VK
VIKRAM KHANNA
Founder · Head of Quant Research
Most people see a trade result — a number on a screen. They don't see the machinery behind it: the signals that fired at 3am, the volatility filter that rejected eight other setups before this one, the risk engine that sized the position precisely. This article breaks all of it down. No fluff. Just the system.
THE FIVE PILLARS
AlgoVik's strategy is built on five non-negotiable pillars. Every trade that goes live must pass all five. A single failure is a hard veto — the algo doesn't trade, and neither do we.
Phase 01 — Market Scan
Every morning before market open, the system scans Nifty 50 structure across three timeframes: daily, hourly, and 15-minute. It looks for one thing — structural confirmation. Three consecutive closes above or below a key level tells the system there is a genuine bias. One close is noise. Two is correlation. Three is conviction.
3× Close Confirmation
3 Timeframes
4:30am Execution
Phase 02 — Volatility Filter
India VIX is the gatekeeper. The system refuses to sell premium when VIX is above 18 — the environment is too expensive to protect the long leg, and the spread math breaks down. Below 15, we're in our preferred zone: compressed premiums, stable structure, optimal risk/reward ratios. Between 15-18 is a conditional zone requiring additional confirmation signals.
VIX < 15 = Green Zone
VIX 15–18 = Conditional
VIX > 18 = Hard Stop
Phase 03 — Structure Selection
Once the directional bias is confirmed and VIX is cooperative, the system identifies the optimal spread structure. For bullish setups: Bull Put Spreads (sell higher PE, buy lower PE). For bearish setups: Bear Call Spreads (sell lower CE, buy higher CE). Strike selection targets a minimum 1.5:1 reward-to-risk, with the sold strike at least one standard deviation OTM from current price.
Min 1.5:1 R:R
1 SD OTM Short Strike
Weekly Expiry Focus
Phase 04 — Position Sizing
Risk is always defined before entry. The system calculates maximum loss as (spread width − net credit) × lot size. This number must never exceed 1.5% of the total capital base per trade. Lot sizing is then back-calculated from this constraint — not the other way around. Most retail traders size first and think about risk later. We do it backwards, and that's the edge.
1.5% Max Risk Per Trade
Risk-First Sizing
Defined Max Loss Always
Phase 05 — Execution & Exit
Entries are timed to the first 30 minutes of market open, after liquidity stabilizes. Exits are rules-based: (a) expiry — the preferred outcome, full credit retained; (b) 50% profit target trigger for early exit on favourable moves; (c) 2× premium stop loss for early closure on adverse moves. No discretionary override. The algo executes; humans observe.
50% Profit Target
2× Premium Stop
Zero Discretionary Override
THE SIGNAL ARCHITECTURE
Below is the complete decision flow that runs on every potential setup. Every node is a filter. The system exits early and often — most mornings, the correct answer is no trade today.
THE SPREAD STRUCTURE EXPLAINED
A Bull Put Spread is a two-leg options strategy. We simultaneously sell a higher-strike Put (collecting premium) and buy a lower-strike Put (paying premium for protection). The net credit is our maximum profit. The difference between the strikes, minus that credit, is our maximum loss — which is always defined before entry.
Example — May 24, 2026: Sell 23,800 PE @ ₹180.05 · Buy 23,300 PE @ ₹31.75 · Net Credit = ₹96.30/unit · Max Loss = ₹403.70/unit · Across 650 lots = ₹12,090 max profit / ₹52,481 max loss at risk
WHERE AI ENTERS THE PICTURE
The rules above are deterministic. What AI adds is adaptive intelligence — the ability to pattern-match across thousands of historical setups and adjust confidence scores in real time. The diagram below shows how the AI layer sits on top of the rules engine.
RISK ARCHITECTURE AT A GLANCE
Every element of our approach is designed around one principle: defined risk, undefined patience. We never take a trade where we don't know exactly how much we can lose. We never override the system based on gut feel. And we never chase a trade that the algo passed on.
The core insight: In options selling, your win rate matters less than your consistency of process. A system that generates 65% win rate with defined risk will outperform a discretionary trader with 80% win rate and undefined downside — every single time, over enough trades.
WHAT THE SYSTEM DOESN'T DO
As important as what AlgoVik does is what it refuses to do. The algo never chases momentum. It never holds through earnings events. It never sizes up after a winning streak (overconfidence bias). It never reduces size after a loss to "protect capital" (loss aversion bias). The position sizing formula is identical every single time. That's the point.
Behavioural edge: The single largest source of edge in systematic trading is not the strategy itself — it's the removal of the human nervous system from the execution loop. Discretionary overrides lose money over time. The data is unambiguous on this.
* This article is for educational and informational purposes only. Past performance is not indicative of future results. Nothing in this article constitutes financial advice. AlgoVik is a SEBI registered advisory. Strategy parameters are described in general terms and may differ from live implementation.