Hi fellow traders,
Wanted to share some empirical findings and execution data from running automated systematic option selling on NIFTY 50 weekly contracts via Kite Connect API v3 hosted on an optimized Linux VPS.
When transitioning from discretionary trading to fully mechanical rules, there are three structural realities that theoretical models often miss:
1. The Decoupled Stop-Loss Problem in Delta-Neutral Baskets
Most retail traders run straddles/strangles with combined basket stop-losses (e.g., -20% or -30% of total premium). In live markets, combined stops often get whipped out during transient volatility spikes.
We tested decoupled stop-loss triggers per leg:
- Entry: 10:00:00 AM IST (after initial 45-min price discovery).
- Candidate strikes: 40% ATM sweet-spot premium ratio.
- Each short leg is monitored independently with a 1.75x individual stop-loss, while the opposite side is left open to capture theta decay until the 15:15 PM EOD cutoff.
- Sub-25ms order routing ensures the winning leg is never prematurely closed during one-sided trending days.
2. The Expiry Margin Reality on Indian Brokers
There is a common misconception that 1 lot of Nifty option selling can be safely run with ₹60k–₹70k capital due to hedge margin relief.
In practice on live broker terminals (Kite/Dhan), expiry day gamma surcharges and leg-sequence order checks require a realistic buffer of ₹1.40L to ₹1.50L per lot (65 Qty) to eliminate RMS rejections and avoid broker auto-square-off penalties. Running undercapitalized is the single biggest operational trap for systematic option sellers.
3. Execution Latency & VPS Infrastructure
Running automated execution locally over domestic broadband introduces 300ms–800ms socket latency. Running an asynchronous Python daemon on a dedicated Linux VPS co-located near exchange data hubs reduces socket response times under 25ms, which makes a measurable difference in slippage during 10:00 AM entries.
Open Telemetry & Live Monitor
We believe transparency is the foundation of systematic trading. Our live terminal runs continuously during market hours with real-time IST timestamps, live tick telemetry, and full audited session ledgers:
Live Terminal Monitor: http://45.129.86.80/selling
Architecture & 48-Hour Paper Gateway: http://45.129.86.80/portal
Would love to hear how other Kite API developers handle decoupled leg stops during extreme trending days, and what slippage thresholds you model for Nifty options.