Stock screeners with a public track record
Almost no stock screener publishes a track record you can check — most show a marketing win-rate or nothing per-pick. Tapeline publishes its daily top-10 picks, 907 checked picks over 96 market days, each recorded the day we publish it and checked against the S&P 500 (the 500 biggest US companies) the next trading day. Picks are not re-ranked or deleted, and corrections are dated. Losses are kept, and the whole record downloads as a spreadsheet (CSV) or raw data (JSON). This page explains what a real track record has to contain — so you can judge any screener’s claim, not just this one.
What a real track record must contain
Four tests. A screener that fails any of them is showing you a claim, not a record.
1. A dated entry for every pick
Not an aggregate win-rate figure — the actual list, with each pick stamped the day it was published. Aggregates can't be audited; a dated list can. If you can't see individual picks with their dates, there is no track record, only a claim.
2. Losing picks kept, and corrections dated
The misses must be present at the same size and weight as the winners, recorded the day they were published, never re-ranked or quietly deleted. If a recorded value is ever corrected, the correction has to be dated and explained where the record lives. A record that can change silently after the fact isn't a record — it's a highlight reel.
3. Every pick compared with the whole market
When the whole market is rising, almost every pick looks good. So each pick's move has to be compared with the S&P 500 (the 500 biggest US companies) over the same day. That shows whether the pick did anything the market didn't do anyway.
4. Raw data you can download and re-check
A screenshot proves nothing. The full archive should download as CSV and JSON so you can re-run the arithmetic against any price source you trust, and find any error yourself. Downloadable raw data is the difference between 'trust us' and 'check us.'
Tapeline’s record, in numbers
45.5% of our 907 checked picks did better than the S&P 500 the next trading day. The full record is public →
How the record compares
What a downloadable, loss-included record looks like next to what most screeners publish.
| Track-record test | Tapeline | Typical screener |
|---|---|---|
| Per-pick record, dated | Yes — each recorded daily top 10 (four sessions have none; they are listed on the scorecard) | Rarely published |
| Losing picks shown | Yes — kept at equal weight | Usually omitted |
| Picks never re-ranked, added later or deleted | Yes — and corrections to recorded values are dated (prices 25 Aug 2026; scores capped 15 Jun 2026) | No public commitment |
| Each pick compared with the S&P 500 | Yes — the next trading day, per pick | Seldom |
| Raw data downloadable | Yes — CSV + JSON, full archive | No |
Reflects publicly available information; a specific screener may publish more. See the transparency breakdown or the raw scorecard.
Common questions
Which stock screeners publish a track record?
Very few publish one you can actually check. Most show an aggregate win-rate or nothing per-pick, because a public record is a liability. Tapeline is built around publishing one: each daily top-10 pick, dated the day we publish it, compared with the S&P 500 the next trading day, losses kept, corrections dated, and downloadable as raw data.
Is a screener's win rate the same as a track record?
No. A win-rate figure with no picks behind it can't be audited — you can't see which picks, on which dates, against which benchmark. A real track record is the underlying dated list of every pick, wins and losses, that the number is supposedly computed from. Demand the list, not the headline.
Can I download Tapeline's track record?
Yes. The full archive is downloadable as CSV and JSON, with the methodology and the dated corrections attached, so you can re-check every entry against any price source you trust. Tapeline invites corrections — the point of publishing raw data is that anyone can find an error in it.
How far back does the record go, and how big is it?
It runs from 11 May 2026: 907 checked picks over 96 market days, and it grows with each trading day. Entries are not re-ranked or deleted. We have corrected recorded values twice, and said so: prices on 25 August 2026, and scores from 18 May to 12 June capped on 15 June 2026. No top 10 was recorded for 31 August, 2 September, 4 September or 9 September 2026. The counts on this page are read from the scorecard rather than typed in, so the scorecard itself is always the current figure.
How does the track record compare with the S&P 500?
45.5% of our 907 checked picks did better than the S&P 500 the next trading day. The full record is public at /scorecard.
- Do stock screeners actually work? — a five-step way to check whether a screener’s picks are worth acting on.
- The most transparent stock screener — what “transparent” actually means, line by line.
See the record, then the scores
Six-factor scores on the full scanner, on a 30-day Premium trial — $0 today. The downloadable record stays public either way.
Sign up →Not investment advice. Scores and the record are descriptive, rules-based information only — not a recommendation to buy or sell, and not indicative of future results.