I've been transitioning away from iron condors. My new trade has been working well this year, but I'm not counting on it being the
trade yet. Neither should you count too strongly on continued success into the future if you've been trading a strategy less than about thirty months.
Why is that? Remember when you tossed coins in grade school to calculate percentages of head and tails? If you tossed a coin only six times, chances are that you might not get that 50 percent heads and 50 percent tails expected result. You might find that tails come up four out of six times. You might formulate all kind so reasons why that occurs. You might decide that the heads side of the coin was heavier, for example, so that the coin tended to land heads down. If you made predictions into the future based on those six trial tosses and your reasoning about why you got the results you did, you'd likely be wrong. If you've had six months of successful trades and you think it's because of the wickedly smart adjustment or entry you've developed, you could be just as wrong. You just might not have tossed enough trades at the market to get an accurate take on how they will perform over the long run.
Active traders think we need to toss our trades at the markets about thirty times before we can expect our results to be reliable into the future. What's so special about thirty months? Actually, statisticians would probably tell us that traders need far more than thirty data points before a particular result should be deemed trustworthy. Those of us who trade strategies that take a month to unfold can't wait for hundreds or thousands of such trades before we decide how much we trust our trades. We need a smaller number.
Choosing thirty months comes from anecdotal observation more than statistically important numbers, as far as I've been able to ascertain. As it turns out, thirty months tends to be a time period that provides considerable variations in market behavior. Most randomly chosen thirty-month periods would include at least one sharp decline, relentless rally and stifling sideways move with declining implied volatilities. We can't count on seeing that kind of variation in six, ten or maybe even twelve months of trading.
A thirty-month period proves important in another way. In John Gapper's How to Be a Rogue Trader, Gapper notes that rogue traders tend to react to losses by increasing their risk, doubling or tripling the size of trades to make up earlier losses. He points out that such behavior can carry on for a year or two before the rogue trader crashes and burns. Before you're certain that a trade works for you, you have to be certain that it 1) works through all types of market conditions and 2) you aren't tempted to increase risks to make up for losses when they start mounting. You might find that you're more tempted to increase risks on one type of trade than you had been in the past.
To test out what randomly chosen thirty-month trades might show us, I asked a non-trader who does not watch markets to name three random months over the last eight years. New traders open accounts and begin trading at any time, and experienced traders might begin trading a new strategy at any time. Choosing the months randomly echoes what happens in real life.
The non-trader chose April 2007, July 2009 and October 2011. My plan was to examine charts of thirty-month periods beginning with those randomly chosen months to see what could be seen. Of course, I can't examine a thirty-month period beginning with October 2011, but including that shorter time period will allow us to determine whether the trader might have garnered the same degree of experience with different market conditions.
Thirty-Month Performance Beginning April 2007:
I've chosen the Wilshire 5000 to chart since it's such a broad index. To make the thirty-month duration legible, I've chosen monthly candles. They minimize the market action, making it seem less problematic than it was. For example, we see that when the first candle, for April 2007, was formed, the Wilshire 5000 was climbing. The climb was actually quite a sharp one. Certainly, if we were to examine a daily chart of the April-December 2007 period, we would notice several challenging periods for traders, although the action looks rather tame and stagnant on the monthly chart. Traders who traded through August 2007's sharp downturn and equally sharp rally off the low, displayed in the small-bodied candle with the long lower shadow or candlewick, might have thought they could handle the worst the markets could deliver. If they'd traded April-December 2007 with a profit, they might have patted themselves on the back, thinking they had successfully managed a position through hairy market conditions.
The relatively tame sideways movement of those monthly candles, compared to what was to come, shows that their trades had not been given the thorough test traders might have believed they had. The sharp decline to come would have tested a trade's ability to withstand sharp declines with rising implied volatilities, wreaking specially harsh havoc on negative-vega trades such as iron condors or butterflies. Calendars are positive-vega trades that seemingly benefit from a rise in implied volatilities, but they're hurt by price movement of that extreme, as prices drop through lower expiration breakevens and keep going.
What about traders always on the lookout for stocks they might acquire through a strategy of selling naked puts? Perhaps in June 2008, as this index was again testing support that had been holding, some of those traders were utilizing the increased implied volatilities to sell naked puts on their favorite stocks.
They likely got put those stocks at the next monthly expiration unless they bought back their sold puts. Those stock prices were likely to keep heading lower, too.
Trading conditions like those create heavy losses in some trades and test the trader's instincts when such losses are incurred. Is the trader loathe to let go of a losing trade even though the trade now depends on the "it's got to reverse" hopium for any improvement? Is the trader holding on and taking a maximum possible loss when a smaller loss might have been possible? Was too much money locked up in the trade? Is a margin call possible? Is the trader finally bailing when the loss gets too painful, too big to bear, but then jumping right back into a new trade, determined to make it up?
Beginning in March 2009, traders faced new obstacles: relentless rallies with falling volatilities. Falling volatilities help iron condors and butterflies, right? They do as long as prices stay within certain ranges. Many iron condor traders have plans that include rolling an in-trouble credit spread out of the way by buying back that spread and selling a further-out one. However, a rabid rally produces conditions that make that tactic almost impossible to execute to any benefit. I can tell you from personal experience that during rabid rallies with falling volatilities, it may not be possible to roll an SPX 10-point credit spread more than 10 points higher and collect any workable credit at all on the spread. In order to even begin to make up the locked-in loss experienced when the old credit spread was bought to close the position, the trader might need to sell 1.5-2.5 times the number of contracts at the new strikes, only slightly above the in-trouble spreads. For example, if it had been the iron condor trader's practice to buy-to-close an in-trouble SPX credit spread when the absolute value of the delta of the sold call reached about 22, the trader might find that she had to sell the 16-delta call in order to collect any decent credit and sell two times as many as she originally had sold. She's taking on twice the risk to take only a baby step away from danger.
Many iron condor traders learn to fear rabid rallies more than they fear steep declines, and with reason. It may be impossible to defend sold call credit spreads in a relentless and rabid rally. Can the trader trust herself not to try to take on more and more risk to make up what she's lost?
I found out that I could trust myself not to risk more and more, and I found that out through experience in the period that was the non-trader's second random choice for a beginning month. This next chart includes the month in 2010 when I experienced a larger-than-expected loss in a relentless and rabid rally. That was a painful period, but that period did prove that I could trust myself not to jump back in immediately with the intention of "getting" the markets and wringing my money back out of them.
Thirty-Month Performance Beginning July 2009:
This thirty-month period included both sharp rallies and sharp declines. Even though it encompasses thirty months, however, it provided less experience with range-bound markets. They require some skill, too. For example, trades that require premium selling, such as iron condors, may not return enough credit to the trader to justify the risk being taken on. Risks may be higher than seems apparent, too, as prices finally break out of those tight ranges. Calendar traders might find that their trades suffer from the dreaded "vol crush" when markets settle into a tight range and implied volatilities drop. Those traders used to buying straddles or long calls or puts might find that tight ranges hit them with unexpectedly steep theta-related decays. I note a few small-bodied, no-long-wick candles during the period displayed on this chart, but very few. Traders who began trading in July 2009 might not have experienced enough of such periods to be certain that they or their chosen trades could handle all trading situations, even with a thirty-month trial.
The October 2011 beginning point provided even fewer variations in market behaviors leading into the current time period.
Shorter Time Period Beginning October 2011:
Most of our subscribers will have been trading during this period. They will remember difficult market conditions, particularly when characterized by markets that gap more at open than they used to move an entire week. They'll also remember May's steep decline. However, when we compare this chart to the previous charts, we see that this decline didn't measure up in magnitude or endurance time to some of the other declines of the past. Traders who began trading in October 2011 have certainly been tested, but the markets can deliver harsher tests than have been delivered during this period.
I've been trading my new trade during this period. It's performed almost as well as my back tests assured me it would, a good thing. I'm gaining more and more confidence in both the trade and my ability to handle it with equanimity. However, how would it behave if we go through another period like that in the mid 2000's, when the VIX kept dropping? Would this trade setup work then? I haven't traded it through a gut-wrenching relentless decline, either.
Barring a sharp mental decline, I don't think I have to worry about going rogue. I've proven that to myself over countless opportunities to go rogue. I'm happy with the trade, and I'll keep trading it and will even continue to gradually increase the size. I'm not, however, going to put all our savings into the trade. I'm still treating this trade as if it's an unproven quantity. I hope you'll consider doing the same if you've been trading only a few months or have been trading a new strategy only a few months. Of course, you won't want to trade with a single contract for thirty months before you size up. Of course, as you manage difficult months with aplomb, you'll feel more confidence and gradually allot more money to a trade. Just don't be certain that your trade covers all market conditions until it has done just that.