Wednesday, March 31, 2010
Considering that the Qs daily ATR has now dropped to .52 (from .54 on Friday) a $.40 daytrade in 90 minutes is more than acceptable. A year ago when I first starting posting this setup we were looking for .01/minute net return as a performance target. That goal became unrealistic months ago . . a reflection of dwindling volatility. Yesterday's trade was a unique and isolated situation, albeit a happy one.
BTW, if you take a look at the midpanel technicals on the Qs Rotator chart you would most likely surmise that the Qs are going __. This view is confirmed by Frank, who has gone far beyond the Blogger's call to duty in preparing his latest multi-system market forecast.
On the currency front the UUP was the star yesterday, clinging to the upper LR30 channel band. As long as the channel remains upslope the prospects for UUP remain positive, so we'll be keeping a close eye on this situation.
Just as a point of reference on the 1/3 ATR trailing stop mentioned last week . . the Qs, with a beta of 1.11 have a daily ATR of $.52, while UUP, with a beta of -.27 has a daily ATR of $.16 so, from a practical standpoint, using 1/3 ATR or $.05 for a trailing stop with UUP is probably more a recipe for frequent premature exits than a smart risk management tactic.
Tuesday, March 30, 2010
The NYAD VIXEN setup generated a similar short signal, although it fired a bit earlier at 11:35 on both the 2 and 5 minute bars, a rather unusual coincidence. and was quickly followed by a bearish 8high/8low channel signal.
Of course the Qs surged up in the last 2 minutes and are up another dime after hours at 48.34 so I may regret my (possibly premature) exit. The good news is I can always buy at back and hopefully that will occur at a significantly lower price.
Monday, March 29, 2010
Friday, March 26, 2010
I've turned the 36 cell Strategy Matrix upside down, so while the SPY is reading +13, -23, the VIX is reading +30, -6. This 2 sided use of the Strategy Matrix is a new idea and I'll be following the results closely to see if the beta differential between the SPY (.99) and the VIX (-3.26) is reflected in the Strategy Matrix signals.
Thursday, March 25, 2010
Occam's Razor .
Using 2 day bars for the daily MLR sorting tends to deliver selections that are displaying both trending and rising momentum behavior. Those readers that have followed the MLR selections over the past month will have noticed that the top ranked picks consistently perform on the following day, or if failing that, show minimal disparity with the actual basket top performer. Since the big basket contains the majors (and a few sector wild cards) we have to anticipate that there will be some degree of arbitrage as the lagers play catch up to the leaders.
If nothing else, the Rotators clue you in to what's hot and what's not and the recent "bunching" behavior of the top sort values has been a good indication that the meltup has been broad and not just a one trick pony.
As mentioned previously, the need for trailing stops and the implementation of such is still a work in progress. One area of investigation is a 1/3 ATR(8) as mentioned in the Occam Razor post. If it works for the big prop shops there must be something effective about it.
Wednesday, March 24, 2010
SPY, IWM, Qs and XLF remain bunched, reflecting the markets broad advance, even at a low volume pace.
On a % change basis Brazil was the clear leader and may make further gains in the term term as it runs up to the LR30 mean as suggested by the underlying technicals.
The Euro and Pound continue to suffer consolidation pains, with the technicals pointing to likely further weakness in the near term.
Tuesday, March 23, 2010
Last week I mentioned the consequences of a low VIX value on the ATR and the resultant problems that situation presented for traders. Over the weekend I also noted that the Clueless One posted some updates to a few of the IWM and Qs Dirty Dozen portfolio of TradeStation systems that I posted a while back. I don't know if that's a good or a bad indication of Clueless's trading performance . . that he's had to resort to dredging up my old studies, but I know him to be an astute trader, smarter than the average bear, and almost as risk adverse in his trading as I am so I suspect he's just spring boarding off my humble little market probings.
Among other things, Clueless notes the flattening of the equity curve of many of the systems. Now I actually trade a few of these systems to supplement my daytrading so I'm pretty much up to speed on how's they're performing and I've noted this flattening myself so I decided to examine the performance metrics a little closer.
Above is the composite equity curve of the top 5 Qs Dirty Dozen Systems going back 16 months. The chart reflects the average performance for each of the systems.
And here's the surprise: the systems are actually chugging along with pretty much the same ratio of winning/losing trades as 1 year ago BUT the net return per trade has steadily declined as the VIX has fallen, followed by the ATRs.
To use an analogy. . there's only so much juice you can squeeze out of a dried up orange, and it's getting tougher and tougher to generate per trade returns similar to those of a year ago from the same orange (Qs or IWM).
This drying up is also painfully evident in the options market. Up until just month month ago it was pretty much a no-brainer rolling out ITM GE covered calls to generate 2-3% a month. That's just a distant memory as of this expiration, with the April 17s yielding 1-1.5%. Of course, you could sell the ATM 18s for 3%, but that's a little riskier situation IMHO. Again, the point is that for the last year you could have sold the slightly ITM GE calls for a pretty much risk free 2-3% per month . . a world away from the piddling 1% money market returns and 1.5% APR CDs, if you can find them.
From a practical standpoint I'm finding it harder and harder to identify attractive longer term trading situations. Most of the majors are still sitting at 6 month overhead resistance levels as the low volume melt up continues and, per the comments above, low volatility has significantly reduced the hedging edge previously offered by using covered calls or side spreads. Too risky to go long, to early to go short IMHO.
Time for Plan B.
Monday, March 22, 2010
For the next refinement of the MLR model I'll activate an equity curve stop based on daily bars and a trailing stop to lock in accrued gains. Tuesday Ill explore some general implications of a shrinking ATR on the shape of equity curves that may help explain a consistent recent trend to see these curves flattening out.
On a side note: Don Worden featured UUP as a LONG pick on Friday based on completely different technical signals . . so it will be interesting to see how this pick plays out over the next several days. The currencies have a tendency for longer term trend runs, which accounts for their relatively low price volatility and beta values.
Friday, March 19, 2010
I've juggled the metrics columns format once more and I think this is a keeper for a while. The yellow column is still the daily % change, while the column to the right "Price" is, in fact, a Worden Brothers algorithm that calculates a proprietary price volatility value. It's like beta, but it's not. The next column to the right "5 Day' is the 5 day performance of the ETF relative to the performance of the SPY. Beta is , of course, beta. I don't know why Worden can't figure out the VXX beta. . it's currently at -2.16 according to other data vendors.
Thursday, March 18, 2010
Above and below are 10 minute Qs charts with QQV and VIX overlays.
Is anyone a bit surprised to see the correlation?
And, just to confirm a longer range correlation of the QQV and VIX, the lower 2 Qs charts are based on daily bars. There's really only a single divergence in these 2 charts as reflected in December, when Qs volatility rose, but VIX declined. What happened there?
So, despite the complex formula of SPX put and call reconciliation involved in calculating the VIX, is it, in fact, just a reflection of Qs volatility? It certainly has an uncanny relationship to the QQV and that must be considered a major factor when evaluating the utility of the VIX as a trading tool for the Qs.
Of course, the VIX is a measure of the volatility of the SPX not the NAZ. For that we have the $VXN, which looks a little bit different and I'll be probing some of those implications later this week.
Wednesday, March 17, 2010
The fractal Strategy Matrix closed bullish with 33 of 36 signals on a BUY.
GE rose another 4.5% yesterday on 228M shares . . 3 times the Qs volume and has now retraced up to November 2008 levels. Keeping in mind that the old high was 42+, there's still a lot of ground to make up, and GE looks to determined to show its stuff short term.
Tuesday, March 16, 2010
The ATR8 has fallen from 6 in April to less than 1 today, a modest 83% drop. Reversing yesterdays' ATR/price ratio logic makes the situation appear a bit rosier but you can only put so much lipstick on a pig . . bottom line . . it's still a pig.
A little aside here... the 20 day beta of the VIX is -3.51, while the beta of he VXX is only -2.16, so a thinking man might expect the VIX to reflect a larger drawdown than the VXX. Not so.
I've tacked on the AROON indicator in the lower technical panel just for fun. This rather esoteric momentum signal needs to be "tuned" to reflect the beta of the study target and this is about the best we can do with VXX. Not a lot of BUY signals here and the door is wide open for further declines so thinking twice is the first caution here before getting bullish on the VXX.
Monday, March 15, 2010
Until the next black swan appears.
But, as part of this dramatic volatility decline, the daily ATRs have also decreased dramatically. Above and below are the SPY and Qs charts with the 8 bar ATR shown in the lower tech panel.
Since last April the SPY's ATR has dropped from 3.00 to the current 1.10 and the Qs have dropped from 1.00 to .52.
Now these are absolute values, not percentages, so the ATR of the Qs is actually large than the SPY. Also to be noted is the fact that these ATR values have to be considered in the context of the underlying price change. Hence, in April, an ATR of 3/price of 70 = 4.0%, while a current ATR of 1/price of 115 = 0.8%, only 1/5 the April value. You can run the numbers of the Qs, but they turn out similar.
The net result is that those 90 minute daytrades that use to produce a $.75 gain are now yielding a measely $.15. So while you may have improved your daytrading skills over the last year, you had to improve your performance by 500% just to keep up. They call economics the dismal science . . maybe they should call trading the dismal profession.
VIX ATR profile below looks exactly the same as the SPY, Qs and basically all the majors. So the VIX is kind of a victim of it's own success, and if you want to peek at the REALLY dismal consequences of this swoon, just check out the chart of the VXX . . the VIX futures based ETN. But more on that at a later date.
Neverending Story of the bull market. Friday was, well, a day to be doing something else. The NYAD hung on a neutral 1.00 value like static cling after a brief opening pop&go. TICK was neutral, volume was non-confirming and the closing stats , as shown below, were flaccid.
setp GE was the shing star of the day with over a 3% gain on sustained accumulation. This was an easy peasey trade if you track the GE/VIXEN setup chart on the blog or if you've got the setup on your own platform.
Every so often, and generally without much technical warning, GE blasts off , and pays off . . which is why IMHO it pays to monitor this little prima donna.
Friday, March 12, 2010
Not a lot of disparity between the top 4 and, with EWC slightly lagging 5th place, the old leaders continue to lead.
Yesterday's VIX action was a classic example of how frustrating it can be to try and capture VIX dynamics with VIX options. Although VIX did make a nice pop at the open, followed by a parallel action in the calls, for the remainder of the day the whole scenario went bizarro. After an early dip in the VIX it proceeded on up to new daily highs. . a situation that was accompanied by the plunge in call values to the low of the day. Now I understand that there are only 5 days left until expiration but I'm talking about the 20's here, not the 24's.
Following that 9:30 am pst high, the VIX plunged straight down, losing 1.34 points or 7% from the daily high to close at it's low of the day.
Meanwhile, the VXX actually closed up .4% for the day, further muddying the waters. Some days you just can't get a break with the VIX and yesterday makes the point in spades.
Thursday, March 11, 2010
Ultimately is was a strong day with only GLD and 20 year bonds displaying any weakness. Even the VIX and VXX were in the green, leading to another unstable EOD as the VIX accelerated into the close.
Thursdays have the greatest statistical probability of being short term pivot high days
so it will be interesting to see how that bet plays out.
With virtually all the majors extremely overbought it's hard to fathom how the markets can continue uptrend much longer. Nevertheless, the market seems inclined to inflict the maximum pain to the greatest number of traders on occasion and I'm foregoing any new longs until we get a couple steps back. Meanwhile, my portfolio, which has been net long since mid-February has been re-aligned to reflect a more defensive fully hedged posture.
Wednesday, March 10, 2010
While the market majors all closed in the green, so did the VIX and VXX and, as mentioned previously, this signal of market instability in typically resolved in favor of volatility.
Keep in mind that the MLR sort is based on 2 day bars . . were it based on only one day bars it would look like the sort below. The VXX and VIX made nice rallies into the close and Project Z now has the VIX as a BUY, while the TS pivot low signal generated a BUY on Monday.
Tuesday, March 09, 2010
This brings up caveat #2. The performance results of this study should only be considered in reference to the number of successful/winning trades. Since the VIX is a statistic and not a tradeable entity, there's a few monkey wrenches that have to sorted out here. The dynamics and pitfalls of entering these signals as VIX options trades are well beyond the scope of this post and each trader has to assess his/her risk exposure comfort level which, in turn, will determine how these signals are translated into trades.
Side note: the yellow dots along the price chart are pivot high/pivot low signals generated by the TS default PH/PL algorithm set to 4/2. I use the signals as a second set of eyes to confirm entries and guard exits.
I view this current study essentially as a jumping off point for further research but it's encouraging to find this initial foray into VIX/VXX alignment so consistent.
Monday, March 08, 2010
75%, with considerably fewer rally episodes. Of course, the VIX is a statistic and the VXX is a tradeable ETN so we're not really talking apples and apples here. But potential VXX traders should take note of the VXX chart as it probably reflects underlying volatility sentiment in a truer fashion than the VIX. I've run the VIX against VXX in a Project Z study that I'll post later in the week to support that contention.
For now the downtrend in VIX/VXX remains, with VXX hitting lower lows each day and with the VIX retracing Friday to a previous low of Jan 11th. . .which was followed by a month long double top rally. This is really a critical breaking point for the VIX and this week should reveal what the near term trend will likely be.
Friday, March 05, 2010
And, while waiting for the consolidation to break, you can mosey over to the real time trackers on the right side of the blog panel to see how the Qs and GE are doing. I've recently updated these 3 vignettes, as well as the GE VIXEN and the full Qs tracker further down. All 5 FSC charts are now set to 60 bars, which improves the definition of the signals and avoids possible misinterpretation. The Volatility and Qs Prognosticators display a little algorithm that hopefully captures turns in both trending and trading range markets, while the lower (and bigger) GE and Qs charts contain a different set of indicators designed to display momentum.
I use these simple 3 real time vignettes every day to guide my Qs trades. When I see an alignment of the windows as shown below I pay attention because the short term 6 bar (yellow) linear regression lines have all tipped into a confirming pattern that translates into a high probability setup from my experience.
Thursday, March 04, 2010
The VIXEN did a nice job of catching the big break at 10:35 and the trade was confirmed by the parabolics, the 8/8 hi/lo channel and the mid panel MAs so this was a high probability entry.
In like manner, the parabolics, 8/8 channel and 3 MAs lined up at 12:02 to signal a clean exit. This rather quick 90 minute trade yielded a respectable $.23 or 1.4%, a sizable portion of GE's current daily ATR of .34.
Reviewing the first 60 minutes of the chart is also instructive as it reveals a common VIX/price dynamic that often provides an alert for things to come. While it is easy to call that early VIX cross over and back a head fake type of move, the fact is that I've seen it hundreds of times and in about 85% of the cases it leads to the situation that unfolded at 10:30. This may be a testing of the water, so to speak, before the big move, or it may simply be the prop shops dropping the bid so they can pick up shares cheaper ahead of the bigger run up. Bottom line: my oft repeated mantra that "the second mouse gets the cheese" applies to this setup in spades.
Wednesday, March 03, 2010
The MLR Rotator now has a slightly different look and an adjusted portfolio of components. I've initiated the component changes based on continuing Project Z studies and will eventually merge the output of single stock/ETF Project Z system studies with the MLR ranking in order to evaluate the rotation model using 2 diametrically different timing models.
First, the deletions . . DBA, EFA and DIA. DBA has never fared well as a rotator candidate and DBC effectively covers the commodities market for my study. EFA tracks technically very similar to EEM and I prefer the robustness of EEM. DIA is, in my opinion the toughest of the majors to trade technically. The massive gaming by major prop shops of DIA components makes trading this ETF a challenging proposition and I think there are many other ETFs/stocks that respond to timing models in a more reliable manner. Project Z actually does a great job of timing the DIA (85%+) but I think it's 30 stock base makes it too susceptible to surprises. That's just me.
New components include GE, BAC and EWC. While those selections may seem a bit odd at first glance there's really some method to my madness. GE is like Caesar's wife, all things to all men.
It's a financial stock, it's a tech stock, and it tracks more like an ETF than a single stock. With daily volume that often exceeds the Qs and IWM, GE is a high beta juggernaut that provides some attractive trading opportunites for short-timers like myself. Ditto for BAC which routinely trades 200M shares a day and which tracks very well technically. While it has no tech side like GE, BAC does have a beta 1 point higher than the XLF and for rotation evaluation purposes I believe it will prove useful to see how BAC ranks relative to XLF on the daily rotator update.
EWC . . OK, here's a wild card but, after many, many hours of testing, I've found EWC provides an interesting relational sounding board for both EEM and GLD, and EWC technically tracks very well (92%) on Project Z.
Side note: I've realigned the Rotator columns with % change (yellow) moved closest to the sort value. To further clarify: the "5 Day" column is looking at the stock/ETF's 5 day performance relative to the 5 day SPX.
Tuesday, March 02, 2010
Monday's surge put EEM in slot #1 with EFA close behind. TLT may have seen the end of it's recent run as the market looks technically poised for further gains although volume remains non-confirming. With the VIX sub 20 (19.26), the daily odds for a retracement continue to loom. Most of the indices are sitting at overhead resistance so IMHO traders betting the long side are best served by small, cautious and hedged positions.
I've pulled back on the blog right panel BZX chart, instead turning on the BZV2, which tracks 2 minute volatility based on a blend of the VIX, VXN and VXX. The BZV does not reflect actual prices or indices values, but underlying momentum and trend.
Project Z is currently FLAT on EEM, having exited on 2/19 a Long trade initiated on 2/9 for a $ 1.38 gain. The EEM butterfly trade continues to improve but after a month in the trade I'm putting a safety net around the accrued profits and will likely exit if EEM falls below my 39 strike target.
Monday, March 01, 2010
I've also been fiddling with the previous XLF component Rotator and have added a few related ETFs into the mix in an effort to better discern likely momentum patterns. New to the XLF (now Financials) model are KBE (regional banks), KIE (insurance), TLT (20 year bonds) and SKF (ultra short financials). All ETFs have good daily volume and narrow spreads making them viable daytrading and/or short term swing trading candidates.
Interestingly, TLT pops up in slot #1 for the Financials rotator also.
This chart in intended to serve as an adjunct to the VIXEN layout and is designed to help forecast short term market momentum from a somewhat unorthodox perspective.