I often trade SPX options. Lately, I've been trading skew and having decent results. However, in an effort to continue improving my trading, I'd love to hear any thoughts on peoples' preferred choice of volatility to plug into their model in order to delta hedge. I hedge with Black-Scholes-Merton. Yes I know its wrong, but in my experience if I account for the left tail and bound my downside risk, the model is still remarkably robust throughout most market conditions. The question is, do you use IV, actual vol, or some other number? Moreover, if you do use IV, do you use a "locked-in" IV - i.e. do you freeze the IV you purchased/sold each individual option at and compute deltas at that vol over the duration of the trade? Or do you let IV float and constantly update as the market moves?
When I traded, I used market implied vols but smoothed the wings. The Ivol gets too high at market values for far OTM short duration strikes.
Thanks. So a quick follow-up question then....as you rehedged, did you update implied vol as the market moved? Or did you lock-in the original IV you bought/sold at and calculate hedge ratios off of that static value for the duration of the trade?
I would reset to market vols if too many strikes were out of line. I always tried not to over hedge, hoping other options would keep me within my limits. My auto hedger was always set to round down. I never felt deltas needed to be exact. Toward the end of day, I would try to get flat. During the day, I cared more about what vol I bought or sold.
I believe hedging at the prevailing market IV will provide a smoother PL vol however hedging at your expected and presumably correct realized is optimal from a $ perspective.
...as to your question about when to re-evaluate vol it gets to the question of when to hedge - it's either based on delta bounds or time intervals. Afaik delta bounds are superior absent trading costs.
Do you guys have guidelines between hedging with options vs the underlying vis-a-vis: long vs short volatility strategies long vs short directional spread strategies when capital is not a consideration? To me, it's rather simplistic. Directional is hedged with underlying (i usually exit promptly). Volatility trades with options to isolate the other greek risks.
I tend to almost always hedge with other options to cut down on model risk. Moreover, hedging with a linear instrument can get nasty in certain situations. Say you have on an ATM butterfly. The stock is creeping higher and you need to hedge your short gamma. So you buy stock. The next day the stock gaps down hard on the open. The long gamma on your downside wing kicks in to protect you, but your hard deltas are pointed the wrong way. If you'd hedged with soft deltas (bought a call) instead of stock, you'd be hurting far less.
Roger, I have been increasing size with volatility spreads lately and trading to expiration. The difference between hard/soft deltas are quite palpable. I almost wiped the profit on a 60 day 1SD short butterfly for lifting a hedge for a single day. well, that and the whipsaw putting it back on. I guess I would be out the door if I was working at a bank. Nice to have you on the board.
Smoothest pnl curve comes from hedging to the actual realized vol. The best predictor of the actual realized vol tends to be current implied vol. I personally believe how you hedge an option is part of the "art" of trading vol. Market makers and flow traders will hedge to implieds as their goal is to minimize their pnl volatility from the inventory while capturing as much bid/offer as they churn that inventory. Buyside/retail/prop traders should be more flexible in their hedging based on their views and the mispricing they are trying to exploit.