Every strategy's rule, universe, entry, horizon and cost model is written down before it produces a signal. Signals are stamped with their session date the morning after the session they came from. Nothing is added or removed later.
Entry for scoring is the close of the first session after the signal. Each signal is scored against SPY over its pre-registered horizon (net of a 50 bp round trip where the rule says "net"). Losers stay on the table. The candidates table is the full list.
In trial: live signals being scored. Verified: n ≥ 100 settled signals across ≥ 4 calendar quarters, net-of-cost positive, reviewed once. Killed in backtest: failed its out-of-sample test; still published, never traded. Passed holdout: Track #0 only — the SPY state machine passed a 2017–2026 holdout.
Everything is free. Candidate names are visible the same morning to signed-in readers and five sessions later to everyone else. Registration exists so we can see which features are used; it costs nothing and needs no card.
No strategy here has a proven stock-picking edge yet; that is what "in trial" means. Market weather, gauges and base rates are context — they never change a rule. We never say how much to buy or sell.
AEGIS is an impersonal financial publication of general circulation — not personalized investment advice, not a broker-dealer or investment adviser. The operator may trade the strategies and names published here with personal funds, after publication. No personal holdings or trades are published anywhere on this site. Data comes from public filings, exchange data and third-party feeds that can be late or wrong; dates are shown so you can judge freshness. Not directed to residents of the UAE.