Summary of key results · M15 Playbooks

Item	Formula or rule	Where
Position size	$Q = \ell\lfloor \lfloor fE_q/	P_e - P_s
Break-even win rate with costs	$p^* = (1 + c)/(1 + W)$ , $c$ in R	M04 (4.5)
Driftless stop-out probability	$2\Phi(-k/\sqrt n)$ , $k$ = stop in daily s.d.; $\mathrm{ATR} \approx 1.596\,\sigma S$	M04A (4A.9), (4A.12)
Cost hurdle intraday	per-trade Sharpe $c/\sigma_t$ ; drag $\sqrt N c/\sigma_t$	M05 (5.25)
Straddle and implied move	straddle $\approx 0.8\,F\sigma\sqrt\tau$ = expected absolute move	M09A (9A.8), (9A.9)
Event variance	$e = \tau_1\tau_2(\sigma_1^2 - \sigma_2^2)/(\tau_2 - \tau_1)$	M09A (9A.13)
Defined-risk margin	scenario margin $\le$ maximum loss	M09A Proposition 9A.12
Trades to detect an edge	$N \ge ((2 + z_{1-\beta})s/E)^2$	M03 (3.2b)

Playbook	Episode	Result (net)	What the wider sample said
15.1 VCP breakout	BEL, 2–16 Mar 2026	−₹9,490.97, −0.95R	7 signals, mean −0.21R, $t = -0.53$
15.2 Pullback in a leader	Coal India, 6–10 Apr 2026	−₹10,708.69, −1.07R	36 trades, mean −0.56R; random-entry null −0.33R
15.3 Intraday ORB	TCS, 28 Aug 2026	+₹1,925.07, +0.19R	56 trades, mean +0.17R, $t = 1.63$ , gone at ~19 bp slippage
15.4 Earnings straddle	TCS, 8–9 Oct 2026	−₹884.67, −0.14R	10 events, long straddle −0.36% of spot, $t = -0.92$
15.5 Weekly condor	Nifty, 28 Oct–3 Nov 2026	−₹4,204.13, −0.43R	one week; edge undetectable in fewer than hundreds of weeks

Lessons that hold across the five: write the rules before the scan; reconcile free data with exchange files; size from the stop or the maximum loss, never from conviction; record R after costs; compare every setup with a null; and count how many trades it would take to know.