Saturday, May 30, 2009

Weekly Update

Even after the run up from March the current chart patterns still look bullish.
Virtually all the technicals remain in a upslope with only minimal evidence of overbought conditions.
A continued advance towards the LR30 channel means could easily tack another 10% to current levels.
IMHO not a time to get aggressively short and I'm also inclined to minimize the butterfly exposure that I was enthusiastic about earlier in the week.
On the other hand, a rather conservative buy/write on the June GE 13s will still return 3% over the 21 days till expiration, which seems like a pretty good R/R at these levels.

Friday, May 29, 2009

Quick Gap Fade

Here's an example from Thursday's open of a gap fade trade setup that I posted a year ago. You never know when these setups are going to occur and with the recent bullish trend for the past few months, these have for the most part, not had a great performance record.
Nevertheless, on the heels of Wednesday's slide into the close, the relative odds favored continued weakness into the Thursday open, so the gap was a bit of a surprise.
I've found that the best risk/reward for these trades typically plays out between 9:32 and 9:45, and while I can't precisely define what's going on during this time period, after watching these things for 30,000 hours or so I kinda got a feeling about them.
This chart is showing 1 minute bars, although I watch 2 minute bars simultaneously on another chart.
The opening bar is quickly faded down, initially suggesting weakness, but I'm not ready to go yet. The ensuing litle hairy top (white circle) followed by the parabolics SELL at 9:46 is my cue to enter and is confirmed by the 1 minute NYAD which has fired a parabolic SELL at 9:43 (not shown).
I'm short at 34.84
My first downside target is the PP pivot, which isn't much of a trade, but the Qs indulge me and quickly blow right through PP, with only a a short pause before cascading down to hit S1 by 10:10.
The Qs then flash a little hairy bottom before the parabolics fire a Cover, and with the NYAD now flat, I decide to cut an run at 34.32 at 10:13 as the Qs pop up off S1.
That little run was only 27 minutes but yielded a nice .52 cents or almost 2 cents a minute, which was better than I'd hoped for. Those who've followed me for a while know that 1 penny/minute is my normal expectation for a return on these types of trades so this was a pleasant surprise.

Thursday, May 28, 2009

INTC Butterfly

Here's another example of the butterfly setup I posted on Tuesday. In this case we're looking at INTC and a July expiration.
The reason I chose INTC for the study is reflected in the stock's conformance to the LR30 channel, which has shown a slight upslope bias over the past 2 months, but which is now riding the lower band of the channel.
The butterfly scenarios for the June expiration are pretty dismal risk/reward situations (and are not shown), but looking out another month to July presents some possibilities.
The 2 setups posted below (3 and 4) have posited 16 and 15 as possible settlement prices for July expiration, depending on your perspective on market momentum. Other higher and lower settlement targets involve a greater initial debit and proportionately less potential returns.
Playing both scenarios 3 and 4 yields the risk/reward displayed in the final lower matrix.
These are just some ideas to ponder. The goal, of course, is to find setups where the net debit is extremely low and the breath of positive payout strikes is wide.

Wednesday, May 27, 2009

Qs MACD SIgnal Line Revisited

Here's an update of the MACD signal line system originally posted back on March 4th. The study uses my typical 16 month floating lookback period to gauge consistency of the equity curve.
Back in March the optimized MACD settings were 4,16,5. The current optimized settings are 3,14,3, a reversion to the MACD settings of approximately 2.5 years ago when I developed a fairly consistent daily bars trading system using the 5,20,3 and the 3,14,3 MACD crosses.
As far as the Qs go this recent study confirms the applicability of a the 3,14,3 timing model (for daily bars).
There was actually an error in the March 4th study results due to the way I optimized the BarsSince Entry functions. That error resulted in the Short cover exits optimized at 9 days, when in fact 4 days produced equal results with 5 days less risk exposure. As a result the current optimized values for len1 and len2 are 4 and 3 respectively.
Final optimized inputs for the system are 3,14,3,4,3.
The max consecutive losers for both longs and shorts is 3, the upper limit of my comfort level and the max intraday drawdown is respectable (IMHO) for both sides of the trade.
With the revised settings limiting position hold time to only 3-4 days this is a time frame that appeals to my short term risk exposure comfort level.

Tuesday, May 26, 2009

GE Butterfly

Last December I posted a number studies using a butterfly spread to generate a low risk income stream. The XLU example was one of the cleanest, but the original XLE butterfly case study generated a great R/R for several months.
The underlying strategy is that you risk a limited fixed amount in order to target a substantially larger return based on the price of the stock or ETF staying within a relatively narrow range.
After the March to present surge, we may be looking at a consolidation period before the markets move higher/lower (pick one). In such a scenario, playing a few butterflies can generate a few bucks with not a lot of exposure.
Here are 4 possible setups using GE puts ...2 for June and 2 for July. This is just to show the range of possibilities and is not intended as a recommendation.
I like to use the LR30 (30 period linear regression channel) to gauge possible price range for the upcoming month and that channel is shown in orange on the GE price chart above. The common practice is to look for a level LR30, but an equally productive tactic is to simply look at the slope of the channel and gauge the likely range of the price 30 days out. I often combine a LR30 view with an LR75 (not shown) view to step back and get a larger risk management view.
Many brokers, including Schwab, let you trade a butterfly (a 3 legged trade) for a single commission, and although there may be a few pennies in slippage, that's usually more economical than 3 separate entries. I typically enter these are limit orders at a par value (midway between the bid and ask) and let the market makers come to me. That works most of the time unless momentum is trending rapidly intraday.

Friday, May 22, 2009

More Overnighters

Here are a few more examples of the Goodbar setup long, one short and one that sure looked like a short but didn't work out.
This time around I'm looking at 5 minute bars and focusing on a stock...CLF, which has been a trading favorite for several in the REWIND chat room.
Keep in mind that a high beta stock like CLF is going to act considerably more squirrelly that a buffered ETF like the Qs, but curiosity got the better of me and I decided to check it out anyhow since my goal is to create a robust setup whose rules can be applied to a variety of stocks and indices.
We had 2 out of 3 good trades here. The middle setup didn't pan out even though all the criteria defined yesterday were in place for an overnight SHORT . . . .with the exception of the parabolic SELL.
One difference between yesterday's Qs signals and these CLF signals is the position of the MACD histogram relative to the other MA and MACD trend lines. In the case of the CLF the horizon of the histogram, either positive or negative has tended to carry through into the following day's OPEN with, of course, our middle setup, providing a dramatic exception.
Our trigger signals clearly require more refinement in order to avoid these false setup traps in the future.
Part of the problem may be the use of CLF as a test case, since it's been the subject of considerable recent volatility after selling off 12M shares @$21 on May 14th.
Just another reason I prefer the risk muted ETFs.

Thursday, May 21, 2009

Looking for Mr. Goodbar

This is a little study of my ongoing efforts to develop a reliable close to open trading system and features 4 recent days in the life of the Qs. This version profiles the 10 minute bars although I intend to incorporate these signals into my fractal close-open trading model using 2 , 10 and 30 minute bars in the same way I use RIMM, GOOG and AAPL in the 3 finger lead and reverse systems. The big difference for the Mr. Goodbar system is that I'm using different time frames rather than different stocks to forecast direction.
The chart above displays my standard stable of midpanel technical indicators and the thing that has jumped out at me after scrolling back through a few months of daily data is the following:
When the MACD signals are downslope into the close and
When the SMA 3 and 7 are downslope into the close and
When the SMA 14 is either downslope or level into the close and
When the MACD histogram is at the zero line coming off a positive horizon and
When the parabolics have fired a SELL signal into the close, THEN
SELL the close and cover at he subsequent OPEN.
The Inverse of these signals also apply and I'll be posting examples in the near future.
In the meantime, I'm working on programing these conditional signals into TradeStation to create a close-open forecaster.
The frequency of completely verified trades (ie, meeting all conditions) either positive or negative, appears to be about 1 every 4 days. Initially this system frequency may evoke a ho-hum response, but if you knew that every four days you could place a high probability overnight trade, would you do it? I sure would.
The setup cannot predict the magnitude of the overnight move, but simply the direction so the best tactic IMHO is to take what the market gives you with each of these trades and then wait for the next setup. This initial run will just focus on the Qs. It's highly likely that we'll see disparate results testing the setup with other ETFs and indices since those financial vehicles have different volatility and cycle period characteristics.

Wednesday, May 20, 2009

Tracking the Qs/NYAD

Here's another example from Tuesday's action of how tracking the NYAD in conjunction with the Qs can provide confirmation for trading signals and avoid premature entries and exits.
For today's study I've ratcheted the charts down to 1 minute bars, although 2 and 3 minute yield similar visual results.
One important item that I've mentioned before: Never try and follow the parabolics on the NYAD open for at least 30 minutes. Regardless of the close-open price differential on underlying stocks/ETFs, the NYAD's close-open differential will be wildly skewed about 80% of the time and you're just asking for trouble trying to read anything from that signal.
I've shown a couple confirmed NYAD/Qs entries: one 20 minutes before Monday's close and one just before 10:00 on Tuesday and one confirmed SHORT/COVER at 10:45.
Note the behavior of the underlying technicals and the action of the VIX on the NYAD and the VXN on the Qs.
And note how the parabolics gave concurrent BUY and SELL signals on these 3 occasions.
And this is why I ALWAYS watch the NYAD.

Tuesday, May 19, 2009

30/65 Minute NYAD

This is a little template I'm working on to gauge the probability of afternoon follow through of morning strength or weakness. It also has implications as the basis of an overnight only trading system.
The NYAD is the most important directional and momentum indicator in my trading toolbox and above are the 15 day charts of the NYAD using 30 and 65 minute bars. In the 30 minute case, the 6.5 hour trading day is displayed by 13 bars. In the 65 minute case the trading day is displayed by 6 bars.
The midpanel technicals remain with the same settings, hence the 30 minute chart is a bit choppier than the 60 minute. There's also some discrepancy between the parabolic signals and I'm exploring ways to smooth the curves. The 65 minute model clearly leaves a lot on the table, while the 30 minute model provides some real promise of gauging overnight action. While an ideal overnight model would involve buying or selling the close and then covering at the subsequent open, there's some work to be done here before deploying any serious capital to that strategy.
Such needed work includes refining the use of a moving average, standard deviation or z-score filter to qualify these overnight trades and is an area of research underway.
Of course, you can't actually trade the NYAD, but trading a basket of ETFs that are highly correlated to the NYAD is a tactical approach that initial test runs have shown to be highly consistent, especially if not deployed during the height of earnings season when overnight volatility and direction can yield some unpleasant surprises.

Monday, May 18, 2009

Divergent Qs Signals

Here's my little ETF basket updated as 7:45 AM pst and showing the morning pop led by XLE and XLF.
The VIX continues to ratchet in the low 30s and until we get a break there, we're not going to see anything dramatic happen. Best case, going forward, is a relatively narrow trading range for the week.
From a strictly technical standpoint, the mid and lower panel technicals are all bullishly upslope and with the Qs and IWM at the lower LR30 channel, the most likely scenario is for a return to the channel mean.
And then, there's a case to be made for an impending reversal. For the first time in the last 10 weeks, the Qs pivot bands did not remain upslope. although the net loss last week was minimal.
The Qs pivot impulse bounce off the zero line was the first time we've see that in the 39 weeks duration of the study so this week's price action should be particularly interesting from a strictly analytical standpoint as the implication of the (now positive) impulse indicator is that volatility (and weekly pivot trading range) are most likely to increase.

Friday, May 15, 2009

Qs Double 7s Update

Here's an update to my Qs Double 7s system. As with previously reviewed and updated systems, the input variables require some adjustment to reflect the bullish trend over the past 3 months. The new inputs are (4,4,7,6). See the original positing for the TS2000i code. Adjust BUY, SELL order verbiage as required to import into TS8.4.
This relatively simple concept system continues to generate a nice revenue stream with moderate intraday drawdowns. given the short duration of the average trade (3 days), pyramiding has little or no effect, so that option has not been activated and backtested.
Total number of trades for our floating 16 month backtest period is 90, divided perfectly between longs and shorts, a situation that system users will recognize as somewhat unusual.
Performance metrics for the longs versus shorts are within 2 % points, again somewhat unusual and reflective of the robust character of this little market timing model.

Thursday, May 14, 2009

3 Finger Reverse Update

As long as I was updating the 3 finger lead I thought it would be a good idea to compare the performance of the 3 finger reverse system (3FR) for the Qs same period of time.
For this update I've turned on pyramiding, which requires a maximum risk exposure 0f 600 shares in 100 share increment positions.
As expected by the trend confirming requirements of the 3FR, the system has prevented any short entries since March 5th, when the last pyramided short trade ended dead on the trend reversal.
As with yesterday's review of the 3FL, the performance metrics still show an 80+ % return and the equity curve is shown below. The revised input settings for the system are (46,5,14,10). The code remains the same.
As with the 3FL mentioned yesterday, an ongoing project is the development of fractal versions of the system using 2, 10 and 30 minute bars to generate more trades and, hopefully, enhanced returns over the same timeframe.
And finally, under the heading of "Articles I wish I'd seen earlier", this nugget from Bespoke.

Wednesday, May 13, 2009

Qs 3 Finger Update

This is an update to the previously posted Qs 3 finger lead system. (3FL).
The original system was posted back in Feb 09, when the market was in the beginning of a solid bear run. This time around we're looking at the system from the hindsight of an almost 3 month bull run. The total lookback period remains a floating 16 months.
As such, optimized input settings for the Qs are now 26,12,10,10, a variation from the initial inputs of 30,8,10,8, but still well within range as I optimize in increments of 2, not 1.
The system continues to generate a respectable equity curve with a very attractive consecutive winners:losers ratio -- 4:1 and a very tolerable maximum intraday drawdown.
Also, as with the original posting, testing various 3 finger components candidates still yields RIMM, AAPL and GOOG as the best fit for forecasting the Qs.
As with the original posting, pyramiding is not turned on. This is for the purpose of limiting capital exposure to a single position, as opposed to scaling in to a larger commitment.
I'm continuing to work on variations of the Qs 3FL including fractal versions using 2, 10 and 30 minute bars in order to increase trade frequency and, consequently, net gains.
Since I know someone is going to ask, below is the Qs 3FL with pyramiding turned on (think piling on or scaling in). The system maximizes at a 500 share position, not 100 as with the original.
Total number of trades is almost tripled and net returns are enhanced about 50%.
Performance metrics for the pyramiding version prove to be resilient and respectable (IMHO).

Tuesday, May 12, 2009

Qs Pivot Bounces

Another picture perfect pivot support and resistance day. It doesn't get much clearer than this. The Qs were one of the few indices to show green today and closed dead on the VWAP.
These are 5 minute bars shown, although the pivot bounces were clearly evident in all time frames.
Those 3 bearish chops at 11:00, 11:30 and 12:00 were a pretty good indication that bullish traction was going to be hard to come by.
This was not an easy day to trade (at least for me) as there was a distinct divergence of strength between the Qs and the rest of the major markets, including energy and the financials and I'm never quite sure when mean reversion action is going to kick in. Based on a quick lookback at yesterdays 4 ETF post, it looks suspiciously like everything may be headed back to the daily LR30 means. We shall see.
My one "decent trade" was an afternoon short of the Qs based on a bearish convergence of all my lower panel technicals confirmed by a parabolic SELL and a LR30,14,7 downslope that continued right into the close as the Qs plummeted back to the PP pivot. Covered at the close.
FYI, the last indicator on the lower technical panel is a Time Series Forecast 10+2 on top of the Detrend 30. The TSF smooths the detrend signal line and works well as a compliment to the 4,7,14 MAs.
So you're probably going to say, "Hey!, what about that little bullish turn up in the last 15 minutes?" "Wasn't that a signal to cover?"
Good point, and I was tempted to walk away from the trade at 3:50 but, being this close to the lower pivot and with the other indices in virtual freefall, I decided to risk it and ride into the last 20 seconds of the day. Sometimes the attraction of the pivots is the best tell...........

Monday, May 11, 2009

Going Forward

Going forward into the week, the Qs are the weakest looking component of my little ETF basket, with XLE showing an uber-bought breakout pattern. XLF and XLE are showing continued bullish momentum while the IWM has retreated to the zero neutral line and the Qs are actually trending lower. A little sector rotation in progress here I suspect as the Qs have led momentum for much of the rally.
Below, the ongoing Qs pivot bands study. Keep in mind that the Impulse indicator is not a tell for price direction, but for pivot range expansion/consolidation. Last week's action put the Qs dead on the zero line and, as can be seen by a quick scan of the chart, the Qs have not performed a zero line rejection for the entire term of this 37 week study, so we are expecting the Qs pivot range to continue contracting into the week.
Finally, here's what ChartsEdge has to say about the week. Check out the free "Daily Forecast".

Friday, May 08, 2009

Futures Mag

This rounds out my magazine reviews for the week. Futures comes in two formats and the website provides an incredible depth of materials and resources, not only for futures traders, but for FX, equities and options traders as well.
I've mentioned the Futures I-Trade shows before. Just go to the site here and sign up for the May presentations, which are also available on an on-demand basis so you can watch them whenever you please. The previous batch of videos included Larry Willams and Al Brooks with some revealing insights that are normally reserved for seminar attendees who fork over $1500-3000, for the same info.
Some surprisingly informative links from the drop down menus also. I've highlighted the TECHNOLOGY menu here so you can get an idea of what's available.
This is just a portion of the site as shown and you can spend a lot of time here checking out the educational and strategy links.
Well worth your time for both new and experienced traders, IMHO.....

Thursday, May 07, 2009

FX Anyone?

Continuing my shameless shilling for free e-magazines..... here's another one. Like the F&O Trader profiled Tuesday, just go online and sign up for a free subscription.
Not just for FX traders, there are frequently articles, systems and trading methodologies explored that are directly applicable to the equities markets, or, at the very least, should stimulate some tactical approaches for your own exploration and research.
Of particular interest to me in this issue was article on forecasting by Barbara Rockefeller which highlighted the use of both linear regression lines and linear regression channels. Those of you who've followed me for a while know that these are 2 of my favorite support/resistance forecasting tools and the article details some of the unique features of the LRs.
The article on intraday timing cycles is also of some interest, although keep in mind the FX markets are 24 hour off of several global exchanges. Nevertheless, the study does tickle a few ideas.

Wednesday, May 06, 2009

Down But Not Out

Tuesday's action was what a consolidation day looks like. The Qs performed their usual dance around the pivots, with S1 providing a distinct support level and the PP and Monday's close (the solid yellow line) providing resistance. I just managed a couple small trades on Tuesday as I was engaged in some ongoing market research projects and was not 100% devoted to the market dynamics. Sticking with my weekend update, I'll still feeling a bullish bias in the Qs at least until Thursday or unless Cisco really tanks on Wednesday. That being said I picked off a couple longs at 11:10 and another one dead on 3pm.
The first one tied up my money for 60 minutes and yielded but .18. The second trade was actually a combination of the convergent parabolics, 3/7 MA crosses and the S1 rejection along with the last hour bullishness that has been a continuing anomaly in the market for the past several weeks. The later trade was more successful and returned a net .27 for another 55 minutes of exposure.
The net result was that although the Qs total high-low range of the day was .43, those 2 little 1 hour trades picked off .45 , a number I could have doubled by also playing the short side, but, as mentioned earlier I'm still in the buy the dips mood. It's days like this that remind me why I prefer daytrading.

Tuesday, May 05, 2009

Latest F&O Trader

Here's the latest from Futures and Options Trader. A great free e-magazine. Sign up for a permanent subscription on line.
This month's feature article is by Ken Wood, otherwise known to thousand of traders worldwide simply as Woodie. A truly humble man with a unique trading setup (he never looks at price).
In this article he reveals some nuances of the "ghost" trade. Well worth a careful read. His CCI 6/14 signals can be mimicked on most platforms by minimizing the price chart and just focusing on a technical study window. For those wanting to know more about Woodie, here's the link to Woodie's Club. He's got multiple live chat/trading rooms depending on your focus, and is the real deal, which I say from personal experience working with him.
Guy Cohen's take on straddles and strangles is also an interesting read as he dissects the pro and cons of these trades and well worth your time if you're exploring alternate revenue streams.

Monday, May 04, 2009

Messing with my Shorts

I haven't run the LR30 weekly update for a while so I thought I'd give it a look after last Friday's amazing 8 minute wonder-close that put an otherwise red market solidly into the green.
While the RSI2 continues to sail along in uber-overbought territory, the LR30 and the 3 MAs (3,7,14) are all suggesting more bullish activity likely.
The MACD in all the ETFs looks like its ramping up for another bull run and MoneyStream (MS) continues to indicate the bulls are in control.
I consider the lower LR30 channel lines to be make or break points and am hesitant to get short very aggressively until those channel lines are violated to the downside.
Those who follow my real time trades in the Market Rewind chat room know that my midday shorts of the Qs at 34.37 on Thursday and Friday based on the Cheaphooker Short signal were both closed out for less than .25 gains into the Thursday and Friday closes, as I chose to cut and run until another day.
I'm not quite sure what it's going to take to derail this bullish juggernaut ride that began 2 months ago. I do know that there are a LOT of disappointed and frustrated Shorts lying around wondering the same thing.
Last week I mentioned that my swing trading view was that it was too late to get long and too soon to get short.
For safety's sake I'm maintaining that view and just picking up dimes and quarters on the daily VXN/Qs crosses and VIX/NYAD crosses. So far so good....
Here's Vertical Solutions weekly forecast for the S&P.

Friday, May 01, 2009

My Little Buddy

I've been using these new format charts recently for the blog posts so I thought I'd mention a little bit about them.
They used to be, but the name has now changed to, although the product is identical an still free. Having been at this for a few decades I've never found real time charts for free even approaching this level of refinement and robustness.
As I have 3 dual core laptops and 3 external monitors in play I just set the chart on a separate screen an use it as my primary 2 minute trading screen during the day.
The settings as shown include the VWAP, the parabolics (.04,.2) and the pivots on the main chart along with 3 linear regression lines (30,14 and 7).
The 4, 7 and 14 SMAs and the Detrend (30) are on the middle panel technicals with a 12,26,9 MACD histogram shaded in the background.
These are a little faster settings than those I've profiled before but even with relatively narrow range days like Thursday, the 4/7 MAs provide great risk management signals and work closely with the parabolics to trigger entries.
As I've discussed previously the 3 linear regression studies really act like an early warning system for impending trend changes and once that LR7 changes slope from the LR14 line you'd better be paying attention.
These settings are optimized for the Qs. My IWM and XLF settings are a little bit different and you're advised to craft the settings to reflect the volatility and ATR of whatever you like to trade. Signals on the double and triple ETFs also benefit from a little tweaking.
There are a lot more programmable technical studies available on the platform, plus news, a watchlist and a whole array of T technical indicators that you can monitor in real time. An incredible platform for those who are currently limited to EOD data, and even those who have real time feeds, AND FREE !