3. Volatility Trading

Trading volatility itself as the asset, separate from price direction. Vol has its own behavior: mean-reverting, regime-shifting, and predictable in ways price isn’t.

Realized vs Implied volatility

Realized Volatility (RV) — actual, historical std dev of returns.
Implied Volatility (IV) — vol the option market is pricing in (extracted from option premiums via Black-Scholes).

The vol risk premium: on average, IV > RV. Option sellers earn this premium for taking the risk of volatility surprises.

The IV surface

For a single underlying, plot IV across:

Strike (x-axis) → smile / skew
Expiry (y-axis) → term structure

Together → a 3D vol surface.

Skew

For Indian indices (Nifty, Bank Nifty), OTM puts have higher IV than OTM calls. The market pays a premium for crash protection.

A steepening skew = rising fear. A flattening skew = complacency.

Term structure

Plot ATM IV vs days to expiry:

Contango (front IV < back IV) — normal, calm market.
Backwardation (front IV > back IV) — panic, near-term event risk.

VIX going into backwardation often marks short-term capitulation bottoms.

India VIX

The NIFTY Volatility Index — IV implied by Nifty option chain (similar to CBOE VIX methodology).

India VIX level	Regime
< 12	Complacent / quiet — short vol favored, but watch for surprise
12–18	Normal
18–25	Elevated — directional bets risky, vol selling lucrative
> 25	Crisis — sell vol carefully (or buy if you expect more panic)

VIX is mean-reverting (long-term ~15). Spikes typically retrace quickly. This drives many vol strategies.

Vol cone

For each maturity (e.g., 30/60/90/180 days), plot historical RV percentiles. Compare current IV to the cone:

IV at 80th percentile → expensive → favor selling.
IV at 20th percentile → cheap → favor buying.

The single best entry filter for option trades: what’s the current IV percentile? Buying options at the 90th percentile is usually a losing trade no matter how good the chart.

Strategies

1. Short Strangle / Iron Condor

Sell OTM call + OTM put. Profit if price stays in the range. Edge from IV > RV + theta decay.

Risk: unbounded (strangle) or large (condor).
Best when: IV percentile high, no major events, expecting range.
Indian context: Bank Nifty weekly short strangles are a favorite — but blow-ups are violent. Always cap risk with wings (condor).

2. Long Straddle / Strangle

Buy ATM call + ATM put. Profit on a big move (either direction). Pay theta + vega.

Risk: premium paid (limited).
Best when: IV is low, expecting expansion (pre-event setups, breakouts).

3. Calendar Spread

Sell front-month, buy back-month, same strike. Profit from term-structure normalization (front decays faster).

Best when: front IV > back IV (backwardation) and you expect IV to normalize.

4. Diagonal

Like calendar but different strikes. Mix of directional and vol view.

5. Ratio spread

Buy 1 ATM call, sell 2 OTM calls. Negative gamma, positive theta. Capped upside.

6. VIX-based directional plays

VIX spikes → market often bottoms within days. Buy index futures.
VIX collapses → market complacent — sell vol or hedge.

(India VIX itself isn’t directly tradable for retail; trade Bank Nifty/Nifty options to express the view.)

Gamma scalping

You’re long an ATM straddle (long gamma). As underlying moves:

Up → delta becomes positive → sell some underlying to re-flatten.
Down → delta becomes negative → buy some underlying to re-flatten.

You collect realized vol by re-hedging at extremes. Profit if realized vol > implied vol you paid.

This is professional market-maker bread-and-butter. Hard to scale as retail because of:

Transaction costs eating gamma profits.
Need for tight, frequent re-hedging.

Vega-neutral strategies

Long one vol leg, short another → net vega ≈ 0. You’re betting on relative vol between expiries / strikes, not absolute level.

Examples:

Calendar spreads (vega-neutral if sized right).
Sell front strangle, buy back-month wings.
Skew trades (sell OTM puts vs buy OTM calls if skew flat).

Volatility forecasting models

GARCH(1,1) — captures vol clustering. $\sigma_t^2 = \omega + \alpha r_{t-1}^2 + \beta \sigma_{t-1}^2$ .
EWMA (RiskMetrics) — exponentially weighted variance. Simple, robust.
HAR-RV — uses daily, weekly, monthly realized vol components. Outperforms simple models.

In practice, even 5-day rolling RV is a decent baseline. Fancy models add ~5–15% accuracy.

A live vol setup — pre-RBI policy

Setup: RBI MPC announcement tomorrow.

Observations:

Bank Nifty IV percentile: 78th. (High — vol bid up.)
Term structure: backwardated (weekly IV > monthly).

Trade idea: Calendar spread

Sell weekly ATM straddle.
Buy monthly ATM straddle (same strike).

Why it works:

After event, weekly IV crashes faster than monthly (event risk priced into front).
You profit from vol normalization (term structure flattens).
Net vega ≈ neutral, net theta positive (front decays faster).

Risk: Sharp directional move beyond breakeven hurts (gamma risk on the short leg). Hedge if needed.

Vol regime transitions

Vol regimes don’t last forever. Watch for:

Signal	What it means
VIX hits multi-month low	Complacency → short vol risky, buying tail hedges cheap
Skew steepens during a rally	Smart money buying puts → reversal warning
Term structure backwardates from contango	Stress arriving → short-vol shorts unwinding fast
RV surges past IV	Sellers bleeding — close shorts, consider buying vol

Common vol-trading mistakes

Selling vol because “premiums are high” without checking realized vs implied.
Naked short strangles — works for years, then one day blows up everything.
Ignoring gamma near expiry — short positions can swing 10× delta in a few hours.
Forgetting margin requirements expand in volatile markets — getting margin-called at the worst time.
Not having a hedge for tail events — buy cheap OTM wings as insurance.

Nassim Taleb’s adage: “Don’t pick up nickels in front of a steamroller.” Vol selling has spectacular Sharpe — until the steamroller arrives.

Reading list

Volatility Trading — Euan Sinclair.
Option Volatility & Pricing — Sheldon Natenberg.
Dynamic Hedging — Nassim Taleb.