It's time to run some calculations. What kind and why?
Years ago, I remember sitting in a seminar for options traders. Most of the participants traded iron condors, as I did. A participant raised his hand and asked the presenter how many standard deviations away did he place the sold contracts on his iron condors.
The presenter shrugged. He didn't know. It seemed fairly clear from his response that he'd never thought about the question. It was left up to an audience participant who had been following his plays for a while to answer, pointing out that the presenter tended to place his sold strike about 1.25 standard deviations away from the current price of the underlying.
Does an option trader have to know about standard deviations to trade successfully? That depends, I'd say. Iron condor traders can use the absolute values of the deltas on their sold options to set up their iron condors so that they have a certain acceptable probability of profit. That depends on what the iron condor trader feels is an acceptable probability of profit, of course, but for me, it's at least 80 percent. Iron condor traders don't have to know about standard deviations.
Knowing how much your underlying might be expected to move during the course of a trade certainly helps, however. I've talked about standard deviations in other articles, and they're stressed in some courses about trading of any sort, whether of options, stocks or currencies. The idea is that about 68.2 percent of the time, an underlying's moves will be contained within a standard deviation either side of its current price. For example, if a security has a standard deviation for a single day of 2 points and if it's currently at 180, about 68.2 percent of the time when such conditions are in force, the security's prices would be contained between 178 and 182. About 94.5 percent of the time, a security's moves will be contained within two standard deviations of its current price. Knowing these facts about probability analysis can help traders set profit targets and also make sure that their stop loss levels are neither too tight nor too loose.
Traders don't have to calculate standard deviations on their own. Bollinger bands are based on standard deviation calculations. Although most charting programs set the default parameters to 2 standard deviations above or below a 20-sma, most modern charting programs allow a trader to reset those defaults.
Some value might accrue from running those calculations manually, however. From my earliest times as a trader, I began gravitating toward trading a familiar underlying, following its action so closely that I had a feel for how it would move under certain circumstances. The implosion of my first choice of a preferred vehicle--Enron--proved to me that it was probably best to have more than one familiar trading vehicle. However, the value of getting a feel for how those several vehicles move remained with me. When you manually calculate standard deviations for your preferred vehicles for a day, week, two weeks or month period and repeat that calculation as conditions change, you get a feel for how your preferred vehicle moves.
My March 20, 2009 Options 101 detailed the method of calculating a standard deviation, with the formula culled from a CBOE webinar presented by CBOE instructor and book author James Bittman and frequent CBOE presenter Dan Sheridan. The PowerPoint from that presentation included the following formula:
I had mentioned in that article that the "I.V." referred to the implied volatility of the at-the-money option for the expiration month of the trade. In other words, if I were going to put on a 60-day iron condor, I'd use the implied volatility of the at-the-money option for the options that were two months out. Most traders use the at-the-money calls rather than the at-the-money puts. In addition, some traders worry about whether to use calendar days or trading days in those calculations. Do what's easiest for you. The calculations will differ, but not by much.
However, the Thursday before option-expiration week in December, I was reading something that Jane Fox had written on the Market Monitor, the live portion of the site. Jane was talking about futures contracts, saying that if one didn't use the continuous contract, it was time to roll forward into the next series, the ones with the "O" moniker rather than the "H" for the December expiration futures. Jane made the recommendation because volume tends to decrease in the front-month futures contracts as their expiration nears.
At the same time, I had been watching and comparing the implied volatility changes in the RUT DEC and JAN 600 puts, part of a calendar position that I had open at the time, noticing how the sharp down days that we'd had some days tended to plump up the implied volatility in the soon-to-expire DEC puts but not in the longer-dated JAN ones. That all concurred with something I had just been reading in Jeff Augen's The Option Trader's Workbook about how the differences in implied volatilities might be exacerbated in European-style options.
Where is all this leading? I started wondering whether we ought to, at some point shortly before expiration, start looking to the back-month ATM options for our calculations for price movement in the underlying, just as futures traders start watching the next-month contract for guidance.
For example, let's go back to that RUT DEC/JAN 600 calendar, with the DEC option set to stop trading in a week at the time that Jane made that comment about futures. Thursday afternoon, the RUT had joined other indices in dropping rapidly. What if, as Augen says sometimes happens, the implied volatility of the DEC options had jumped, but under the premise that we'd had a lot of short-term-and-soon-to-be-reversed downdrafts, the JAN's hadn't seen a corresponding rise in implied volatilities? I try to get out of my calendars by the Friday before option expiration week, but what if that wasn't my practice? What if I'd planned to hold over through option expiration week and wanted to calculate a standard deviation for the RUT over the next week?
Would it have been best to have used that perhaps temporarily inflated DEC implied volatility, something Augen suggests can be exacerbated in some circumstances in European-style options, to calculate a standard deviation for the RUT over the next week? Or, should I have, at about that point, rolled forward to the JAN options for calculating that standard deviation?
Of course, I don't know the answer to that question yet. I don't even have a guess. Here's what I'm thinking, however. If, within a week of the time that an option is likely to stop trading, there's a big move either direction, with either a pop or a collapse in implied volatility in that front-month option, and I have a need to run a standard deviation calculation, I might use the at-the-money volatilities for both the front- and back-month options. If I'm making that calculation for the purposes of determining the most profit I'm likely to make, then I might use the smallest of the two calculations. If any move away from the current price is likely to result in a loss, and I'm making the calculation for the purpose of determining how big that loss might be so I can take the proper steps to minimize that loss, I might use the biggest of the two calculations. In other words, until I've studied this long enough to know whether it's ever valid to roll forward to the next-month calculations, I'm going to be conservative in the amount of profits those calculations tell me that I might make and be extra cautious in protecting against the losses they tell me might be possible.
One caveat must be mentioned: earnings. If a company will report earnings during option-expiration week, all bets are off. Long before the Thursday or Friday of option-expiration week, market makers will have already hiked up implied volatilities in the front-month options. Unless the underlying moves more than is typical or expected after earnings, those implied volatilities are going to quickly revert to normal levels after earnings. So, the Thursday or Friday before option-expiration week, they're telling us how much the markets expect the underlying to move after earnings, but that expectation is based on a blending of knowledge of past behavior after earnings, rumors about earnings and such matters.