Sentimentor uses optional analytics cookies to understand visits and improve the product. Essential cookies remain enabled either way.
Privacy policyFull Forward is Sentimentor’s validated forward track. Selecting Validated requires a server-verifiable protocol-v2 pass through discovery, rolling validation, embargoes and the final lockbox, then runs that exact frozen PineNode strategy on bars that arrive after activation. Users without a matching pass can still start a clearly labeled exploratory runner. Either evidence mode can record modeled trades inside Sentimentor or, when eligible, submit paper orders through a connected Alpaca Paper account. Broker-integrated with Alpaca Paper today. Live-money trading is planned for December 2026, with additional brokers to follow. The historical backtest remains a separate baseline.
The live runner verifies the final-lockbox record on the server, then uses the same saved code, symbol, timeframe, parameters, capital, sizing, costs and execution settings. Material code or configuration changes invalidate the pass and require a new validation record, because changing rules after seeing outcomes would contaminate the test.
Active ticker runners are evaluated independently as new bars become available. Exchange-traded assets follow applicable market hours while supported crypto markets can run continuously. Each evaluation records timing, direction and the strategy’s stated reason.
The history includes completed trades, entry and exit timestamps, prices, holding period, profit or loss and signal explanations. Sentimentor-modeled fills and Alpaca-reported paper orders remain separately labeled so a simulation is never presented as broker execution.
Forward testing is valuable when it reveals divergence: fewer signals, worse fills, larger drawdowns, regime dependence or a thesis that no longer works. A short or inactive record is inconclusive, and favorable results do not guarantee future performance.
A forward record becomes informative gradually. Review whether the live runner received enough new bars, generated the expected number of signals and encountered market conditions comparable with the original test. Compare fills, drawdown, turnover and benchmark-relative return with the frozen backtest, but do not overinterpret a short window or a small number of trades. If the rules or configuration change, evaluate the revision as a new version so later performance cannot be blended into the earlier record.