, backtesting, and stress testing — is published at earlysig.com/methodology.

Operator subscribers receive 30 run credits per month for triggering daily, weekly, or quarterly analysis runs. Additional run packs can be purchased from your Account page. Unused base credits reset at the start of each billing cycle.

Yes, you can cancel anytime from your Account page. Your Operator access continues through the end of your current billing period. There are no cancellation fees.

Operator users can export CSV files from any analysis run (signals, recommendations, portfolio weights). You can also download all your personal data from the Account page under "Download My Data."