Posted in My Daily Journal · 15 July 2026
Six stock strategies on 48 NIFTY-50 names and a 275-stock universe, the first live day of the paper fleet, and the discovery that two-thirds of the straddle's profit comes from one day a week.
Does the index pivot system work on individual stocks?
No, decisively. The live configuration on 48 NIFTY-50 stocks produced 30,286 trades with negative expectancy in every period — −9 to −17 basis points a trade against +35 to +40 for the index — and not one stock was positive in all three windows. The index is a noise-cancelling machine: fifty idiosyncrasies average out so a pivot break is a regime signal; a single stock keeps its noise and the 0.30% box filters nothing at 2% daily volatility. Also a moat: the rules do not port.
What about holding stocks overnight, buying dips, breakouts on volume?
The hierarchy that came out of the night, on eight years and real costs: momentum rotation (monthly rank by six-month return, hold the top 10–20) +41% CAGR at −31 to −34% drawdown, about double the honest survivorship-matched benchmark, with an 18-month "momentum winter" in 2025–26 as the warning; volume-spike breakouts +13.9% CAGR at −47%, a real convex edge strictly dominated by plain ranking; RSI-2 dip buying +4% CAGR, a real phenomenon eaten by 1,900 trades a year of friction; cash BTST −72% over eight years because delivery tax exceeds the 0.23% nightly edge; the pivot negative. The law: stock returns scale with holding period and die with transaction count.
Opening-range breakouts on stocks?
The perfect null. Four variants on 48 stocks over two years all came to −7 basis points a trade — exactly the cost model, gross edge 0.0 — with two to four stocks positive and every year negative. The opening range carries no information on single names.
Where does the 9:20 straddle's edge actually live?
On expiry day. With a 25% leg stop it earned +₹689 per expiry day against +₹70 per ordinary day — ten times, 64% of the total profit from 15% of the days. The contrast holds at a 50% stop and even without a stop (positive on expiry days, negative on ordinary ones): a pure settlement-day theta collapse, positive in every weekly-era year from 2019 to 2026, with an expiry-only maximum drawdown of ₹14k and worst day ₹7k — a profit-to-drawdown ratio around 16, the best in the programme. The "hero-zero" idea of buying ₹5–20 options at 14:00 on expiry was falsified alongside it: 10–24% win, −₹180 per expiry forever, best jackpot in eleven years ₹15.9k.
What did the first live day of the paper fleet teach?
The straddle paper bot made its first slippage measurement on a stopped leg — +0.42 points, 0.3% of entry — the one number the backtest assumes. It also made money on the exact day the directional books bled, which is the negative-correlation thesis visible on day one. And a deployment was missed: an upgraded bot never reached the trading machine the night before and the old version traded a day in paper. Lesson written down: a boot check the morning after every change.