The questions a critical reader should ask.

No. Version 1 is an end-of-day structural test of the decision core. It examines consistency across many instruments and oscillators. Exact intraday ordering, realistic stops, costs, slippage and full APS integration belong to Version 2.

Andreas Plagge is a NinjaTrader® Ecosystem Vendor. In the vendor context, NinjaTrader® is presented as the preferred platform, NinjaTrader® Brokerage as the preferred brokerage provider and Kinetick® as the preferred market-data service. Plagge Consulting remains an independent third-party vendor. Full Ecosystem Vendor information →

Ninety-four real instruments were evaluated with sixty oscillators under the same underlying logic. 94 × 60 creates 5,640 individual tests. They provide breadth, but are not fully independent because markets are correlated, oscillators can be related and all tests share the same core architecture.

It is the number of winners divided by the number of losers and has no unit. 1.0 means equal counts; values above 1.0 mean correspondingly more winners than losers. It is explicitly not a profit factor and not a return. The specific documented Version 1 values are shown on the results page together with their limitations and the full hypothetical-results disclosure.

Monetary profit factor is gross profit divided by gross loss. It reflects the size of gains and losses, while the winners-to-losers count ratio reflects only their number. The two metrics can differ materially.

A control group helps determine whether an unusual result could arise from the test construction itself. Six synthetic random datasets with 360 tests were evaluated using the same underlying logic. The specific comparison values and their limitations are shown on the results page.

It translates market structure, analytical values and filters into a fixed, versioned decision. It does not determine later position size and is not intended to be autonomously rewritten by AI. Execution and APS risk governance remain separate layers.

APS is risk and money-management logic. It does not decide what or when to trade; it governs how much capital is exposed and how long increased exposure may remain active. It uses stages, high-watermark protection, warning states and tactical rules.

No. Its mechanics, development history, historical simulations and individual comparison paths are extensively documented. Its incremental contribution to the overall framework is intended to be remeasured under identical intraday execution and more realistic cost assumptions in Version 2.

Otherwise it would be unclear whether differences come from new technology or a domain-rule change. The new implementation should first reproduce the reference state. Only then should extensions be introduced one at a time.

AI is intended as an assistance layer for documentation search, explanation, quality checks and development support. Core domain rules remain versioned and should not be autonomously altered by a model.

No. They are sensible current working directions for compute and data layers. The durable requirements are reproducibility, testability, clear interfaces and operational robustness. An institutional partner may have different infrastructure constraints.

Instead of evaluating a system only on one fixed historical sample, rules are repeatedly assessed on subsequent data windows that were not used for the respective fitting or adjustment step. This helps test whether a finding persists outside its original development window.

Because limitations are part of a scientific and engineering result. Serious review requires knowing both what has been tested and what remains open. That matters more for an institutionally reviewable system than presenting a flawless marketing image.

Yes, as a planned additional validation stage. Once Version 2 is sufficiently defined and reproducible, a professional trading and research office is intended to operate an approved system version under real market conditions. Version, decision, order, fill, fees, risk and outcome should remain traceably connected. Such a live track record does not yet exist today as institutionally audited evidence and would not replace the preceding methodological validation.

It is a possible later commercialisation option, not the current core project. A simplified analysis and decision-support layer could be derived from the protected institutional framework. Features, pricing, distribution model and regulatory classification are not fixed today and would be reviewed separately before any launch.