So far it's almost been like clockwork with this new trading system.  I mentioned that it averaged about a trade a month, and so far it's done that almost on the nose. Since this is a mean reversion system, the recent downdraft has flagged buys across the board. The averaging part of that statement should be paid attention to. Sometimes when the market is dive bombing, the system will flag a couple of days in a row as opportunities. Then in straight up mode, none for months. So I would consider this luck, so far. It could flag a trade or two this month.

So back to the trade at hand, SPY, QQQQ, IWM and DIA are all flagged. I reviewed all four and the credits on SPY and IWM were the best even though they weren't necessarily the most volatile. In fact, I would say QQQQ is probably the most volatile of the bunch but it yielded the worst credit. The second thing to consider is that the SPY NOV trade is still on, so do I want to expose myself to additional SPY risk? Well, IWM and SPY are pretty tightly correlated and the SPY has a little less volatility, so this becomes a pretty tough call. My plan is to take the IWM trade today and maybe the SPY trade tomorrow if we continue lower. Here's how they look.

 First IWM.

iwm-vertical-spread-dec-2009.png (43.70 kb)

And now SPY

spy-vertical-spread-dec-2009.png (45.00 kb)

Hopefully you're getting used to seeing the P/L structure of verticals. Each of these yield about the same credit and the distance between strikes is $2.00, which puts about $1.67 at risk.

Now the next trick is getting filled. The futures are pointing up slightly this morning, so a gap up may put these credits out of reach, but hopefully the market will rattle around enough today to fill. Things are changing rapdily as volatility has come back.

Will keep you up to date on the progress.

DISCLAIMER: This is not a recommendation to anyone to buy or sell.  I'm not a professional and I don't have much sense, so following me is hazardous to your finanical health.  The point of the post is to share ideas and hopefully get some feedback so that we can all improve our understanding of the subject.

I have a trading system that I retired right about the time the downturn started.  It was in the final phases of live testing, but I decided that I didn't like the actual results enough to continue committing capital. It also had a lot to do with the fact that another trading system I was testing was showing better results. It operated against individual Dow components, so before too long, I found myself holding 1/3 of the Dow.  Then whoosh, the bottom fell out.  The trading plan didn't have a stop loss, only a bar time out.  So I decided to "outsmart" the trading system and hold on to these guys because I was certain the market would bounce back.  Well, it has bounced back (although I didn't expect to be sitting on these positions for over a year).  Like most, I'm a little nervous about being excessively overweight in equities, so this has provided me with a great opportunity to unload some of these speculative equity positions from market cycles past.

So two things are accomplished here, one, I get to fix mistakes from the past and second, I get to lower my equity exposure.  Another win/win.

October expiration has passed, so it's time to analyze closed positions and think about new ones for November and December strikes.

Position Recap

The GE calls @ 18 expired worthless.  So I keep 100% of the 0.15 premium.  Not exactly juicy, but not too bad considering the stock hasn't done much but bounce around since I sold the calls in the middle of September.

The IWM trade flagged on September 2nd and taken on September 4th, brought in an 0.18 credit.  These options also expired worthless allowing me to keep the full credit.

November's SPY 92/94 vertical spread is well out of the money.  In fact, I tried to close this on Friday for 0.04, but didn't get filled.  With the original 0.31 credit, that would've been a fast move.  Usually things have to get way out of the money (since it's a spread) to be closed this quickly.  Well, the SPY has done so, but not enough for the market makers to give me 0.04 to close the position.  If there is another decent move up, I should be able to close it out, which would be nice.

The DEC XLF calls are showing a profit, but I'm getting less comfortable with these as time passes.  These calls were essentially a way for me to lock in some gains on my underlying XLF shares.  I felt pretty strongly that the financials were in for a set back.  Due to earnings, volatility shot up on the XLF and for a short time, the position was at break even.  The original credit was 0.41 when the XLF was around 15.35.  The XLF is now around 15.15 with the calls trading around 0.18.  I'm going to look to close this out around 0.10 and see if I can't catch a wave up and write some calls at 17 or 18 for December.  Stay tuned on that.

Looking Forward

Writing calls against GE is my top priority.  I don't like the credit offered for November at this point, I think the numbers weren't as bad as the market thought and shows some stability in the company. Maybe it won't happen this year, but I think people will slowly begin to buy up GE again. Hopefully it's next week and I can write a call or two.

The trading system that I write the verticals with isn't on a calendar, it's a mean reversion system, so it could flag trades at any time, but gets triggered on pullbacks. I'm comfortable enough with the wider strikes as I did with the 92/94 put credit spread on SPY. I'll stick with a 2 strike distance and see how that goes for awhile.  So far the trading system is 3 for 3.  Bill trying outguess the trading system is 0 for 1.  Imagine that.

And I'm still paying for the sin of writing puts against C back in 2007. If you weren't with us, my puts were exercised @ 22.50. That's not a typo. So with credits since, my cost basis is still somewhere around $21 which is obviously nowhere near the $4.60 C trades at today. I never bought any more C because I had little faith in the price and my original take was that Citi was big enough to absorb the losses without too much of a hit.  I was wrong and have been paying ever since. I decided that since I was wrong that I would not average a loser and held true to that. Now I feel that Citi is stabilizing in price and it's time for me to get some premium. I'll be looking at the $4 puts on another move down or a spike in implied volatility. The worst case is that I'm assigned at $4 which will lower my cost basis for C substantially and possibly get me out of this hole. It will be a happy day seeing C out of my trading account.

I spent a portion of the weekend getting caught up on older issues of Futures & Options Trader magazine.  If you're not a subscriber, you should be (it's free).  It can be downloaded from http://www.futuresandoptionstrader.com/.  Current issues are free, back issues are not.  I'm going to be referring to the May 2009 issue, so if you don't have it, I'll summarize the relevant bits.

Each month, the magazine has a feature entitled Options Trading System Lab.  The major players for the software OptionVue are the main contributors.  The feature takes a relatively well known strategy and runs a back test on historical data using the OptionVue software.  The results are usually disappointing. I don't mean that in a disparaging way.  If you've done backtesting for any length of time, you'll know that most of the ideas put forth are real stinkers.  In fact, I'd say over the course of many years, even before my days of working on RightEdge, I'd tested hundreds and now thousands of trading systems.  To this day, I use about 3 trading systems out of thousands tested.  There are some systems that have shown signs of promise in the lab and some that just plain stink, but I took an eye to May's Iron Condor results.

Strategy Summary

If you're not familiar with iron condor strategy, it's really just a fancy name for two vertical spreads put on in tandem.  It's basically an out of the money bear call spread and out of the money bull put spread. Both are credit spreads.  The P/L structure for an iron condor is below. 

spy-iron-condor.png (47.83 kb)

Most iron condor traders are looking to put the short strikes around one standard deviation.  In the P/L structure above, you can see that one standard deviation is marked by the vertical red dashes.  As a result, I put the short puts at 99.00 and the short calls at 115.00.  The long options can be placed anywhere you feel comfortable.

Risk

Now here's the catch on iron condors, my maximum reward is 0.44 less commissions.  My maximum risk is 4.00 - the credit.  This puts the risk/reward ratio at around 1 to 10.  That's not 10 to 1, that's 1 to 10.  This means that 1 losing month can wipe out 10 months of winners.  Couple this with the fact that one standard deviation means that it's probable the price has a 68% chance of staying within your shorts on either side, it sounds like a mathematical loser all the way around.  You need 90% winners in a case where statistically the price will only stay within your boundaries 70% of the time.

Backtesting the Strategy

Normally, this idea would be discarded especially with the pitiful risk/reward ratio.  Admittedly, it's not one of my favorite strategies, but let's see the results anyhow.

The test period started in January 2001 and ran through April 2009.  The starting capital is $10,000 and each month an iron condor was opened one standard deviation out.  They used 5 call and 5 put spreads. The system is always in the market and cycles the position (i.e. closes the current month and opens a new one for the next month) on the second Friday of each month.  The system had a 75% win rate with an average win of $763 and an average loss of $1,314.  What's noteworthy is the annualized return of 28.7%.  This would give you a return of 236% for the time period tested.  The S&P 500 had a negative return over the same time period.  Wow.

Discrepancies of Note

OptionVue constructs positions based on historical data.  I find the further out of the money the spread, the harder it is to get filled.  At best, the bid/ask spread on these positions is pretty lousy.  This probably wasn't accounted for.  Even dividing the bid/ask spread in half usually won't get you filled.  Second, the authors list the risk/reward ratio of 1 to 4.  I think that's generous.  I've seen in real life the ratio as poor as 1 to 12.

My Take

Given the discrepancies noted, if you were to consider putting on ICs, I strongly recommend using 1 contract positions and also putting them on the SPY ETF and not the big S&P 500 futures contracts (which are leveraged instruments).  Of course, never put on anything until you know what you're doing. Even in light of this impressive test, I'm still not very interested in iron condors because of the horrible risk reward.  I do, as most readers know, put on bull put verticals on the indexes when my trading systems flag them as a buy.  This has turned out to be successful so far in my first phases of live testing.

Newest trading system flagged a buy on most of the indexes since it was a reasonably big down day.  It isn't uncommon when all of the indexes to flag a buy for there to be another down day, although usually not as dramatic as the one before.  This sort of behavior gives one the idea to "outsmart" their trading system.  I've learned better over the year.  However, logistically, I really can't add positions for all indexes, so I'm going to increase my exposure on what I feel the best spread is from an risk/reward perspective.  The best of the bunch was a SPY vertical with the short strike at 94.00.  To increase my exposure and to further test my trading system, I'm going to open up my strike distances by one strike.  So instead of a 93/94 put credit spread, I'm going to do a 92/94 put credit spread.  See the risk graph below and as usual, click to enlarge the picture.

spy-vertical-spread-november-2009.png (48.37 kb)

This graph is an approximate credit amount, if we gap up this morning (which doesn't seem likely thanks to continued weak payroll numbers), I might hold off.  If we gap lower, I'll probably take the increased credit as opposed to lowering my strikes.  Will update with fill information.

Usual disclaimer: This is not a recommendation to anyone to buy or sell.  I'm not a professional and I don't have much sense, so following me is hazardous to your finanical health.  The point of the post is to share ideas and hopefully get some feedback so that we can all improve our understanding of the subject.

UPDATE: Filled @ 0.31 per spread.

Update on the last post, I took the IWM spread trade and went with it live.  I got filled with a 0.18 credit per spread.  I figured at the very least, something fun to watch until October.

Looks like the latest round of dips got my newest trading system all hot and bothered.  Today a trade in IWM was flagged.  To recap, I have an ETF trading system that I'm using to flag option trades.  The ETF trading system is a low frequency/high probability setup and I add to my winning % by selling put credit spreads one standard deviation out of the money.  A standard dev for the IWM is about $51 (current IWM price is 55.80), so almost 10% out of the money.  If I put on a put credit spread 1 standard dev, here's what the P/L graph looks like over time.

(click to enlarge)

At the time of analyzing the position, I'm looking at a credit of 0.18, before commission.  This sets my maximum profit of the credit and my maximum risk being the distance between the two strikes - the premium (0.82).  I could do a few things, first, I could widen my strikes, if I did a 49/51 put credit spread, I could obtain a credit of 0.33 per spread but would also double my risk.  I could also move the short strike closer to the money.  If I did a 54/55 PCS for example, I would get a 0.39 credit, but if $55 was breached at expiration, the position would lose.  So higher probability of loss, but bigger credit.  Options are always a trade off and allow you to fine tune your position to what you're feeling.

I haven't decided if I'm going to go with real money or do another paper trade.  I don't recommend anyone put this trade on.  Doing so will likely be hazardous to your account value's health.  This is in testing at this point.

As usual, just sharing an idea and always welcome feedback.

So the headline today is that the Dow is having its best month in 20 years. It sounds like a statistical outlier, some extremely rare event ... something you shouldn't weigh too heavily in your speculative analysis. Hopefully you're detecting my sarcasm. You can look at it a different way, I started testing my new system this month and wouldn't you know it, we have some statistical improbability on the very first month! If you're a trading system developer, surely you feel that the moment your new system goes live that the market is out to prove that your system doesn't work any more. heh.

Good thing I decided to cap my risk with a bear call spread. It's tempting to go with naked options to keep more premium, but if the indexes have thier "best month in 20 years" you're screwed. With the spreads, my risk is always 100% defined. I also opened way out of the money, so while the indexes have run straight up, my 56/57 spread is still out of the money (although not by much). The only time I ever consider running naked (sorry for the bad visual) is when I don't mind the pure directional bet that assignment might put me in. In the case of IWM, I put on a put credit spread to the downside because I didn't want to own more IWM. The bear call spread to upside is because I don't want to be short IWM. If either of those statements were inverse, I would consider going naked.

So the plan at this point is to wait. August is typically a slow month thanks to summer vacations. I expect a bit of mean reversion. It's a weird feeling when you don't really care which way the market heads. My only real care is if the market heads down sharply. But then again, if that happens, I buy more long term holdings. I think a wise man once said, trade what's there, not what you think is there. So I stress, once again, to have a plan for all cases.

In the previous article, I shared my 56/57 bear call spread on IWM.  Yesterday's close was 55.09, putting my short calls at risk. Even though it is statistically unlikely for the short calls to finish in the money, it wasn't impossible by any stretch. The impressive winning streak over the last couple of weeks has really made for interesting times on the speculative positions which are nearly all statistical. When things deviate outside of their probabilities, my speculative positions usually hurt. It's not always the case though. I do have some speculative positions (just plain stock) that have ratcheted some nice wins in the last couple of weeks. They were sold at their respective profit targets. It's also interesting because my long term portfolio (which holds some long IWM) is now down $103 total. I consider this a full recovery. So when the market moves up, I'm happy. When the market moves down, it hurts, but not quite as much.

For some reason I like to spend more time on creative positions and analyze those that are losing. The key here is that I don't want surprises. If the position is behaving differently than I anticipated with whatever the market delivered, I obviously don't understand what I'm doing and need to close the position. Thankfully, that's not the case here. In fact, I've been so happy with the way things are acting, I've decided to double down a bit. I opened another spread at 59/60. This is another 5+% away from my original bear spread. Again, I'll be elated if we run to 59 or 60 by August expiration. But if we don't, I have a little locked in cushion. I'm also interested in watching this one behave in the wild as well. And for those keeping score, it was another 0.12 credit.

A week ago yesterday, I completed the Iron Condor trade I mentioned in that morning's article.  At the time I wrote, IWM was at 50.50 but steadily rose throughout the day.  With the IWM at ~51.40, I felt compelled to take a stab at the 56/57 bear call spread and was filled at 0.12 (same price as my put credit spread).  This is also ~10% out of the money on a position that was already about 7% ahead.  This allows me to eliminate most run-up's.  For the IWM to run 17% and hold it for one month would be unusual, but certainly not impossible, hell, it happened this year. I just consider highly unlikely.

So the numbers to watch are IWM 42 and IWM 56.  These are where my short options strikes are.  A close between 42 and 56 equals full profit ($12 credit for each side = $24 per Iron Condor spread).  A close above 56.24 or below 41.66 would mean a loss.  My maximum loss now is $66 per spread and my maximum profit is $24 per spread.  It really helps up the reward side on credit spreads when you do Iron Condors.  This is why most folks put them on immediately instead of "legging into" them as I did.  Since only one side can expire in the money, you hedge your maximum loss.  But instead of 1 way to lose, now you have 2 ways to lose.  There's always a trade off.

I may or may not convert into Iron Condors in the future.  The idea is to continue with the put credit spreads whenever my trading system flags a hit.  That's going to be automatic.  Whether or not I convert that into an Iron Condor is going to be at my discretion. I say its not likely in most cases unless there's a significant run up.

Right now my IC is in the red as the market has continued to move up day after day. It would still take a pretty extraordinary move (5+% to breach my short options), but it's certainly within the realm of possibility. I've not wagered much, so this is going to be a fun one to watch.

If you trade Iron Condors, I'd love to hear your thoughts on this trade.

So my new system flagged a trade on Monday.  The market took a brief dip and then went higher on Monday and a bit more on Tuesday.  The put credit spread is firmly in the black for now. The next thought might be to complete this position as an iron condor. Briefly, an iron condor is an option position which really is comprised of two vertical spreads. Usually a put credit spread and a call credit spread. This is a market neutral strategy where you expect the underlying to stay fairly put so that both of your credit spreads expire worthless and you obtain maximum profit. Of course, markets don't always stay put and iron condors have pretty bad risk/rewards with extremely high success probabilities. I've read several cases where folks have been making consistent profits month after month for a year or two only to have one bad month wipe out all of the profit and then some. Iron condors are quite dangerous like that. You are short a ton of gamma.

My initial trading system is bullish only because I've never had any success being short of the indexes and being consistently profitable. So why would I entertain the idea of putting on a bear call spread now? Two things, one, if the underlying (in this case, IWM) moves much higher, the distance between my bull and bear spreads are substantial making the probability of both expiring worthless much greater. Second, I have a pretty good idea what the profit distribution looks on my existing trading system to hopefully discern a good spot to be relatively 'safe'.

(click to enlarge)

This profit distribution shows that most of my positions expire profitable and they're between 0-5%. A few make it as far as 10-15%. The 30-35% outliers concern me slightly because I think I might have a data error somewhere. But let's assume I don't and that 8 of the positions out of 100 or so make it to 20% or better. It would be relatively safe to put on a bear call spread when the underlying hits over 5% profitable. Most of my trades are ~10% out of the money, so that takes care of about 90+% of the cases of my bear call spread being a loser. And of course, this is offset by the fact that my bull put spread will certainly expire out of the money which adds to my bear call 'padding' should the short call get breached.

As of right this minute, my put credit spread was put on with IWM trading around $48. It's currently trading at $50.50 which represents about a 5% gain at this point. I haven't figured out my plan on this yet, so I may miss the boat to try it, but if it continues upward, I may give it a go and see what happens.

I'll keep you posted (after the fact).   Do not follow me. I could lose you lots of money.

Finally got a bite on the new trading system. I used RightEdge to develop the system. The goal of the system is low frequency / high probability. There are two reasons for this. One, I like winners more than losers, even if the R/R is skewed, it's better to win more psychologically. Let's say that you have a system that is 50% winners and 50% losers. The average winning profit is 5% and the average losing profit is 4%. This has a positive expectancy. Let's say you could take the same system and adjust it to be 60% winners and 40% loser with a winning and losing profit of 5%, which would you prefer (numbers aside). I like option B. Now, I'm completely making those numbers up, but hopefully you see my point.

Now the second reason for low frequency is that I can use options. Options require additional time for two reasons. One, once a trade is flagged, there are a number of strategies to pick from. Second, options are pretty illiquid so if you need to be in and out of a trade, options will let you down time and time again. My plan is to let the trade run until expiration (about 1 month per trade). Oh, and the options use far less capital, so I can let them go longer and give myself exposure to the same symbol a few times over without worrying about risk too much.

So on to the trade, my first entry was IWM. I went with an AUG 42/41 put credit spread. I was filled for 0.12 credit. The 42 strike is just over one standard deviation from the current price of $48. My backtesting shows that combined with my filters and indicators and a one standard deviation out of the money credit position that I should expect around a 95% success rate. Here again we have a high probability / dismal risk/reward type setup. Selling out of the money put credit spreads isn't a bad strategy by itself. You win some, you lose some. My idea is to couple this OK strategy with a pretty good trading system and see if the two make sparks.

I'll keep you posted. And as always, never, ever follow my advice or trades. These are works in progress designed to invoke thought and constructive conversation. Your ideas are welcome. 

Sing it with me ...

Tom Petty said it best for sure.  I've got my new trading system up and ready for testing.  As I mentioned before, it's an infrequent trader.  To reveal some more details, it's a mean reversion type system. At least it will expect that the current price when a position is opened will hold less than one standard deviation. My trading system has a filter in place that results in less trades, but from all of my tests so far, roughly the same profitability as a higher frequency system, because it has a higher winning percentage. Naturally, this produces a higher Sharpe Ratio and more restful nights.

So there is my empty "Scan" tab in RightEdge.  Nothing doing when it seems like so much is going on.  This is another bit of psychology that a trading system developer must overcome. It was a lesson that I learned the hard way in my earlier years. It's easy to try and outguess the system, but once you do so, you've invalidated all of your work as its no longer based on something systematic, but discretionary. Reward yourself by following your system as closely as possible. The testing phase should turn up differences between reality and backtesting. If there is a sharp contrast, the system must be chucked. But don't try and compensate, anticipate or outsmart your system. Make it a point to follow closely and change the system if necessary, not your behavior.

In my June 1st posting I state:

And in trading system news, I believe I'm about 1 month away from testing another trading system.  I haven't figured out if I want to use options or just plain stock yet.  It's a quick "in and out" type trading system with a maximum hold time of about 10 sessions.  It may be tough to overcome the option spreads in so little time. This is what testing is for though.  I may test with both vehicles and see which one makes the most sense.  I'll share results here. 

Welp, I'm ready to go live and I've decided to go with options.  I've made some slight modifications to the system to support more liquid contracts which resulted in less symbols in the watchlist which increased the hold time and reduced the number of trades.  The selection of options over stock has also increased my hold time because part of the adaptation results in being short theta (i.e. drawing benefit from time decay).

So the testing will begin.  As I've mentioned before, backtesting results do not always translate well into the real world.  There are a number of reasons for it, curve fitting, data mining and emotions all come into play.  So I like to take it slow and enter in with single contract positions.  I'll share my entries and exits along with the options strategy.  I'll lag my actual entries because I don't want anyone to stupidly follow me.

This is the testing phase!  It will probably take several months (if I'm lucky) to put any real money from my speculative pot into this idea.  This is also a low frequency trading system.  Again, most of the benefit is drawn from letting short options expire.  When I enter the trade, I'll feel that there's a high probability that the current price will hold, increase or decrease slightly over the next month.  I expect the system to average one trade per month.  And I do mean average because we could go 5 months without a trade and then get 5 trades in a month.

Be warned, few of my trading system ideas make it to the testing phase and even fewer make it beyond testing. 

Someone made a good point yesterday regarding technical analysis and market timing among the asset allocators.  These are the buy and hold folks (like me on my long term holdings) who all of a sudden know and practice technical analysis.  They start with something simple like the 200 day simple moving average and then graduate to SMA crossovers and then when those two seem shaky, it's the EMA that's the REAL holy grail.  This new found love for TA just so happened to flourish when the market tanked.  Up until that point, most buy and hold guys despised technical analysis.

What doesn't help one bit is that the 200 day SMA did provide a great exit signal this time around. Imagine if you gave your kid $5,000 to invest.  The kid picked a bunch of bio techs that went to the moon and doubled the money to $10,000. You know how that kid would feel? He's a better stock picker than Cramer (well, he probably is, but that's not the point). So he continues to take wild risks thinking that he knows the next greatest thing. There's a high probability that he will lose in the end. Hopefully, the experience humbled him.

This has been the basis for my studies posted over the last couple of days.  I feel the need to debunk a lot of the crap floating around and get people to understand and test for themselves.  Don't shoot the messenger, but these extremely simple strategies have been known for decades and have been proven random by the system builders many times over. To think that all of a sudden they work is fooling yourself.

Back when the VIX was at 10 in 2006, long term holders began to pile into risk.  There was supporting evidence that emerging markets, small caps and even a diversified individual stock portfolio was the key to outperforming.  Now the crowd discussion has turned to commodities, gold, bonds and even currencies.  It will be interesting to see if this is also cyclical.  The VIX will some day hover in the low teens, risk will be forgotten, TA will be a way to "surely underperform", high risk asset classes will be all the rage.  We can only wait and see.

In the meantime, I sincerely hope those that are newly crowned technical analysis experts do take the time to study and understand how system building works and how to test trading systems for randomness.  This doesn't mean reading other people's studies! This means doing your own homework and understanding the risk you're taking.

I lost a lot of money in my early days of system building thinking that I knew what I was doing and found the grail. It's easy to get the comptuer to lie to you or fool you with data. Please exercise caution!