Analyst · previewYou're seeing the public read on the BG2 cohort. Analyst unlocks per-host transcript receipts on every position + Let-It-Ride drawdown timeline + alerts when Gerstner / Gurley open a new position. Trader adds cross-pod join with All-In (who else said what about this name) + custom methodology BYOB + the API for pulling the live BG2 ledger.Compare tiers →
BG2 Buys · Gerstner + Gurley
Brad Gerstner's BG2 book. Bill calls, framed as portfolio.
46 BG2 episodes, 66 graded calls. The default below is BALANCED — the same published default every book on this site runs, so numbers never disagree across pages. But BG2's data-best methodology is Let It Ride: this book is dominated by long-runway AI compounders (NVDA, META, TSLA, AAPL) that punish early harvest — our trim sweep across 100+ configurations couldn't beat “no trim at all.” One click in the chooser shows the gap.
Data-optimal default: moderate stop, medium hold, stepped gain harvest.
The cohort numbers above are the result of running the engine with this preset's exact parameters (stop loss, hold window, gain harvest schedule, conviction floor, portfolio mode). The recipe itself isn't published. Subscribers can see the full parameter readout and replicate the methodology end-to-end.
BG2 Portfolio vs SPY — 586 trading days · raw engine book · $200K bucket
0 shorts · 3 expiries · matched-pairs SPY mirror
Host portfolio$NaNSPY mirror$NaN$NaN (NaN%)
Currently Open · BG2 Book — 1 positions · 50% return · BALANCED book
Top Holdings · 2 slices
NVDA$1,1510.4%
SPY (uninvested)$298,94499.6%
Trim-sweep verdict: no harvest wins. We ran 100+ trim configurations against Brad's BG2 book — gain-harvest milestones from +25% to +400%, sell amounts 15-50%, time-decay glides 3-9 months, all combinations. Every single one underperformed “no trim at all” (+25.18pp baseline). The closest harvest variant lost only 0.2pp, but the worst (time-decay combos) crashed to +5pp. Confirms what Brad and Bill articulate on the pod every week: let your winners ride.
Bill Gurley — 6 graded mentions, only 1 INITIATE. Most of his framing is bear-OBSERVE (Cisco analogy on NVDA, healthcare-tech skepticism) without explicit short language, so it doesn't survive into a tradeable position. He plays moderator + skeptic, not picker. Not yet a standalone book.
