Research
Stochastic events — when AUC 0.85 is an artefact
On EURUSD, raw stochastic crosses inflate the model. After fixing sampling, AUC drops: weak regime filter, not a strategy.
- process-first
- sampling
- stoch
- eurusd-lab
Thought notes and research write-ups — eurusd-lab series and methodology notes.
On EURUSD, raw stochastic crosses inflate the model. After fixing sampling, AUC drops: weak regime filter, not a strategy.
Amplitude-based segments on EURUSD TIMB: D(h) peak +0.55, C4 stack permission R=0.56 valid → 0.34 OOS 2024.
A 3-state HMM denoises multi-horizon OLS slope. Stable soft-fade structure, useful mid-episode hazard — no entry gate or net EV.
On EURUSD TIMB, short vs long trend conflict exists and predicts OOS — but net EV collapses after costs.
V1 std_logp onset is lagged; V1b restores structure. Incremental vol vs slope+ER ≈ 0 — context, not a trigger.