Changelog

A dated record of every change to the rule, the schedule and the site. If something changed what the screen admits or when it runs, it is listed here, so a result from a given date can be read against the rule that was actually in force.

Last updated By Izaias Cavalcanti

Anything that changes what the screen admits, when it runs, or what the published figures mean is listed here with the date it changed. A result from a given date should be readable against the rule that was actually in force on that date.

Cosmetic changes to the site are not listed. This is a record of the rule, not a release log.

2026

  • Signal desk moved to its own address at signaldesk.izaias.xyz. The previous address continues to work.
  • This learn section, the methodology page and this changelog were published. The full rule had been in force for some time; it had not been written out publicly in one place before now.

  • Price refresh reduced from two runs a day to one, now at 21:20 UTC on weekdays. The morning run had no stable meaning: at its scheduled hour both markets were shut, so it re-wrote the previous close that the evening run had already recorded, while making the page claim prices had just refreshed. The new time clears the closing bell in both markets under either daylight saving regime, so it cannot land mid session.

  • Both business filters became a hard entry requirement. Previously they were checked when a signal fired; now a stock cannot join either watchlist at all unless it passes them on freshly fetched data. The signal time check was kept as well, so staleness is still caught.
  • The short term trend check now fails closed. If the price history needed to compute it cannot be retrieved, the stock is rejected rather than admitted. It previously failed open, which meant a stock could be admitted on a check that had silently not run.
  • The monthly scan's last day of month guard was re-enabled, so the job is a no-op on any day that is not the final trading day of the month.

  • Filter B was rewritten. The rule became: at least two of BlackRock, Vanguard and State Street must have increased their position and the third must not have reduced it. Any of the three missing from the data counts as a failure. The previous rule required specific named institutions to have increased, which rejected almost everything for reasons that were about data reporting rather than about the stocks.

  • Filter A's source changed. The growth forecast is now the average of the current year and next year consensus earnings growth estimates. The previous source had become unreadable. The threshold stayed at 10 percent a year, but the figure it is applied to is measured over a different horizon, so readings before and after this date are not directly comparable.
  • Existing watchlists were re-tested against the current rules and entries that no longer passed were deactivated rather than deleted.

Keep reading

MethodologyThe complete public rulebook. Which stocks are scanned, the exact StochRSI computation, both About Signal deskSignal desk is built and run by Izaias Cavalcanti, a product designer and design engineer. It LearnPlain explanations of the ideas Signal desk runs on. What StochRSI measures, what a reading of

Never find out three days late.

The signal scan already runs every weekday at 22:10 UTC, whether anyone is watching or not. Premium is how it will reach you.

Free while in early access. No card, no spam.

Signal desk is a public record of a mechanical screen. It is not investment advice and it is not written by a licensed financial adviser. The percentages shown are plain price changes with no positions, sizing or costs behind them, so they are not a real world return.