Sure. If you buy a straddle for 5 pts today and sell it for 5 pts 3 weeks from now, then any scalp you made during those 3 weeks is pure profit. You are not paying any decay on it. Every scalp you make is a profit on the offset.
RegT isn't much of an issue with a long gamma scalp; moot with futures options as the hedge decreases requirement under SPAN. The biggest problem with gamma trading tends to be psychological, missing hedges or trading weak. Suppose you buy an XYZ 100 straddle at 25vol with your 1/2/3 daily sigmas at 102/104/106. The position is already "long trend" but there are always traders who are looking for more. They'll see the 1sigma trigger at 102 in a strong market and push it, expecting 103 or more. The market trades to 102.20 and reverses to close at 101... missed on the reversion. The weak trader is less common, but tends to be on the retail-side and more high-frequency oriented. They are reversion-types that can't stomach unbounded risk strategies which puts them in the long gamma camp. The mean-reversion mentality and need to trade actively is expressed in weak-deltas and early scalps. They will invariably sell 5d at 100.80, 5d at 101.30 and so on... completely neutering the strategy and guaranteeing a few bucks on the scalp, but losing a multiple of the scalp in the underlying gamma-position. Another instance plagues true mechanical scalpers that live and die by the model. They have buy and sell limits in the market at their 1/2/3 sigmas, vol-adjusted in r/t. They spend more time canceling/replacing limit orders than actual trading. Problems arise when the model-vols rise and the 1SD hedge increases from 102.00 to 102.12... trader places his 20d sell at 102.12 and the market trades to 102.04 and immediately trades 102...101.88...101.60... missed scalp.
Would one HAVE to buy/sell a straddle? Couldn't one simply buy either puts or calls and then hedge with the stock/futures (buy 2 atm calls/sell 100 stock; sell 2 atm calls, buy 100 stock)?
Any strategy that involves stock is better in a haircut environment as you put up no capital to put on the stock. You always have to put up 50% in reg t even if there is no risk in doing so! Thank the boys at the SEC for that nice one. Futures options are a moot point to my recommendation trading long gamma into earnings as it's not applicable to indices. No way would I try this on index options. Only scalp gamma into high vol event plays.
Perhaps someone can clarify this for me, but buying both sides (the plain vanilla straddle) and then scalping seems superfluous when one can simply do a synthetic and scalp (as described in Natenberg). Either way, one is trying to pay for theta decay while being prepared to take advantage of rising vega ("buying vol") and hoping for the occasional 2 or 3 sigma gap move.
Risk , the frequency of adjustments is not important so much (except when it's comes to commissions). IOW , the price changes on the 100$ stock will trigger "X" 20cents , "Y" 40 cents and "Z" 80 cents adjustments , where X ,Y and Z PnL is almost the same. More of the personal/mental choice. Some individual stocks carry a small edge for gamma scalping : unusual intraday hi-lo / close-to-close ratio. While an average is around 1.5 , a few (SNDK , MSTR) hitting 2. Stocks with high monthly numbers of up(down) grades are great too. After reading all posts , looks like I'm wasting my time on indices.
Good points. More of an issue with short gamma scalping though even with the long gamma scalp, haircut treatment on the long straddle would be preferable etc. RE: futures options, indeed is much better. Retail Joe has access to SPAN too. Therein lies the rub. The strategy requires skill. Count me out then LOL. I would be bold enough to suggest this is true mechanical or no. Get too greedy (read: wait too long) and you miss the scalp. There's that pesky skill again. MoMoney.