How is this different from a stock screener?
A stock screener filters stocks based on criteria you set — low P/E, high growth, etc. SynapseAlpha is a prediction engine that learns which criteria actually predict sector outperformance. It uses ~120 calibrated parameters across 8 signal groups (valuation, growth, momentum, financial health, earnings quality, quality, mispricing, and event exposure), and these parameters self-tune monthly based on measured outcomes. The system scores 564 stocks weekly across 11 sectors and 3 market regimes, producing a ranked list — not just a filter.
What does "sector-relative" mean?
We don't predict whether a stock goes up or down — that's mostly driven by the overall market. Instead, we predict whether a stock will outperform its own sector ETF over 12 months. A "Strong Buy" on a tech stock means we expect it to beat the Technology Select Sector SPDR (XLK), not that it will necessarily go up. This isolates genuine stock-picking signal from market beta. It also means our picks are useful in both bull and bear markets — in a downturn, a successful pick falls less than its sector peers.
Why a 12-month horizon?
Fundamental signals — valuation, earnings quality, financial health — need time to play out. Short-term returns are dominated by noise and sentiment. Our walk-forward testing across multiple market regimes shows signal strength peaks at 12 months, with supporting evidence at 6 and 24 months. We use 12 months as the anchor horizon (weighted at 0.45), with 6-month and 24-month horizons as calibration support. Every pick is also reviewed at 3-month and 6-month checkpoints to track early signal quality before the primary 12-month review.
How are predictions kept honest?
Every prediction is published in advance with an immutable record — once saved, it cannot be edited or deleted. Each weekly edition locks 10 picks with their exact pick prices, sectors, decisions, and confidence levels. Reviews happen automatically at 3, 6, and 12 months against actual sector-relative returns. All past editions, including picks that underperformed, remain permanently visible. There are no retrospective changes, no cherry-picking, and no hidden failures. The full track record is public from day one.
Can I trust a system with no track record yet?
You shouldn't blindly trust any system — including ours. The engine was trained on 644,831+ data points and validated using walk-forward testing across multiple market regimes, where top-ranked stocks outperformed bottom-ranked by ~8–10% over 12 months. But out-of-sample testing is not the same as live results. Forward tracking begins April 6, 2026 with full S&P 500 coverage. The first 3-month checkpoint is July 2026, the 6-month checkpoint is October 2026, and the primary 12-month review is April 2027. Until then, use SynapseAlpha as a research input alongside your own analysis, not as a sole decision-maker.
What do the decision tiers mean?
Strong Buy = highest conviction that this stock will outperform its sector over 12 months. Buy = solid conviction with good signal quality. Buy Small = moderate conviction, worth a smaller position. Tiers reflect confidence level, not expected return magnitude. A Strong Buy doesn't mean "will go up more" — it means "we're most confident this one beats its sector." Only 10 stocks are published weekly from a universe of 564 to ensure concentrated signal quality.
How are financial stocks handled?
Banks, insurance companies, and asset managers operate with fundamentally different balance sheets — they use leverage as a core business model, not a financing choice. Standard valuation metrics like P/E, EV/EBITDA, and debt/equity can be misleading for these businesses. As of April 2026, financials are included in the scoring universe with banking-specific features: price/tangible book value, net interest margin, efficiency ratio, return on tangible equity, and tier-1 capital ratio. These tailored metrics allow the engine to evaluate financial stocks on their own terms rather than forcing them into a framework designed for non-financial companies.
Is this free?
Free for all S&P 500 stocks: Quantitative scores, radar charts, category breakdowns, and feature analysis for every S&P 500 stock — updated weekly, served from our database. Free for weekly picks: The top 10 stocks also get a full AI deep-dive analysis with thesis, key drivers, risk factors, and narrative explanation. We plan to offer premium features in the future — on-demand deep analysis for any stock, daily alerts, and portfolio tools — once the system has a proven forward track record. We won't charge for predictions until we've demonstrated they work.