Educational market research
We write down what the crowd thinks — then check whether it was right.
Each market day we score the US stocks people are actually talking about, publish those calls before the open, and then check them against what the market did.
This is educational research, not investment advice. Nothing here is a recommendation, no account creates an advisory relationship, and we never place trades or handle money.
Why this is worth reading
Signals are recorded before the outcome
Every research signal is written down and timestamped before the market opens. There is no way to quietly revise it afterwards, which is the whole point.
The misses stay on the record
Calls that went the wrong way are kept and counted alongside the ones that went right. A research record that only shows the wins tells you nothing.
The method is written down
The scoring formula, the inputs, the assumptions under test, and every change to them are documented and versioned, so you can judge the method rather than trust it.
Assumptions get killed in public
Each assumption has a stated condition that would disprove it. When the data meets that condition we record the assumption as dead rather than reinterpreting it.
What members get
The full ladder — the ranked companies, their scores, and the outcome history behind them — is for signed-in members. What follows is the method itself, which we publish in full whether or not you ever create an account.
- —A daily ranked list of the most-discussed US stocks, split into the strongest positive and negative readings
- —A score per company built from discussion volume, how one-sided that discussion is, its direction of travel, and the quality of the sources it came from
- —Market-cap segmentation, so a microcap reading is never pooled with a megacap one
- —Realized outcomes attached to each past call at several horizons, measured against the same window for the S&P 500
- —A per-source accuracy ledger that tracks which feeds have actually carried signal
The scoring formula, its inputs, and every version change are documented in the methodology, which we publish whether or not the results flatter us. Past readings do not predict future ones, and the record includes calls that went the wrong way.
Why we built it
Financial commentary is graded on how convincing it sounded, never on whether it was right. We wanted to know whether the loudest opinions about a stock actually carry signal — so we built something that writes its research signals down first and grades them against what the market did afterwards, including when the answer is no.
More about Sentifyx and the people behind it →See the daily record
Create an account to see today's ladder and the outcome history behind it.