I'm following this thread closely, if for no other reason than it is on options, which I have limited knowledge and experience with but a (perhaps dangerous) curiosity for, and I am an ex-Ohioan (now basking in 80 degree CA weather... I'll try and send some over Howard). My experience with manual back-testing of using long options in place of the underlying for my directional trading of equities has been rather disappointing (the insurance+leverage doesn't seem to be worth cost). So naturally I am more interested in non-directional spreads based on these testing experiences, particularly if a strategy does not correlate to my directional equity or commodity trading. At any rate, thank you so far for your posts Howard. My feeling is, if there is a positive expectancy in here, it is due to your trade management rules. The main reason I have been avoiding options (which I feel have a lot of potential due to inefficiencies and a lack of user-understanding), is that I haven't found an agreeable backtesting platform. Please let me know if you have found something useful. In my (amateur on options) view, selling volatility is not naturally a winning solution over the long haul. It would need to be augmented by some predictive ability as to future vol. At any rate, carry on, I am enjoying the discourse and I look forward to more.
Back-testing For heavy duty back testing I wrote a PHP application that runs on my workstation and on the server that hosts my web sites. I have a very elaborate procedure to avoid "curve fitting" during the search for a strategy worthy of downstream evaluation. The key issue is the source of data. Since my strategy uses a proprietary ToS (ThinkOrSwim) metric I saved data files from the ToS ThinkBack application for my app to chew on. For light duty back testing I use paper and pencil. Here, I use ThinkBack and just record the EOD results for each spread. Selling Volatility Not many identify this strategy as selling volatility. Even some who seem to have a lot of experience with options. So I congratulate you with your self-described amateur status and bringing up the subject. Before attacking volatility head on as an issue, let me remark that in the over five months that I have been trading this strategy with real money in the account that I report on here, we have seen one down month, one sideways month and three up months. There has been not detectable difference in performance. With respect to volatility, the entire five months have been at low volatility. Higher volatility periods remain untested with real money. The key question is, will the PoT (probability of touching) metric adjust properly to higher volatility periods? The answer is, it should, because volatility is one of it's components. If the PoT does not adequately represent the true chances that I will have to close the trade before expiration, then are my trade management rules sufficient to keep me from losing significant sums of money. I believe they are, but this remains untested using real money. Next we must ask if during higher volatility periods, will the credit received at the specified PoT be acceptable? If unacceptable credit is available, then we will find no acceptable trades. This means we are out of the market. We are not earning, but we would not be losing.
As a veteran of this strategy I will post a few relevant thoughts: 1. Iron Condors always look good when VIX is 15-20 range and stays there for some time. Current market since August has been a slow grind higher which makes this strategy work well if risk is managed correctly. The real risk is thinking the risk management set up that works 90% of the time will protect you the other 10% of the time. Backtesting with end of the day data has its own risks as it does not reflect the wide b/a spreads intraday that would kill a position and force you to close out by your rules even though end of day last mark would not trigger such an event. 2. There is no edge in this strategy. The profitability comes from reading the market, strike selection and, most importantly, risk management. A 40% blow out wont kill you but it will take 2-3 times as long just to make that back assuming no other losses (i.e. lose 30-50% in one year and could take 2-3 years to even get close to even.). 3. Theoretically probabilities like touching or expiration are static numbers based on theoretical models so use those numbers with caution. A strike could still have a 20% probability of touching but IV spike could cause a huge swing in premium. 4. The position is only delta neutral the minute you open it. After the market moves, one side begins to be at risk right away. 5. When the market explodes and VIX pops, best risk management is to reduce exposure and wait and watch. I stopped trading these in August 2007 when VIX environment changed and stayed that way for 3 years until recently. 6. Theta is your friend and Vega is your enemy but most of the time they cancel each other out so basically all that matters is that the market is not near your strikes as expiration approaches. Pure strike selection and risk management. 7. Risk/reward calculations are always somewhat skewed due to leverage of the positions. To be honest all that really matters is cash at beginning of period and cash at end of period and how much it grew. You might make 20% return on money risked in the spread but if the spread was 40% of your total cash then be realistic about what the actual return to your account was. Best of luck to you but always be aware that there is no true effective risk managment when VIX spikes and/or the market moves fast. Do not put all your eggs in one basket and get out when the market environment changes no matter what p/l the spreads are showing.
I've always wondered if these iron condor strategies that just try to manage 'risk' based on metrics and rules and not on market regime/environment/vola/macroeconomics could actually work. The strategy is real sexy in the details...it reminds me of a girl with lots of makeup, different color bracelets and bands that look like they're from India, a small dap of glitter on her cheek, and some sparkly stuff right between her cleavage to get you looking, and she crosses her legs just right and her posture is nice, even her feet look nice from afar in her heels...but you get to talking to her and she has no depth. All these rules to manage an IC in a slow vola environment, trying to get your risk management just right...but the only real 'trading' in an IC is how not to get blown up in the big move (and big is just relative to your own position, it could be relatively minor on a larger timeframe but eat up a 60 day vertical).
Thank you for some wise counsel. Aug & Sep VIX was > 20 and my returns were the smallest for those two months. There were some additional circumstances, but it is an important caution, none the less. Back-testing has loads of limitations, only some of which can be accounted for. One of the things I did was record bid, ask, daily volume and open interest so that I could adjust accordingly. Back-testing is only the first of four steps before I get into production (trading with serious money). I think that proper management provides for taking advantage of additional opportunities, such as rolling, as well as managing jeopardies. This further mitigates the cost of severe market moves. I'm a little confused here about the classification of the probability estimates as static. A spike in volatility would appear as an increase in PoT and would likely trigger an exit with or without a roll to a more appropriate short strike. The speed with which this comes on is an important concern. Very true. I agree. What remains untested is whether PoT combined with credit required will cause the "wait and watch" or I have to this on top of the rest of the rules. Agreed. Don't forget roll opportunities for the spreads on the other side of the IC. Spot on. That's why, in addition to reporting the yield on individual spreads, then on the spreads combined into Iron Condors, I also report month on month return. I prefer not to get deeply into money management here, but it is still important to consider impact of commissions and weeklies in the mix. Great advice. I hope I have the discipline to take it.
For the amateur, asking about selling volatility. Howard seems to be entering spreads with Delta of 16%? The volatility is automatic and instinctive, not needing calculation. You simply sell into an index as it closes on where you want your spread, the volatlility automatically rises and you get the same price. As the completed spead sits, the index fluctuates back away from your entry and you get a decline of volatility and a profit automatically. I´m not too sure I explained that right. I sort of just do it automatically having long passed the stage of figuring it out.
I just noticed the extra tidbits. ( contributions ) From my year of weekly trading, the Option Coach has it right in my opinion. Strike selection and risk management are ALL. What Howard is throwing into the mix is rollovers. However, it seems that if you close a spread that is threatened past his 20% loss point, ( for an Iron Condor ) you MUST be able to make back those loss, plus commission costs and spread costs, by having multiple winning side spread rollovers in the same 60 day period, I would think? I´m at a disadvantage here having only traded weeklies I cannot think yet of rollovers with any background of instinctual experience in a 60 day period. Looking at a P & F chart it should be possible. Whether it is probably I don´t know? The trouble for me is with, the low returns for the risks. To do money management, and let us start with something very small like a $5000 account, you are only going to earn 3% if right, on a 20% of your capital at risk. In other words to make $30 you will risk a $1000. That scares the bejeesus out of me, without something to protect myself, like some workable strategy like Howard´s rollovers. He has been operating in a bull trend, without a really big VIX spike and that is scary. Can his rollover adjustments do it? My only conclusion was; do not do any Iron Condors with a VIX over 22. Only do single credit spreads. The trouble is; the money is really in the Iron Condor strategy, but safety says when VIX is over 22 you should trade single credit spreads. I´m not at all sure I want to get back into credit spreads, unless Howard can show a sure fire way of adjusting for threatened closing of spreads and work back to at least a zero month with no profit, but no losses after being threatened and closing a side of the Iron Condor. In the meantime, I think I am sticking to straight buying of options in which the risk exposure is more easily managed. The rewards are the same, or greater. For me the risk is much less as well.
howards path of collecting credit on options is in the right direction, the risk is the trading variable, the problem i see is complacency,a belief that this works on paper ,is easy ,and therefore attractive,like any profitable trading strategy,or any job in general,the harder you work at it,in this case the closer you monitor it,the safer it is,an aggressive non complacent management of this would be to watch for the 3 trend days a month and spread off the weak leg,at this point you would become a directional trader which is what the complacent safe money bettor is trying to get away from,like it or not ,the risk is in the direction and needs to be figured in.. a riue of thumb for options before the other 90% figured out the greeks was countering with an option of equal value or add 2 together to counter the price of one
Excellent post and very true. Howard, ToS probability of touching is good for entertainment. The number doesn't give you accurate PoT. The real distribution has much more kurtosis. Compute the difference between ToS PoT and the actual distribution around your strikes over 5 years. If the difference has normal distribution you strategy has an edge (or postive expectancy). However it isn't the case as you will see that your difference has fat tails. I would recommend to make sure you can withstand two 40% loss in your account, especially if you need to support a family. Ninna
19 JAN 2010 Trading Plan Opportunity - Takeoffs are optional Column: IC Spread #61 is unpaired. Opportunity to form an Iron Condor. Column: Spread P/L Spreads #69, #59, and #61 at more than 80% of capped profit. Opportunity to roll. Jeopardy - Landings are mandatory Column: Probability of Touching Spread #35 shows at 24% PoT. With only 1 day until expiration, it looks likely this spread will be closed, perhaps at a loss. Column: Days Until Expiration Six spreads are 1 day from expiration. Column: At Risk P/L No caution indications. New Opportunity None