Gamma Scalping NIFTY: Delta-Hedge Your Long Gamma

OBSERVED: Traders buy NIFTY straddles for "volatility" but bleed theta waiting for the move. Gamma scalping is the discipline that converts long gamma into small profits by re-hedging delta as price moves — you buy low, sell high, repeatedly.

SOURCE: Standard options Greeks math (Black–Scholes delta/gamma), applied to NSE NIFTY weekly options. Your nifty50-xgb-engine and dhan-options-analytics already pull live Greeks; this article frames the scalping loop around them.

DERIVED: A practical scalping rule + a walk-forward backtest design you can run on your own data.

1. The Core Idea

Long a straddle = long gamma, short theta. As spot moves:

• Delta goes from 0 → positive (up move) or negative (down move)
• You sell the moved delta (lock profit) and buy back when it reverts

Each re-hedge captures gamma × (move)² minus transaction cost. If moves are big enough, scalping pays the theta.

2. The Math (No Hand-Waving)

P&L per scalps ≈ 0.5 × Γ × (ΔS)² − θ × t − cost

Where:

• Γ = portfolio gamma (straddle gamma ≈ 2× single-leg)
• ΔS = spot move since last hedge
• θ = daily theta (straddle theta is high)
• cost = brokerage + spread per hedge

Scalping wins when 0.5×Γ×ΔS² > θ×t + cost. That needs realized volatility, not just any move.

3. NIFTY Implementation

1. Buy ATM straddle (next weekly or monthly)
2. Track delta from dhan-options-analytics (live Greek feed)
3. Re-hedge when |delta| hits a band (e.g. ±0.25)
4. Hedge with NIFTY futures or the underlying ETF (NIFTYBEES)
5. Close at expiry or when Γ collapses

4. Backtest Framework (Walk-Forward)

# pseudo: simulate scalping on historical NIFTY
for bar in bars:
    delta = straddle_delta(spot[bar])
    if abs(delta - hedged_delta) > band:
        pnl += -delta * (spot[bar] - spot[bar-1])  # hedge trade
        hedged_delta = delta
    theta_cost += straddle_theta[bar]
# walk-forward: retrain band per regime (see Walk-Forward article)

Use walk-forward (your corpus standard) — fixed band decays in trending regimes.

5. The Mistake Everyone Makes

• Hedging too often → cost eats gamma. Band too tight = death by commission.
• Hedging too late → missed the move, theta wins.
• Ignoring spread → NIFTY option spreads widen near expiry; scalping there is negative EV.

OBSERVED: Optimal band ≈ where 0.5×Γ×band² ≈ 2× cost. Your dhan-options-analytics can compute this live.

6. When Gamma Scalping Shines

Use GEX (your GEX article) as filter: pinning regime = scalping gold; acceleration regime = avoid.

7. Risk Management

• Size straddle so theta < 1% capital/day
• Cap hedges/day (cost budget)
• Stop if 3 consecutive negative-EV days

8. FAQ

Q: Futures or ETF to hedge?
A: NIFTY futures for size, NIFTYBEES for small. Both liquid.

Q: Which expiry?
A: Monthly for stable Γ; weekly only if you watch spread.

Q: Capital needed?
A: Straddle + hedge margin. 1-2% risk per trade.

Q: Advice?
A: No. NISM-Series-XII educator, not SEBI RA.

8. Worked Example: One Scalp Cycle

Suppose NIFTY 24,800, straddle Γ = 0.04, θ = ₹120/day, band = ±0.25 delta.

T0: spot 24800, delta 0, hedge 0
T1: spot 24900 (+100), delta +0.22 (below band, wait)
T2: spot 24950 (+150), delta +0.27 (BAND HIT)
    -> sell 0.27 delta via NIFTYBEES @ 24950
    -> scalp P&L = +0.27 × 150 = +₹40.5 per unit
T3: spot 24900 (-50), delta +0.10 (revert)
    -> buy back 0.17 delta @ 24900
    -> P&L = -0.17 × (-50) = +₹8.5
Net scalp: +49, theta cost ~₹40, commission ~₹5 => +₹4 positive

Tight band (0.25) → small but frequent. Wider band (0.40) → fewer, bigger. Optimal band = where 0.5×Γ×band² ≈ 2×cost.

9. Cost Model (The Real Decider)

breakeven_move = sqrt(2 × (theta×t + cost) / Γ)

If Γ=0.04, θ=120, cost=₹10/trade: breakeven ≈ √(2×130/0.04) ≈ √6500 ≈ ₹80.
Spot must move >₹80 between hedges for profit. Below that, theta wins.

DERIVED: Low Γ (far expiry) needs huge moves; high Γ (near expiry) scalps faster but Γ collapses.

10. More from Shakti

• https://shaktitiwari.in
• https://optiontradingwithai.in
• Repos: nifty50-xgb-engine · dhan-options-analytics
• Related: GEX on NIFTY · Volatility Skew