Delta hedging

Definition

Delta hedging holds an offsetting position in the underlying (shares or futures) against an option position’s delta (greek-delta), re-adjusted as delta changes, so that P&L no longer depends on small directional moves. What remains is the VOLATILITY position: a delta-hedged long option earns when realized movement exceeds what was paid in theta (gamma scalping); a hedged short earns the reverse. Delta hedging is how options trading isolates volatility as the traded quantity.

How it works / structure

• Mechanics: position delta = Σ(option deltas × multiplier); hedge = opposite underlying position of that size; as the underlying moves, gamma (greek-gamma) changes delta and the hedge is rebalanced — buying low/selling high for long gamma (each rebalance banks part of the move), the reverse for short gamma (each rebalance locks in a loss — the structural pain of hedged short options).
• The P&L identity: hedged option P&L ≈ ½ × gamma × (realized move² − implied move²) summed over rebalances — realized-vs-implied volatility is the whole game (indicator-realized-vs-implied-vol).
• Parameters (engine-executable): rebalance trigger (delta band, e.g. ±0.10 per contract-equivalent; time interval; or move threshold), hedge instrument (shares vs futures — friction and financing differ), band width vs friction trade-off (tight bands track cleaner and pay more friction).

When it applies

Volatility theses that must not be directional (long gamma into expected turbulence, short gamma harvesting the volatility risk premium); neutralizing unwanted delta drift in multi-leg books (port-exposure-netting); market-making. On this platform it appears both as a strategy component and as the lens that explains WHY straddle P&L is a realized-vol claim (strategy-straddle).

Risk profile & failure modes

• Gap risk breaks the hedge: delta hedging assumes rebalancing is possible along the path; gaps jump over the band — short-gamma hedgers take the full gap loss on maximum size (the structural reason short premium’s tail cannot be hedged away, only sized).
• Friction vs fidelity: every rebalance pays friction; tight bands bleed costs, wide bands leak direction — the band IS a P&L parameter, not plumbing (ms-slippage-friction).
• Pin/expiry instability: near-the-money gamma at expiry makes deltas swing violently; hedging the last day is chasing a coin (mgmt-hold-to-expiry gamma cliff).
• Model dependence: delta comes from a model and an IV input; wrong vol = wrong hedge ratio = residual direction (opt-implied-volatility).

Evidence & limits

The hedging framework is Black-Scholes (1973) — its central construction is exactly this replicating hedge. Bakshi-Kapadia (2003) measured delta-hedged option returns and found systematic negative returns to hedged LONG options — direct evidence of the volatility risk premium that hedged sellers harvest and hedged buyers pay. Continuous-hedging assumptions never hold in practice; discrete hedging leaves tracking noise and gap exposure that no parameter removes.

Falsifiable-thesis examples

Illustrations only, not signals:

• “X’s realized volatility over this option’s life will exceed the entry implied volatility (hedged long position profits)” — falsified by the realized-vs-implied comparison.
• “Widening the rebalance band from 0.05 to 0.15 delta improves this program’s net replay P&L (friction saved exceeds tracking cost)” — falsified by the paired replay.

Cross-references

• The exposures traded: greek-delta, greek-gamma, greek-vega, greek-theta
• The quantity isolated: indicator-realized-vs-implied-vol, opt-implied-volatility
• Strategies it underlies: strategy-straddle, strategy-strangle (unhedged versions carry the direction)
• Book-level use: port-exposure-netting