FAQ

Mechaniq is a systematic signal service for US-listed stocks and options. Every strategy is a pre-registered hypothesis: its rules and pass/fail gates are frozen before any data is touched, then tested out-of-sample across market regimes. Most fail — that record is public. The few that survive every gate go live, and members get the trade signal — symbol, level, entry and exit rules — by Telegram. No commentary, no opinions — just the tested trigger.

Screeners give you tools. Mechaniq gives you conclusions. A screener has 800 parameters and asks you to choose, configure, and interpret. Mechaniq picks the strategy, runs the scan, applies the regime gate, and delivers the result. The work has been done — you decide whether to take the trade. We are not a better screener; we are a different product.

Yes — and we publish the methodology. Every setup is validated through walk-forward testing across multiple market regimes before it can be promoted, with per-metric attribution (which components carry the edge, which are noise) recorded in the lab. We do not show CAGR screenshots without context. Every strategy passes through the same lab pipeline — and its full test history stays public on /lab — before it goes live.

A normal backtest tunes parameters on one dataset and reports the result on that same dataset — it tells you how well the strategy fit the past, not whether it will hold. Walk-forward splits the data into chronological folds, tunes on fold N, and measures performance on fold N+1 (which the tuner has never seen). If a signal looks great on the training fold but fails on the next, it was overfit. Walk-forward is the only honest test.

Markets behave differently in uptrends than in downtrends. SPY above its 200-day moving average = trend regime; below = defensive regime. Mechaniq's regime gate is a hard, mechanical rule: when the regime says sit out, the signals go quiet rather than fire into conditions where the edge historically breaks down. The signal knows when to keep its mouth shut.

By design. Mechaniq is a pure information service — we publish signals, you decide and place the trade yourself. We never see your brokerage credentials, never place orders on your behalf, never modify positions. Our own Tastytrade data connection is strictly read-only and runs against our own account, never yours. This keeps us out of regulated territory and keeps your account control with you.

No. The content on Mechaniq is informational and educational only. It is not personalised investment advice under the Dutch Wet financieel toezicht (AFM). You are responsible for your own trading decisions, including whether and how to act on any signal. If you need personalised advice, speak to a licensed advisor.

By Telegram. Members opt in per setup and get the trade signal — symbol, level, entry and exit rules — the moment a setup is live. The full public record (every setup, every gate, every failure) is on /lab, free; the actionable rules are the members' side. Optional email summaries too.

Mechaniq Pro is $45/month, or $450/year (2 months free — works out to $37.50/month). The first 14 days are a free trial on either plan — no charge if you cancel before day 15. Founding members (the first cohort) get a price-lock for the lifetime of their subscription. No hidden tiers, no upsells, no 'pro+' add-ons.

Yes. One click in the billing portal. Your access continues until the end of the period you've paid for, then stops. No phone calls, no retention scripts.

On purpose. Mechaniq starts the public clock at zero: every setup is recorded forward with its outcome on /lab, from the day it's pre-registered. We could show ten years of in-sample backtests — everyone does — but in-sample numbers prove nothing. The honest answer is to build the record in public, failures and all.

An engineer, not a trader. The habit that built it: you don't argue about what a system will do — you model it, make every assumption testable, and check it against reality. What survives is real; what doesn't is data. More on the /about page.

OHLCV bars from Tiingo EOD (Power tier) and Yahoo Finance (yfinance library) as fallback, fundamentals from yfinance with Finnhub as fallback, US-universe ticker list from Finnhub (~5,100 common stocks filtered to ~2,600 by ≥$5M average daily dollar-volume). No paid data feeds beyond Tiingo Power — the edge is the method, not the data.