Track Record

Quantustik publishes a public, auto-updating track record: every past signal and confidence band is scored against what actually happened once the forecast horizon has elapsed. Nothing is curated after the fact — misses are shown alongside hits, the same honesty standard applied on the calibration page.

What this page reports right now

Performance numbers are on hold We found data errors in how these trades were recorded: some entry prices were never actually tradable, and a signal re-issued on consecutive days while a position was already open was booked as a separate closed trade each time. Both defects can distort the result of an individual trade, not just a total built from many of them, so we've pulled every return figure (the summary numbers above and the per-trade result on each row below) until we can rebuild the book correctly. The rows themselves stay visible as evidence; we just won't compute or add up a result from entries we know are wrong. We'd rather show you nothing than show you something wrong.

Why this page is different from a “trust us” backtest chart

Most forecasting services show you a curated performance chart after the fact. This one is generated automatically: every signal is written down — entry, stop, and take-profit levels — before the outcome is known, and the record updates itself after every scan. Nothing here is hand-picked or edited after the trade closes.

Why the losses are on this page on purpose

The log covers every closed signal, full stops included, not just the winners. If the numbers look unimpressive some weeks, that is the point: an auto-updating record cannot quietly drop its bad trades the way a marketing snippet can. The worst calls are ranked and shown, worst first, so an embarrassing trade cannot be swapped out for a less-bad one.

Does this record demonstrate anything yet?

This is a simulated (paper) record. We make no claim that the model has an edge until the published cohort clears a pre-registered statistical bar. Small samples swing hard, so treat every number here as evidence accumulating over time, not a verdict.

How this differs from the calibration page

This page answers “did the direction and timing work out?” For the companion question — when the model says 90% confident, is it right 90% of the time? — see the calibration page, which reports coverage separately for each quarterly forecast start date, including the windows where the band missed badly.

Where the underlying data comes from

Live per-ticker scoring is served from https://quantustik.com/api/v1/signals/track-record.

Educational research only — not investment advice.