If you've been trading options long enough to have studied complex options trades, you've probably heard that iron condors have a lousy risk/reward setup. The way most iron condors are set up, the trader risks losing far more than the trader can make. Why would a trader even consider a setup like that?

That poor risk/reward setup is balanced by the high probability that the sold strike of an iron condor will not be violated at expiration. There's a high probability that the trade will be profitable at expiration. The way I and many others set up iron condors, the trade will theoretically be profitable at expiration about 81-88 percent of the time. Theoretically. Remember that word.

How do I know what the probability of profit might be? I could calculate standard deviations and use what I know about the likelihood of prices moving beyond a certain number of standard deviations. I could look at a Hoadley calculator. If I'm on think-or-swim, I could throw a simulated trade up and look at what TOS tells me the probability of profit would be. Ditto something like OptionVue or the freeware OptionsOracle. Or, I could just perform a quick-and-dirty calculation looking at an options chain. Here's how. If I'm selling a call option with a delta with an absolute value less than 10 and a put option with a delta with an absolute value of 8 or 9, I can use those numbers in the calculations. The formulas used to determine their deltas mean that the expectation is that the call has a less than 10 percent chance of being at or in the money at expiration, and the put, an 8 or 9 percent chance. Total those and subtract the total from 100 percent, and, the trade has at least an 81 percent chance of being profitable if held until expiration. In my case, it's usually about 82 percent these days, although it tended to be closer to 86 percent a year or two ago when I sold cheaper calls and puts.

Theory and real life don't always coincide, of course. All of us know that. We're wise if we temper our expectations of how trades will progress when those expectations are based on theoretical results. A couple of reasons explain this disjunction between theory and real life. One is that a large sample size is required before data conforms to a normal distribution. Most subscribers flipped a lot of coins at some point in a grade school math class to determine whether 50 percent of those were heads and 50 percent, tails. When you were flipping those coins, you probably experienced a run of many heads or tails in a row. If you flipped enough coins, however, the results approached that 50-50 distribution. You'd have to flip a whole lot of trades before you could be sure that your results would approach 81-88 percent profitable statistic. You might luck on that result your first year, or, you might find that along the way, you'd hit a run of several losing trades all in a row. If you were really unlucky, you'd start out with a run of losing trades, seriously undercutting your trading capital, on your way toward trading enough iron condors to get anywhere near a normalized distribution.

In addition, some doubt that market action is normally distributed, no matter how many trades one flips. There's all that talk about "fat tails" that you've heard from me and others. Market prices tend to spend more time moving toward and into the tails of a normalized distribution (the flattened-out part of a bell curve) than we expect them to do in a normalized distribution.

That unexpected movement seems more prevalent the last several years. It seems that markets have zoomed one direction or the other more often than I've seen happen in the last decade. I know traders who are currently losing on their iron condor trades for the third month in a row, and I just don't remember that happening previously.

It was in this environment that I was thinking about the whole theory that these type of iron condors, put on at those delta levels that many of us use, are profitable 81-88 percent of the time at expiration. Remember that "at expiration" term because it's an important distinction.

I decided to put this whole idea to the test using recent data. I back traded SPX iron condors with expirations from January, 2008 through April, 2010. These 28 months of trades were set up in as standard a manner as I was able to set them up in the software I was using, given that I wasn't using an automated system and I also didn't want the trades to overlap. They were set up from 28-30 days before expiration. The 10-contract call spreads were set up by selling the first call with a delta with an absolute value under 10, and then buying the call 10 points higher. The 10 contract put spreads were set up by selling the first put with a delta with an absolute value under 9, and buying the put 10 points under that. The software I used is proprietary and not available to the general public, so I can't show it to you or offer you a link. What I can tell you is that the data used for historical options prices was imported from the CBOE. I consider the CBOE a sound source, but I of course can't guarantee that the data was always 100 percent accurate.

I wasn't testing a trading style I intended to emulate. Instead, I was testing this whole theory under which we trade iron condors. For that reason, no adjustments were made. Some of those iron condors were fully profitable well before option expiration, but I was testing the idea that a certain number could be expected to be profitable at expiration, so I did not close them even if they were fully profitable 10 days into the trade, for example. Would a trade that was profitable 10 days into the trade necessarily be profitable at expiration? No. Prices could run the toward one or the other of the sold options the next 18-19 days of the trade. I did close the trades at the end of the last trading day before expiration. My intention was to avoid manually calculating next-morning settlement values for all the options.

I back traded 10-contract lot sizes, therefore putting $10,000 minus the credit received in a particular month at risk each month. What did I find? The 28 months of trades resulted in a theoretical cumulative profit of $22,025.00 on a maximum margin for any one month of $9,240.00. Losses were experienced two months out of the 28, or 7.14 percent of time. Although I'd expected to find that the percentage of losing months was higher than the anticipated 12-15 percentage, that percentage was actually smaller for this period of time. Surprise!

However, that end-of-test cumulative theoretical profit doesn't tell the whole story, and it shouldn't when you're evaluating a strategy, either. What were draw downs like?

One month's loss was $8,395.00 or 99.58 percent of the margin requirement for that month's trade. That loss wiped out more than five previous months of profit. It will likely be no surprise to anyone that loss was experienced for the iron condor expiring in October, 2008.

Even that big loss doesn't tell the whole story. We should distinguish between where prices were at expiration and where they were during the duration of the trade. At one point, the February, 2008 iron condor was down almost 69 percent of the margin withheld, theoretically down $8,260.00. Prices never touched any of that iron condor's sold strikes during that month, but prices did violate the sold strike in September, 2008. Only a last-minute bounce saved that trade. The September, 2009 trade came close enough to violating the sold strike that a trader would have been holding his or her breath. That wasn't the only time that some breath holding would have been necessary. The March, 2010 iron condor was in trouble, too, when the SPX traded within $0.14 of a sold strike two days before expiration.

Now that we've talked a little about results, we have to talk about traders' emotions. Would you have been able to stand by and let those trades go so wrong in February, 2008; September, 2008; October, 2008; September, 2009 and March, 2010? Me? I think not. A trade that comes within $0.14 of a sold strike two days before expiration: would I just sit there and let that close call occur with full faith in the statistics that tell that 81-88 percent of the time these iron condors will be profitable at expiration? That may be you, but that's not me.

However, only one of those months ultimately turned in a losing performance. If you or I had stepped in and opted to close those worrisome trades, we would have changed the whole premise of the trade. We would have locked in losses in some months that actually turned out to be winning trades. The statistics would no longer apply because we would no longer be applying statistics relating to probability of profit at expiration. We weren't waiting until expiration.

I know my personality enough to know that I wouldn't have been able to endure all those unrealized losses and temporary violations of sold strikes, even if I knew the statistics that said that most of those losses and violations would be reversed by expiration day. And, if the anecdotal evidence is true, neither can many other people. They think they can but they can't. They plan to withstand them but when push comes to shove, they cave. This method isn't for the majority of people, and that includes me. Other methods are available for trading iron condors while managing losses, and they can be tailored to each trader's style as long as the pros and cons of each are understood. Yes, they change the probability of profit, but, if employed skillfully, they result in results that are evened out a bit: smaller but perhaps more frequent losses.

Imagine that September, 2008, when the sold strike was violated temporarily and the trader employing this method saw unrealized losses mount up. A bounce saved the trade, but no bounce saved the trade the next month, when more than five months of profits were erased. The trader would likely have been traumatized and might have overreacted a month later when an unrealized loss reached above 12 percent.

This period of time did not produce back-to-back losses, but such losses can't be ruled out. Neither can a period of 36 straight months of profitable trades. Expectations of profitability don't promise that that the trades will start out with a number of months in which profits nicely build up to cushion the loss in a single month of losses, followed by a run of profitable months to again cushion the next isolated loss.

I am not writing proposing a methodology that I would never feel comfortable using. What I was doing was testing the premise under which we trade iron condors. I wondered if the strange market behavior of the last couple of years had undone that premise. I half expected to find that these past couple of years had seen altered results, although I knew that two years wouldn't produce enough trades to be statistically significant.

I was in fact surprised by the results. Instead of finding an anomaly--a time when prices had spent so much time in the fattened-up tails of a normal distribution that losses were more common than and bigger than anticipated--I found rather tame results. In fact, if I had been able to stomach this method, I would have been better off these last few months, when I mismanaged some iron condors and experienced a too-large loss.

So, am I suggesting that traders abandon all efforts to manage losses, move spreads or take other actions when trading iron condors? Should we all abandon those efforts in favor of just letting them run? No. Let's take my case in particular. Of course, I wonder why I'm working so hard to manage my iron condors when these theoretical results turn out this way. Then I take stock again of my personality, my preferred trading style, and my situation. I know that I can't tolerate even one almost 100 percent loss without having my confidence totally undermined. This isn't the right methodology for me right now, and I don't know a lot of traders who would choose to let a trade run up such a big loss without opting to at least close the trade at some preset maximum loss level.

If you're tempted to trade this way, spend some time thinking over all these issues and more. Anyone contemplating this method should remember that this was intended as a test of the theory under which we trade iron condors, not of a specific trading methodology. I did employ methodology, but only in the trade setup: the date on which to place the iron condors, the vehicle to use, the number of contracts, the spread between sold and long options and the deltas of the sold strike at trade initiation. Beyond that, there was no other methodology.

Traders should be aware that, although commissions of $1.50 a trade were included, no efforts were made to account for slippage or bad fills. The idea wasn't to provide proof of exactly how much money could be made but rather to determine whether those losses occurred as seldom as the probability of profit suggested they would. Traders should also be aware that what happened in the past--particularly when the "past" encompasses only 28 months of back trades--does not guarantee what happens in the future.