Monday, July 1, I closed my July butterfly trade and locked in my profit. Because I'm entering my trades 37 days to expiration, that meant that I had some downtime during the holiday week. It wouldn't be time to enter my August trade until Wednesday of the next week.
Downtime didn't mean wasted time, however. I knew that if I didn't engage my interest elsewhere, I'd be tempted to enter my August trade early and build up extra decay over the holiday. That would mean more event risk, however. Instead, I started testing a theory I'd wanted to test.
I trade a butterfly established below the current price and hedged with a deep-in-the-money long call position, call debit spreads, or a mixture of both, depending on various factors. I originally learned about this setup from a trader named John Locke, but as we should all do, I've adapted it to my own needs. For example, although I obviously still work, my husband is retired. We're in that age group that cares more about preserving the money we've saved than making money. There's nothing wrong with making money, but we want as smooth a return as possible. Therefore, I always buy an extra out-of-the-money put as Armageddon insurance. That practice most definitely zaps my profit, and I have to accept a smaller return on my investment than others who trade this type of trade. That's okay with me. It might not be with you.
Another way I've adapted the trade is by back testing different hedges for the upside. When implied volatilities are nearing 20 and especially when they're over 25, the deep-in-the-money long calls I like to buy are often too expensive. I typically buy a long with a delta of about 95-99, and I want the extrinsic value to be near or even under a dollar. I typically pay something like $116-$120 (times the 100 multiplier, of course) for the DITM long. However, when implied volatilities are high, I might be buying a long call that has $2.00-3.00 ($200-300 after the 100 multiplier is applied) of extrinsic value to get a delta that high. If the underlying climbs as time passes, implied volatilities are likely going to drop, and I'm going to lose that $200-300, sometimes losing most of it as quickly as a day or two. When I can't get the delta I want without paying for all that extrinsic value, I often hedge with call debit spreads instead.
Although I'm experimenting and making this trade my own, I still know of several other traders who trade similarly. I had noticed that another trader was a later riser than I am, and he tended to come to his trading desk about mid-morning. He usually started new trades midmorning or after, while I had tended to start mine about ninety minutes after the open. What used to be called "amateur hour," the first hour of the day, had passed. Prices had often settled a bit after the initial volatility. There was less chance of starting the trade and having the underlying take off immediately the opposite direction before I could get the hedges lined up.
I kept noticing that other trader paid less for his butterflies than I did, however. Sometimes he paid much less. It wasn't for lack of trying on my part, either. I tended to set up flies for days before I intended to enter so that I would have a good understanding of the current prices. I watched spread orders through TOS so I knew what other would-be butterfly traders were offering. I could see which trades disappeared from the screen, probably filled, or disappeared only to reappear at the different price. I felt that these and other due-diligence efforts meant that I wasn't just being a sap and paying too much. Was there something about the time of day that we were placing our orders that meant he paid less than I did?
That first week in July was the perfect time to start testing that theory. I could have back tested it through the software I use for back tests, OptionNet Explorer, but I decided to test it live by entering a paper trade at about 10:00 am and another near 11:20-11:30 and just watching how they performed over the next week. It was my intention to place each butterfly 20-25 points below the then-current price and then hedge the upside as needed to flatten the current profit-and-loss curve. I would repeat this exercise over several days. As it turned out, the RUT didn't move much on July 2 and 3, so all four butterflies that I had paper traded by then (9:55 and 11:22 on July 2, and 10:02 and 11:32 on July 3) were centered at 970.
July 2, 9:55 Version, Showing Risk Profile at Close 7/03, Calculated by OptionNet Explorer:
This chart shows the expiration graph, the Greeks of the trade, and, in the sidebar on the left, the options comprising the trade. The accrued profit (after accounting for two-way commissions, since I've set it up that way) was $70.00. This was calculated using actual options prices, not theoretical prices, but of course I don't know if I could have executed the trades at mid-price. Lately, it's been difficult to do that due to low volume and perhaps the changes since the RUT options have gone to the CBOE. I could have gone in and changed prices to add in extra costs for slippage, but since I'm just comparing times, I didn't go to that trouble.
What about the other three trades entered by the close of trading on July 3? How were they doing compared to this one?
Portion of Open Position Grid from OptionNet Explorer Software:
The first and third trades are those established at 9:55 and 10:02 (Central) on the two mornings. The second and fourth are trades established at 11:22 and 11:32 am on the two mornings. In each case, the trade established in the late morning was doing better than the one established around 10:00 am, although the differences weren't huge.
Maybe I was on to something, and I should wait until later in the day to enter my flies. Maybe I wasn't, but the time spent could prove helpful in future trading results. Update: I wasn't. The profit or loss depends more on the implied volatilities and the movement of the underlying as I'm placing the various components of the trade. More testing ahead. I also could have run this test on TOS or on any brokerage platform that allows for simulated trading, so such testing is available for most traders, even if they don't want to spring for a software such as the one I use.
I want to run a back test, too, as soon as I carve out time for that test and that's a trickier endeavor without such software. Think-or-swim has "thinkBack" but I haven't used it and can't comment on how workable it might be. I don't know of similar setups on other brokerages, although some may have some.
What back test interests me? I've always preferred the SPX as a trading vehicle over the RUT, but the electronic fills on the RUT led most people to prefer the RUT for this particular strategy. Now that the RUT options have migrated to the CBOE, fills function more like the SPX's. I'm thinking of running back tests covering a 30-month period to determine if the SPX is an appropriate vehicle for this type of trade. This software allows me to go in and trade with actual option prices during that period. If the IRS ever questions whether I work fulltime, I can show them my back tests. More importantly, I can uncover faulty thinking before I pay for it with actual money.
Downtime doesn't have to be wasted time. Before you jump into a trade because you're bored, think about trading theories you always wanted to test but never had time to run.