The asymmetry nobody in this market will tell you about
Every PSX educator publishes what worked. Almost none publish what they tried and abandoned, because a list of failures reads like an admission. It is the opposite: a tested failure is a permanent, transferable result, while a success is usually one sample of one method in one market character.
ZEME's own working notes carry 153 documented tested-and-rejected approaches. Each one built, measured against real PSX history, and killed by a specific number that is recorded next to it. This course teaches a selection of them.
The rule that makes them binding
Do not re-propose a rejected idea. Each was killed by a measurement, and a redesign will be tempted by at least six of them.
That rule exists because rejected ideas come back. They come back wearing new names, and they come back most strongly from smart people who have just thought of them for the first time. The only defence is writing down the number that killed them.
What you get out of this course
Not a strategy. Something better: a list of expensive experiments you now do not have to run. Five of the ideas in this course are things almost every PSX trader believes right now.
Evidence · History · Confidence · Counterargument
Evidence, 153 rejected approaches catalogued across the working notes, each with a named source file and the measurement that ended it.
History, accumulated over roughly a year of building, mostly on 10.5-year backtests.
Confidence, high that these ideas failed as tested, on PSX, on this data.
Counterargument. The honest limit: "rejected here" is not "impossible everywhere". Several of these were tested without trial-count adjustment, deflated Sharpe or PBO, so the testing regime itself has documented gaps. A rejection is strong evidence, not proof. Where we know the test was weak, the lesson says so.