Over the next few weeks, I'll be trying to find time to run new back tests of my current butterfly trade. I want to test it for the period of August, 2004 to July, 2007. Why is that? It's because I'm a "just in case" kind of person and that period presented particular challenges for some trades.
Usually, when the volatility indices reach the levels they have now approached, signified by the horizontal blue line in the VIX's chart below, they tend to reverse, heading higher again. Equities tend to reverse near the same time, heading down. That's what has happened the majority of the time when the volatility indices have dipped to their current levels. That's why I'm being particularly careful with my negative vega trades right now, waiting to ramp up in size until I have a sense of whether that same kind of action will occur this time.
It doesn't always. That reversal did not occur during that period from the summer of 2004 until the summer of 2007, when they mostly traded below that same blue line. That period is marked by the orange rectangle in the following chart. I want to back test my current trade during that time period.
Remember that I rough out my articles early, so this chart does not reflect current levels.
VIX Weekly Chart:
While a glance at this chart tells me that the highest probability move by the VIX once it approaches that blue line is a bounce, scanning the same chart also tells me that I better know what I'll do if the opposite occurs and the VIX submarines beneath that blue line for a few weeks, months or even years.
My current most common trade is a RUT butterfly. That is hedged with DITM long IWM calls so that the profit and loss will hold relatively steady long enough for me to adjust if there's a relentless climb. Because it's set up as a bearish butterfly, it will benefit from a drop back under the highest part of the tent. "This won't work for low-volatility environments," a trading friend warns, but he hasn't tested it for those periods any more than I have. I intend to test.
Why test? If it works in one environment, won't it likely work in another? Maybe, maybe not, but I've been trading long enough that I have traded through many types of market environments, and what works beautifully in one doesn't always work in another.
For example, I was trading iron condors on the SPX, OEX, RUT and MID during that period marked by the orange rectangle, steadily increasing the size of my trades over those years. At the far right-hand side of that rectangle, I was regularly trading 100 contracts a month, split between three or four of those indices. Now, if I sell an iron condor to initiate a new trade, I want to collect at least $1.20 and preferably $1.40 or above. Back then I was lucky to make much more than $0.90 per contract, and, if the volatility indices continue lower now, iron condor traders might again experience difficulty collecting credit.
I typically establish my iron condors about 60-65 days until expiration. I typically sell the first call strike with a delta under 10 and the first put with a delta above -9. Since we're talking negative numbers with the puts, I might sell a put with a delta of -8.45, for example. I typically put the hedging longs 10 points further outside the money. That would have meant that if I were opening a new SPX iron condor on Friday, March 16, 2012, 63 days until the May expiration, I would have tried for a +1505 call/-1495 call/-1230 put/+1220 put iron condor. Because the volatility indices have sunk so low, I would expect that it might be difficult to get the premium I prefer, although I didn't enter an iron condor this month, so I can't be sure. Near the close on that Friday, the mid price or mark for that iron condor would have been about $1.27. Whether I would have been able to get in at the mid or mark is uncertain since I didn't place the trade, but since we'll be comparing the mark at one time to the mark at another time, the comparisons should be valid. It's sometimes easier to get near the mark in one index and harder in another, and those "easier" indices sometimes switch places. Back in that low-volatility period from the summer of 2004 to the summer of 2007, it was often difficult to get anywhere near the mid in the RUT, for example, but now traders think it more liquid than the SPX.
Let's go back to that period in question to determine how an iron condor setup would have worked then. Stopping at random on a date 60-65 days to expiration during that period, I lighted on Friday, March 18, 63 days to that May's expiration. That day did not have enough strikes available to set up the iron condor according to the parameters I use. I well remember that during that period, we iron condor traders often had to call and ask for the strikes we needed to be opened and then wait for a day or two for that to happen. This points to one immediate difference that might occur due to low-volatility conditions. It might have been about during this time that I stopped trading the MID because the group in which I participated back then could never get the strikes we needed. If we asked for them to be opened, they weren't liquid enough for our purposes.
Low-volatility conditions might change the liquidity, then, in the far out-of-the-money strikes if you use them in your preferred trade. That's another difference you might see.
Friday, April 15, 2005 was 63 days until that June's expiration and more strikes were available. Moreover, a steep decline had raised implied volatilities. The decline that day was a 2.40 standard deviation drop. Was it likely I would have entered an iron condor trade that day? Maybe, maybe not. Let's look at the setup. If I'd used my same parameters for setting up an iron condor, I would have been looking at the 1235/1225/1005/995 iron condor and the mid or mark would have been a healthy $1.42, so the premium would have been good, but it was good for a reason. A steep drop and a spike up in the vols meant that the trading environment was riskier, so iron condor traders were getting paid more because they were taking on more risk.
This points to another difference in trading during low-volatility environments: those of us selling premium often looked for those days when implied volatility had exploded higher, making those credit spreads increase in value. That resulted in spreading our sold strikes further apart and further away from the then-current price of the underlying. However, that then-current price was 1142.62, only 1225-1142.62 = 82.38 points away from the sold call spread. This short distance from the current price to the upside credit spread wasn't uncommon. I remember many times when we could get only 70-80 points away from the then-current price of the SPX on the call side. What would happen if a trader had entered that iron condor after such a steep decline on that April 15, 2005, and the SPX exploded upwards?
Fortunately for our hypothetical iron condor trader in this instance, the SPX's explosion higher was met first with a helpful decrease in implied volatilities and then enough of a flattening in early June that the trader could have taken profits, but that doesn't always happen that way. This need to find a rising-volatility day to pad the credit we took in often led to experienced iron condor traders legging into the iron condor with only one side at a time. A big down day, if it brought the underlying right what was considered strong support, might have been used to enter the put side of the trade, for example. A swing higher over a number of days during the entry time period that brought the index up to relatively strong resistance might have been used to sell the call debit spread.
I wouldn't recommend legging into the iron condor for these changed market times, but we often employed it in the past. Even then, we sometimes found the underlying barreling in one direction only, without any opportunity to get into the other side and use the premium to cushion potential loss from the trending move. The trader needed to determine beforehand what would happen if there wasn't an anticipated swing the opposite direction and the underlying trended, with only half the iron condor established and only half the credit received. I used to set a time limit for some of my trades. If there hadn't been an expected swing in the opposite direction so that I could set up the second side within a certain time period, I would just hold my nose and set it up anyway.
This anecdotal evidence doesn't prove anything about my current trade. However, having traded one type of trade--the iron condor--through changing market conditions tells me that I might find that I need to change my vehicles, change the dates or manner in which I enter the trade, and set higher or lower profit targets. It alerts me that changing volatility conditions can change the trade in ways that might not be anticipated. Currently, because I'm the cautious sort and because I don't know how well this current trade performs as expiration draws close and gamma risk rises, I'm trying to close it out about two weeks before expiration. I'm willing to set lower profit targets so that I can do that. Will prolonged periods of low implied volatilities increase or decrease the risk into that last week or two until expiration? I'm currently buying DITM long IWM calls as an upside hedge. They have so little extrinsic value that they're not much impacted by changes in implied volatilities. Will I need to buy more of them or fewer of them to hedge upside risk if the RVX, the RUT's volatility index analogous to the VIX for the SPX, drop below its own support line and stay there? Is their hefty cost totally unnecessary or even more necessary under different conditions?
If you don't have access to subscription charting services that allow back testing, you can use a freeware service such as OptionsOracle to set up a mock trade and then vary parameters such as implied volatilities, time to expiration and price of the underlying to see how your trade functions. If you've been trading years, but those years are since March, 2009, you've been trading in a rallying market environment punctuated by some steep declines. That's not the same as the late 2007-early to 2009 environment, and those aren't the same as the summer 2004-summer 2007 environment. Don't assume you've got this trading thing down. The markets have a way of changing just when you start thinking that.