QualityDip
Get alerts

Public track record

The QualityDip Ledger

Every alert, logged the moment a rule fires: the date, the price, what triggered it, and how it has done against the S&P 500 since. Wins and losses. Nothing backfilled, nothing deleted.

Started Nov 1, 2026 · prices as of Fri Oct 30, 4:00 PM ET close · page updated Nov 2, 4:00 AM ET

Alerts2
Average return—
Average vs SPY—
Beat SPY—
Median—
Worst—

Returns start with the first close after entry (Nov 2). Until then every return is zero, so we show a dash.

Alerts, newest first

ANETArista Networks
QD-0002
Rules v1

Closed 13.3% below what the market has paid since earnings on Aug 4 (Market Value $194.71, 62 trading days).

Entry · Oct 30 close$168.86
SPY entry$755.72
Return—
vs SPY—

Tracking starts with the Nov 2 close.

The 8-K behind Market Value · ANET page
Logged Nov 1, 2026, 9:48 AM ET · fingerprint 1f67a0071a71

AVGOBroadcom
QD-0001
Rules v1

Closed 7.0% below what the market has paid since earnings on Sep 2 (Market Value $350.52, 41 trading days).

Entry · Oct 30 close$326.08
SPY entry$755.72
Return—
vs SPY—

Tracking starts with the Nov 2 close.

The 8-K behind Market Value · AVGO page
Logged Nov 1, 2026, 9:48 AM ET · fingerprint ec697db564a1

The rules

Rules v1 · effective Nov 1, 2026 · fixed in advance

  1. Universe: the companies on the QualityDip watchlist: AVGO, AMZN, MSFT, KLAC, AMAT, ANET, COST, V and BE (BE: founder holds shares).
  2. Trigger: the daily close is at least 5% below its Market Value, and Market Value has had at least 5 trading days to form since the last thesis-changing filing. Also tagged if the close is below the DCF value or the FCF value.
  3. Entry price is that day’s close. The benchmark entry is SPY’s close the same day.
  4. Cooldown: at most one alert per company per 30 calendar days, unless a new thesis-changing filing resets its Market Value.
  5. No exits are called. Returns are tracked since the alert, plus fixed marks at 30, 90 and 180 days.
  6. Changes: a rule change creates a new version. Old alerts keep their version. Nothing is deleted. A correction is a new row that points to the old one.

Reversal Signal Rules v1 (published as “Bottom Rules v1”) · effective Nov 1, 2026 · fixed in advance

  1. Universe: every US-listed stock worth $20B or more that passes four quality checks: it generates cash (positive free cash flow, or operating cash flow of at least 20% of sales), earns a good return (ROE of 12%+ or ROIC of 10%+), sales are not shrinking year over year, and debt is manageable (net debt under 3× EBITDA; banks and insurers exempt). 333 companies today; we also track 667 more for reference (the next-largest US-listed companies, popular names and reader requests), which never trigger alerts.
  2. It must already be a dip. Dip watch: 15%+ below its 52-week high and under its 50-day average, or 7%+ below its Market Value. Deep dip: 25%+ below its 52-week high and under its 200-day average, or 20%+ off the high and 12%+ below Market Value.
  3. R1 Capitulation reversal: a capitulation day (volume at least 2× normal, down 2.5%+, a new 20-day low) in the last 10 sessions, no lower low since, and price trades above that day’s high. Alerted intraday the moment it happens. Entry is the alert price. Stop: below the capitulation day’s low.
  4. R2 Reversal candle at value: a bullish engulfing or morning-star candle at a 20-day low, while the stock is in a Deep dip or 5%+ below its Market Value. Judged on the final daily close. Entry is that close. Stop: below the pattern’s low.
  5. R3 Oversold momentum turn at value: the MACD (12, 26, 9) crosses up while it is deeply negative (below −2% of price), with the stock in a Deep dip or 5%+ below its Market Value. Judged on the final daily close. Entry is that close. Stop: below the 10-day low.
  6. One alert per company per rule per 20 trading days. Returns are tracked against SPY from the alert like every other Ledger row. A stop level is information, not an order; no exits are called.
  7. Backtest, 2017–2026, these ~330 companies: the three rules caught the actual low (no lower low in the next 10 trading days) 53–73% of the time, against about 42% for simply buying the first day of a Deep dip. On 60-day returns against SPY they did not clearly beat that simple baseline (+3.7% to +4.5% vs +3.6% on average), and the test is flattered by survivorship: these are today’s strong companies. The rules are about timing and a defined stop, not a promise of profit.

Why you can trust the record

  • The rules are fixed and public, and every alert is logged under the version it fired on. Nothing is hand-picked.
  • The Ledger began on Nov 1, 2026. The first rows are what the rule caught on the latest close that day. There is no history from before that.
  • Every row carries a SHA-256 fingerprint of its data and the previous row’s fingerprint. Change any past row and every fingerprint after it breaks. The latest is in the footer; the raw data is at /ledger.json.
  • When an alert is posted publicly, the post’s link is added to its row as an outside timestamp.

Paper-tracked, not real trades. Returns exclude dividends, fees and taxes. Past results don’t predict or guarantee future results. Information only, not investment advice.

Get the texts