Position trades are found on weekly charts.

Every Sunday a screen sweeps US stocks for a narrow profile: liquid names trading 45–90% below their all-time high, still underwater on a four-to-five year basis — yet already 45%+ off their low, with medium-term momentum that has turned positive and relative strength showing up on at least one timeframe. In short: long downtrend, bottom in, turn under way. That is the universe — typically 30 to 50 names.

A screen can't read structure, so every candidate is re-examined on two years of weekly bars. Four things must be on the chart:

• A volume surge — one week whose volume exceeds the prior six months' maximum, closed strong. Institutions cannot accumulate quietly; this is their footprint.
• A higher low — after the major low, a confirmed swing low above it. Sellers are done; pullbacks now find buyers earlier.
• The first moving-average hold — the 10- or 30-week line that rejected every rally during the downtrend gets held for the first time. Resistance becoming support is the single strongest signal in the system and defines the entry.
• Quiet action — contracting weekly ranges and drying volume near the level. We buy weakness at support, never strength at highs.

No qualifying surge in 52 weeks, lower lows still printing, or no defensible stop — the name fails outright, regardless of how good the story sounds.

Every pick must offer a structural stop — a shakeout wick low, the just-reclaimed moving average, or the prior week's low — within 5% of the entry zone, ideally under 3%. Position size follows mechanically: fixed dollar risk divided by stop distance. A tighter stop buys a larger position at identical risk. If no such stop exists, the setup is not buyable yet and goes on watch instead — no exceptions.

Institutional rotation hits groups, not single tickers. When two or more screen hits share an industry, each gets a ranking bonus — the cluster itself is evidence. A lone name can still make the letter, but it ships with a visible flag: rotation unconfirmed.

The top three survivors make the letter — each with its entry zone, stop, stop distance, share count at a reference risk, the full evidence trail, and an if/then plan for Monday. The cap is hard. When only one name qualifies, you get one. When none do, you get the market read and a watchlist — a week with zero picks is a signal, not a failure.

Everything in the letter works off weekly closes. Alerts do the intraweek watching; you act when a level is reached or a week closes. Stops are evaluated on a weekly closing basis, trends are trailed with the 4-, 10- and 30-week averages, and holding periods run weeks to months. It is a system built for people whose day job is not watching ticks.

Each published entry is recorded the moment it ships and scored against real prices at 4, 8, and 13 weeks. The full history — wins, losses, and no-pick weeks — will be public from the first paid issue. No cherry-picked screenshots.