These are ZEME's own engines. Each one computes something specific, daily.
Coiled Spring™
Finds stocks sitting in the same quiet, compressed state that preceded their own past upper circuits. The die → drift → fire pattern. It only qualifies stocks with a personal history of firing from that state. Covered in full in The ZEME Toolkit.
Stealth Accumulation™
Quiet loading with no price signature, volume arriving without the price responding.
Market DNA™
Fingerprints today's session on 12 measured components (trend, breadth, circuit locks, participation) and finds the closest-matching sessions in 10.8 years of history, then shows what actually happened next. Walk-forward validated with a 25-session embargo: correlation 0.195 against −0.066 for the base rate, and a positive tercile spread in 9 of 9 years.
You can use it right now: Market DNA.
Lock Radar™
Forecasts upper-circuit locks from 90-day lock history × recency × ignition.
Absorption Count™
Counts events where supply was taken without a price rise, heavy volume, little movement.
The honest part, and it is important: absorption measured negative as a standalone short-horizon buy signal, −0.79% alpha at 5 days, with decile monotonicity of −0.69, meaning it ran backwards. ZEME therefore shows absorption as evidence of where volume went, never as a reason to buy. A term can be real, computed, and still not be a signal.
Operator Radar™ and Exit Risk™
A 5-signal forensic engine, and the reversal/trap grader described in the previous lesson.
Do this now, practical activity
- Open Market DNA.
- Find today's closest match score and compare it to the 95th percentile figure shown beside it.
- If today's score is below that percentile, the page will tell you plainly that the analogues are weak evidence. Read that warning, then look at the forward table anyway and notice how easy it would be to quote those numbers without it.
Measured: Market DNA walk-forward, 25-session embargo, predicted-vs-actual correlation 0.195 against −0.066 for the base rate, with a positive top-minus-bottom tercile spread in 9 of 9 years. Absorption, by contrast, measured −0.79% alpha at 5 days with −0.69 decile monotonicity.