Lab: our own testing, in public.
A running log of how Cephic tests Watchdog and Allocate on accounts it runs itself. Live data tracked by Cephic, published as it happens, including the tests that go wrong.
What the lab is
The methodology page says how Watchdog's rules are meant to behave. The lab shows how they actually behaved on real accounts that Cephic runs itself. Each entry is one test: a strategy and its backtest connected to Watchdog, sometimes with a deliberate fault added to see whether, and after how many trades, Watchdog notices.
Every entry, including the failures
- What was tested, and why.
- The setup: platform, strategy, backtest length and lot sizes on both sides.
- What was changed on purpose, if anything, and at which trade.
- What Watchdog flagged, and when: the status at each step and the trade count at which it changed.
- What failed or surprised us, including mistakes we made ourselves and alerts that came too late, too early or not at all.
- The numbers in a table, next to dashboard screenshots, so you can check them against the methodology.
- What we changed afterwards, if anything.
Tests that make Watchdog look bad are published the same way as the ones that make it look good. Account numbers and login details are never shown.
Entries are coming
There are no entries yet. Nothing is published here until the test behind it has real data, so this page stays empty rather than showing examples.
Questions about a test you would like to see run? Email support@cephic.io.
Analysis of your own data, not financial advice, no guarantee of future results.