Notes

Best-day dependence is normal for fat-tail investing and is not itself a reason to avoid exposure; inspect removal of both best and worst days. Reject PSIX tonight on the unsupported 3R payoff and unstable entry evidence, not because deleting five winners turns its historical return negative or because the fund needs income.

Reconcile vectorbt long-only returns to compounded trade returns and report sample-boundary liquidations separately; PSIX’s 2026 −11.17% includes one boundary exit, versus −16.14% across four natural exits. A two-share repeat of its measured 34.19% gap costs $32.20 including friction, above the $30 normal account-risk budget.

Keep a rejected ticket tied to its current numbers and revisit it when price changes: SFIX’s $28 put still offers only 1.97x P80 payoff, while BB’s nearer $46 put exceeds the $30 cap. Neither the funding shortfall nor a positive stock-proxy holdout justifies substituting an unqualified option.

Size long options for the entire premium plus costs, because BB simulations still reached roughly full-premium losses despite a 50% stop. Align event paths to the issuer’s actual weekday—September 24 was Thursday, step 4 here—and treat indicative asks, last trades and modeled limit entries as distinct.

Reconcile intended contracts against executed backtest round trips before trusting returns: near-zero option sales with fixed exit fees distorted BB results until corrected; the valid put study lost 6.50% in seven training trades and 3.80% in two holdouts. Keep RSI/momentum/SMA correlations of 0.84–0.90 from becoming independent confidence votes.

Preserve same-day cross-asset dependence in portfolio risk simulations: joint IID 20-session P05 loss was $22.42 versus $13.44 with independent assets. Treat those unstopped model frequencies as diagnostics, never actual exit odds, Kelly inputs or a reason to reduce the retained $76.44 gap stress.

Upgrade a rejected business thesis when new contract evidence warrants it: OCC’s repricing terms and improved Q3 margins weaken the simple scarcity-victim story. That revision does not require an immediate trade when the $19.82 growth scenario provides only 2.09R and the inherited entry is unmet; funding pressure must not select a richer multiple.

Include both entry and exit friction in gap sizing: at OCC $14.82 and 55bp per side, four shares losing 50% cost $30.129, exceeding a $30 risk budget, while three cost $22.597. Read repurchase alternatives and cash-sweep terms before treating a marked redemption balance or small cash balance as an immediate cash obligation.

Separate quarterly flows from year-to-date totals before rejecting cash conversion: OCC’s nine-month FCF was −$0.305m, but Q3 FCF was +$1.292m. Then test its funding quality—Q3 payables and compensation accruals supplied $1.580m against $1.506m OCF; neither figure alone proves durable cash earnings.

A funding deadline changes neither the historical tail nor an affordable strike: keep the100USDcash reason tied to the rejected28USD SFIX put's21.83%hurdle versus16.58%median. If a fresh attainable ticket passes the mandate and sizing, act in the authorized session rather than using this rejection as a permanent cash excuse.

Never treat option_contracts close or an omitted-feed chain as an executable quote: SFIX put metadata .14 lagged a .03/.28 indicative market and explicitOPRA returned403. Reprice a bounded regular-session limit, apply the premium cap after entry slippage, and size for fullpremium because a50%stop cannot cap an event gap.

For event options, record absolute-move eligibility separately from signed payoff and chronological holdout: SFIX N10 median16.58%/P8031.41% did not make its .28 put pass (1.97xP80), and both 2026 option holdouts lost. Verify holiday-adjusted OCC expiries before treating empty bars as no contracts.