Rules frozen.
Results open.

WIRE developed the model using 146 headlines and validated it on 576 previously unseen headlines. After validation, WIRE finalized the sizing equation, rules and passing criteria, then collected approximately 4,600 unseen events spanning the initial 2.42-year evaluation. That development history remains unchanged. The results presented here come exclusively from the replacement direct API replay.

This record is a single algorithm replay through the model API used by WIRE's live engine. WIRE reports the historical sample was fully out of sample. The test window runs from Jan 4, 2024 through Sep 11, 2026; the last recorded trade exits Sep 9, 2026. The balance is held flat through the window end. Earlier backtest outcomes are excluded from this record.

The displayed +4727.87% cumulative return includes a fixed 0.10 percentage-point modeled deduction from each directional trade return. The supplied net returns already contain that deduction, which is not charged twice. Raw pre-cost return is +6423.13%, compounded using the same weights. This cost model does not establish actual commissions, slippage, financing, borrow costs or executable fills.

The account starts at 1. Historical positions are capped at 50% each. All positions sharing an exit morning are scaled equally if their combined gross allocation would exceed 150%. We reconstruct unrounded weights from the supplied equation, multiply each by its net trade return, and compound the combined contribution once per exit date across 359 sessions. Gross allocation can exceed account equity, implying leverage. Drawdown is measured at recorded exits and can understate intraday risk.

Live execution uses a different activity input and sizing rate, sizes positions as signals arrive, and stops accepting entries once its gross allocation limit is reached. It also limits each order relative to available broker buying power. The backtest instead allocates across the entire exit-date batch, using the final number of trades. These differences can change which trades execute and their sizes. Direct API calls do not make this an exact replay of live portfolio execution.

The table and download retain the supplied entry and exit prices, net trade returns and outcomes. Position weights and account contributions use the unrounded equation. In the download, portfolio_equity is the recomputed curve. Internal source identifiers and redundant source balances are excluded. Small rounding differences from the source do not affect the displayed cumulative return. Trade win rate and profit factor use unweighted net trade returns.

Net CAGR annualizes final equity from Jan 3, 2024 close through Sep 11, 2026, using 365.25 days per year (2.69 years). Calmar divides that CAGR by the magnitude of maximum recorded nightly drawdown. The day-clustered t-statistic uses an intercept-only CR0 estimate of mean net trade return, clustered by exit date without a small-sample correction. It accounts for within-day dependence, not dependence across days or future performance.

This is a WIRE-supplied historical simulation, not live member performance or an independent audit. Historical prices, modeled costs and batch sizing cannot reproduce live execution. Hypothetical performance does not predict future results. Stocks alerts are part of the WIRE Alerts membership. No orders are placed in member accounts. Read the risk disclosure.