The research system

Continuous re-evaluation

Why buying is never a one-time decision, and how every position competes against every new opportunity.

Buying is not a one-time decision. Every market update triggers a complete re-evaluation of every open position. Holding requires continuous justification — a position stays open only while the expected reward remains superior to the expected risk.

The question asked at every update

“If we did not already own this token, would we buy it now?” If the answer becomes no, the default action is to prepare an exit.

The possible actions

Action	When
Buy	Reward clearly exceeds risk and there is room to add.
Hold	The thesis still justifies the capital committed.
Reduce	Conviction has weakened but not collapsed.
Exit	The reason to own it no longer holds.
Ignore	The opportunity is not worth the lab’s attention or risk.

Every position competes

Each open position competes against every new opportunity for the same capital. No trade is sacred. Capital flows to wherever the expected reward/risk is currently best — which is exactly why the lab favors short-lived, speculative opportunities over long-term holding by default.

Early-token positions move faster

Open early-token positions reevaluate far more often than mature ones, with a richer action set: HOLD · ADD · REDUCE · TAKE_PROFIT · EXIT · EMERGENCY_EXIT · HUMAN_REVIEW. Adding is treated as a new decision — never add merely because the price dropped.

Reevaluation trigger
Scheduled interval	Adaptive cadence — seconds early, then tapering
Price / liquidity threshold crossed	Sharp move or liquidity change
Creator-wallet movement	Creator selling or transferring
Whale distribution	Large holders exiting
Security-state change	New contract or honeypot signal
Failed sellability check	Exit may no longer be possible
Social-momentum reversal	Narrative fading
Exchange / chain incident	Venue degradation
Maximum holding time	Time-stop reached

Emergency exit is a first-class action

A failed sellability test, a creator removing liquidity, or a risk state of halted/emergency can trigger an immediate EMERGENCY_EXIT — independent of the score. Every exit rule is versioned, measurable, replayable and linked to the active strategy.