Live experiment · 0 RLTW positions
Live holdings
Three public books: RLTW (ML), EMC (our momentum book, from 31 Aug 2026) and corrected MC (their published holdings). EMC and MC: one decision a week, Friday close → next session open. RLTW may ask for an action any market day (morning mail, 20-day min-hold).
Live RLTW v3 · 0.19% fee
Seeding in progress: the book starts from today's Top 8.
Live EMC · from 31 Aug 2026 · 0.19% fee
One decision a week: Friday close → next session open. The NEW labels below are the 31 Aug seed, not a rotation today.
EMC seed: Friday Top 8, Monday start.
Corrected MC · 0.19% fee
MC snapshot not available yet.
Notes
- Rank is not the same number across the three books. RLTW uses the ML ranking (RLTW table: buy_priority after quality gates). EMC uses our 13-week momentum (EMC ranking: 60% Stage 2 + 40% last 4 weeks). MC uses the published MomentumCommand ranking.
- OUT / BUFFER ≠ sell today. RLTW keeps a name for 20 days (min-hold). After that it sells if rank > 14 or the name is no longer eligible. A holding can be the best P&L in the book and still be OUT: the model does not keep winners because they already went up.
- Calendar is a different job. RLTW may send a buy/sell any market day (morning mail, 20-day min-hold). EMC and MC take one decision a week: Friday close → next session open (Friday-night mail). Mid-week the momentum books stay put. If Monday is a US holiday (Labor Day 7 Sep 2026) the fill moves to Tuesday.
- Status. NEW = bought in the last 5 days — it is not a new rotation today. The EMC NEW names are the 31 Aug seed (see EMC rotations). HOLD = rank ≤ 14. BUFFER = rank > 14, waiting for min-hold. OUT = failed the ML gates (not in the ranking at all).
Modello di ricerca, non consulenza finanziaria. RLTW, EMC e il libro MC corretto applicano la stessa commissione 0,19%/lato su ogni rotation. EMC è live dal 31 agosto 2026. MomentumCommand sul loro sito resta a 0 fee: qui lo ricalcoliamo con i nostri costi, sugli holding che pubblicano. I rendimenti partono dal giorno di avvio del libro pubblico, non dal backtest storico.