How TrendWake’s nightly scan ranks stocks

We deleted the score

Until July 2026 this scanner graded every name 0–100 — a composite of Supertrend, RSI, market regime, volume and momentum. That score is gone, and nothing replaced it. Here's why, because the reasoning is the feature.

A composite is a number we invented. It looked precise, and it wasn't: it blended one strategy's opinion into a single figure and hid the only trustworthy fact — whether the rules that fired have ever actually made money — behind one we made up. It also hardcoded a bullish Supertrend gate, so the scan could never surface a breakout, a 52-week high, a dip, a meme squeeze, or anything at all on the short side. It was one strategy's view of the market, published as the house view.

What ranks a name now

Every night the scan runs our Daily strategies across the universe and asks one question per name: did a strategy fire an entry it would actually have taken? A blocked entry is a trade the strategy would have skipped, so it doesn't count. An exit isn't a candidate either. If a real entry fired, the name is on the list.

Then it ranks the list by the best track record among the strategies that fired on it — the lead strategy's profit factor, shown as itself:

1. $NVDA — LONG · 20-Day Breakout (PF 1.74) · fired 2 bars ago · +2 others
2. $TSLA — SHORT · Bear Trend (no track record yet) · fired 1 bar ago

Top of the list means the best evidence, not the best odds. It's the name whose firing strategy has the most convincing record — which is a different claim from "this will go up," and a much smaller one.

Three rules we won't bend

A losing record still prints. If the lead strategy's profit factor is 0.62, the row says 0.62. Hiding the bad ones would turn this into marketing.

No record is not a good record. Before a strategy can rank a name it needs a measured record, earned on its exact deployed settings, over at least 10 closed trades. Below that bar the row says "no track record yet" — in words, never a number. A brand-new strategy with two lucky trades cannot outrank a real book, and we won't invent a score to let it try.

Shorts say SHORT. Direction is spelled out in every row. "Beary fired on TSLA" must never be read as a buy.

Confluence is context, not the ranking

When four strategies fire on the same name, the row notes it — "+2 others" — and that's all it does. It never moves a name up.

Several of these strategies are trend-followers that fire together by construction. Ranking by agreement would sell you one idea four times and call it consensus. Four mediocre strategies agreeing is weaker evidence than one with a real record. So the count isn't in the sort — not even as a tie-break. The moment a count can move a row, it's a score again.

What it covers

Daily strategies only. Intraday setups keep the on-demand Scanner, which you run yourself against a strategy. The nightly list ranks by house and published strategies — a private strategy of yours is never read into it. Publishing is the consent.

The universe spans our curated liquid names plus the active watchlist, and includes crypto majors (BTC, ETH, SOL and friends) alongside stocks — crypto candidates carry their own asset tag, and the board can filter Stocks vs Crypto.

Using it well

• Read the strategy and its record first — that's the whole ranking. "20-Day Breakout, PF 1.74" tells you more than any grade could.
• Treat "no track record yet" as exactly what it says: unknown, not bad and not good.
• Cross-check the chart (click any ticker) and the earnings date (Markets → Calendar) before you act on anything.
• A quiet night is a result. If nothing fired, the list is empty and we don't manufacture a name to fill it.

The scan's job is to narrow a universe down to a handful of names and tell you honestly how much the rules behind each one have earned the right to your attention.

Strategy records are simulated — from backtests and paper-forward runs, not live-money results — and include losing strategies. Hypothetical performance is not indicative of future results. Educational only — not financial advice.