Lesson 11: Delta Hedging in Practice
Promise: Understand why real hedging doesn't work like the textbook says, and develop intuition for the judgment calls professionals make.
Why Textbook Delta Is Wrong
Black-Scholes delta is a mathematical derivative: the option price change for an infinitely small move in the underlying. But there is no such thing as an infinitely small move. BTC doesn't move $0.0001 at a time. It gaps $500 in a second.
Modified delta accounts for this. Instead of the theoretical derivative, you measure the actual option price change for a realistic move size, averaged over up and down:
where h is a realistic move size (say, $1,000 for BTC).
For ATM options the difference is small. For OTM options in high-vol environments, it can be significant. A 15-delta OTM call on BTC at 80% IV might show modified delta of 18 for a $2,000 move because gamma is convex in your favor over that range.
In crypto, where single-move sizes are large relative to the underlying price, always sanity-check your hedge ratio against a finite move rather than trusting the infinitesimal derivative.
The Bleed: Greeks Drift Without Spot Moving
Your delta changes even when spot doesn't. This is charm -- the time decay of delta itself.
The pattern:
- OTM options lose delta as time passes. Your 25-delta call drifts toward 20 delta, then 15, then 10.
- ITM options gain delta as time passes. Your 75-delta call drifts toward 80, then 85, then 90.
- ATM options stay near 50 delta but become increasingly unstable as expiry nears.
This creates daily hedge drift. You went home delta-neutral on Friday, and by Monday morning your book has moved even though BTC is flat.
BTC at $100,000, 30 DTE start, 65% IV. Spot unchanged throughout.
Drag the slider to watch deltas drift over 30 days with spot frozen:
Charm is most dangerous over weekends when two days of time decay hit at once with no ability to re-hedge. If you are short OTM calls into a weekend, your delta hedge is staler by Monday than you think.
Vol Changes = Time Acceleration
A vol drop has the same effect as fast-forwarding time. A vol spike reverses it.
Consider a 25-delta OTM call at 80% IV with 30 DTE. If IV drops to 60%, the option behaves as if weeks of charm hit instantly. Your 25-delta call might become a 14-delta call in one session, not because spot moved but because the probability distribution narrowed.
This is the Ddeltadvol (sometimes called the stability ratio). In practice, it means:
- After a vol crush, your OTM positions have lost delta -- you are over-hedged.
- After a vol spike, your OTM positions have gained delta -- you are under-hedged.
Every time implied vol moves significantly, re-check your delta. A 10-point IV move can shift your hedge ratio more than a $2,000 spot move.
Handling Expiration
As expiry approaches, ATM gamma explodes. A 100k call with 1 hour left and BTC at 99,800 has enormous gamma -- small moves flip delta from 0.40 to 0.60 and back. This is where experience matters most.
The smoothness imperative: narrow your hedging increments progressively as expiry approaches.
"Hedging is smoothing." The goal near expiry is not precision but avoiding discontinuities. A smooth series of small adjustments beats waiting and making one large correction.
If BTC is sitting right at your strike with minutes left, delta oscillates between 0 and 1 with every tick. This is where you stop trying to hedge perfectly and instead manage your maximum possible loss.
Transaction Costs vs Hedge Quality
More frequent hedging reduces delta risk but increases fees. Less frequent hedging saves on fees but increases P&L variance. There is no free lunch.
The asymmetry: short gamma traders structurally pay more in hedge costs. When short gamma, you must chase the market -- buy after it rises, sell after it falls. You are always on the wrong side of the spread. Long gamma traders do the opposite -- they sell into rallies and buy dips, capturing mean-reversion naturally.
Mean-reversion in prices benefits long-gamma traders. Trending prices benefit short-gamma traders (fewer re-hedges needed). Your hedging P&L depends on the microstructure of returns, not just total volatility.
Path Dependence: The Order Matters
European options are theoretically path-independent -- they only care about where the underlying ends up. But hedging P&L is entirely path-dependent because you hedge discretely, not continuously.
Take two scenarios with BTC starting at $100,000 and ending at $100,000 with the same realized vol:
- Path A: BTC drops to $95,000, then recovers to $100,000.
- Path B: BTC rises to $105,000, then drops back to $100,000.
Same start, same end, same realized vol. But if you are long a 100k straddle, Path A and Path B produce different hedge P&L depending on exactly when you re-hedged and how gamma was distributed across strikes.
Taleb's experiment: shuffling the same set of daily returns into different orderings produced hedge P&L ranging from -$72,000 to +$72,000 despite identical terminal conditions.
When long gamma, you want big moves when gamma is at maximum (spot near your strike) and small moves when spot is far away. The sequence of returns matters as much as their magnitude.
The Worst Case Is Worse Than You Think
A common misconception: "I bought the option, so my max loss is the premium." Not when you are delta-hedging.
Example: You buy a 30-day BTC call for $3,000 and delta-hedge. BTC trends steadily upward from $100,000 to $110,000 over 30 days in small daily increments.
- Your call gains value, but you are selling into the rally (hedging off the increasing delta).
- Each day you sell a little more at a slightly higher price, but the hedge is always lagging.
- In a smooth trend, the delta-hedge bleeds cumulatively because you are short the underlying against a rising market.
- Total hedge loss: $4,200. Net P&L: call profit $6,500 minus hedge loss $4,200 minus premium $3,000 = -$700.
You lost money despite the call finishing deep in the money. The smooth trend produced low realized vol, and your hedges bled more than the option's intrinsic gain covered.
Premium buyers can lose more than premium in trending markets (hedge bleed exceeds option gains). Premium sellers can lose more than premium collected in whipsaw markets (hedge costs compound). Neither side has a "known max loss" once dynamic hedging is involved.
Practical Rules
- Use modified delta, not BS delta, for risk assessment in crypto-sized moves.
- Check your delta bleed daily. Charm shifts your hedge ratio even in flat markets.
- Vol moves shift your delta. Re-check after any significant IV event.
- Narrow hedging increments as expiry approaches. Smoothness over precision.
- Long gamma: use limit orders. Short gamma: use stop orders. Match execution to your gamma sign.
- Track total P&L including hedges, not just option value. The hedge is half the trade.
Common Mistakes
💡 Tip: Try answering each question yourself before revealing the answer.
Try It Yourself
For a hands-on simulation of the delta hedging lifecycle, see the Delta Hedging reference which includes an interactive simulator.
See Also
- Lesson 8: Greeks Beyond the Basics
- Lesson 9: Reading Your Greeks
- Delta Hedging Reference
- Course Home
Navigation: ← Lesson 10: Market Signals | Lesson 12: Liquidity & Microstructure →