RL: Q411 earnings release May 25. Dividend 0.6% (doubled last payment). Ex-date 6/30? All time new highs?

HUN: Q1 earnings releases May5 with favorable response. Dividend 2.12%. Ex-date 6/13.

HNSN: Q1 earnings was May 4. No dividend.

SWK: Dividend 2.23%. Ex-date 5/27.

Two more risky ones: RAX, JDSU

RAX: Financials not so hot. No dividends. All time highs?

JDSU: Reasonable financials. No dividend. Telecon May 18. No dividends.

Here’s where we are now (with the my original annotations):

So having been luckily saved from entering too early, now the question is does the intersection with the middle Bollinger line? BGU is up 10% a.t.t.

until I look at $SPX in terms of bonds

which is a much cleaner chart. I guess, as always, this could go either way.

Gold has reached new highs recently but note that using looking back at the Fib levels since three years ago nicely fits the congestion points.

Bonds have been soaring

Now the ratios. $SPX in terms of bonds.

$SPX in terms of gold.

And gold in terms of bonds.

And here’s TLT (Bonds):

It seems when one is behaving well the other is not following a pattern. What are the red and green dashed times? They mark the third intersection with a trend line of the ratio. Here’s the ratio:

The trends in the ratio chart seem to be longer time scale. Does this reflect the time scale of movement of money between equities and bonds?

Show stocks where close is near lower bollinger band(10)

and Average Volume(90) is above 50000

and close is between 5 and 250

as well as a back testing engine. I’m still a dedicated user of www.stockcharts.com for annotated charts but I’m very impressed with this screening system – despite not managing to make it work well for me [yet] with only a few hours of playing.

First experience: my filters made sense to me but delivered too many false triggers that I would not have made myself. To debug I realize I need to limit the screen to only apply to one stock (one I’m familiar with) with patterns I want to use as triggers.

Here’s one year of SPY with several triggers (clean popping through the Bollinger band). So how do we program this? We need to avoid those instances where there is little distinction between the short term trend of price and Bollinger band (e.g. late Jul ’09, Dec’09. We also need to filter out events where the band is punctured due to a too rapid change in price (e.g. Jan ’10). And finally we need to filter indecisive activity (.e.g Apr ’09).

In the weekly chart it looks like we are at a critical juncture.

Notice how well the downtrend from the summer highs of ’07 and ’08 predicted the turning point. The other down trending line from the lows of those years forms a wedge which the S&P is pushing against now. This lower line is weak IMO. The dashed line, nearly parallel to the upper downtrend is much stronger. Given that, I see no real support until we reach the 200-day MA (labeled 40-week) which strengthen by the up trending channel drawn from the last winter. I think interaction with that will be the next major juncture. All assuming today marks the beginning of a return to the bear market.

So the buzz on the blogs is all about whether they’ll be a bounce or not. S&P500 punched down through the 1080 support after testing the supporting trend for a week. On previous encounters with the supporting trend, volume was above the 60-day EMA but fell below on the first up-day after the turn around. In this current situation, volume has been increasing (already above the EMA) lending strength to the downward momentum. One point in favor of a bounce is the PPO has flattened already but that’s more likely an after-effect of the drawn-out hesitation a week ago. This situation looks different than the previous four lows.

1070 is a week support. If this fails then technically there seems to be nothing to lend support until 1040 unless the trend line drawn to 1060, originating at the beginning of October, becomes a real trend (three encounters).

We’ve just had the FOMC, iPad and home sales and Thurs/Fri the biggest news are GDP and durable orders which surely the markets are expecting to come in under the mark? This seems like a good time to start a swing trade starting long.

A buy of BGU at 48 with a stop at 47.5 should do the trick. Pending no major boo boo’s from the state of the union speech tonight. Speaking of which, the futures are up during the speech so I guess I’ll be buying BGU at the open assuming it’s below 49…

…the dollar seems to have turned around…

which all suggests the market might have turned a corner (or at least exited the nine month rally).

The three year chart also suggests we the S&P might be feeling resistance…

…while the dollar could have bounced off of support three weeks ago (it might see resistance here)…

…and all the time gold keeps climbing…

.

…as does Europe?

It’s curious that in terms of gold, Europe has not really gone anywhere for two years, except down…

…which is almost true of the dollar in terms of gold which no longer looks bullish (yet)…

…and the S&P seems to not gained any ground during the so called rally?

Eurodollar index vs the dollar index has really been moving but think this double counts the dollar…

…and for completeness, the S&P vs the dollar index hides the recent change in behavior?

The story in terms of gold is most interesting. The bump up against resistance could be due to gold bouncing off of support (again) or to the S&P meeting a long term overhead down trend. Or both. Any of which might well please the bears…

and although the six day zigzag is only at 80% of the median amplitude the three day is at 130%. I think the trend line above will be an important decision point. If it breaks through this could signal a change of regime. Watch the opening hour and if there’s no panic from the unemployment numbers (8:30am) then it could signal a short term buy.

Today we saw a perfect bounce off of a supporting trend line:

Notice the general decline in volume? Does that mean anything on a rising wedge?

The amplitudes of the last high and rapid low are consistent with the overall pattern:

Which begs the question when will this pattern end? I can’t see any meaning the retracement levels so I won’t show them but the de-trended amplitudes are interesting:

What’s notable is the minimum historic change is at around the 61.8% (Fibonacci) of the trend (y=1). For the current situation y=1 corresponds to 5.94% so 61.8% of 5.94% of the last low (1041) is a minimum target of 1079. The middle of the pdf has a target of 1102. The next significant Fibonacci number of 162% has a target of 1141. This is using today’s normalization of 5.94% but five days from now the normalization is 5.72% which would adjust 1102 to 1100 and adjust 1141 to 1137. The further out that one looks then the lower the target is because the amplitude pattern is trending down.

This could all change if there is a low below 1041, wiping out Friday’s low. I think this is unlikely but you never know…

,

This shows the current leg, which had a high on Thursday, is significantly towards the upper part of the trending channel. Normalizing by the trend of the mean:

It seems reasonable to expect a top anywhere around here. It’s not impossible to climb by another factor of two, in fact the market seems to rise for no reason these days, but it is time to tighten up stops for selling.

I’ve drawn on some Fibonacci levels. The blue Fibonacci levels for the overall drop (Oct ’07 to Mar ’09) vaguely has some correspondence to support levels on the way down. I couldn’t fit any Fibonacci levels for the way up better than the puce ones I’ve shown which consider the overall retracement to be the 61.8% level of the big drop, topping at around 1230. It kind of fits but I’m not completely convinced. More curiously, the solid trend lines appear to be converging sometime in November at around 1150.

I’ve also marked in yellor, the lows associated with high(ish) volume, the highs associated with low(ish) volume and in purple the highs on high(ish) volume. There are no significant low volume lows? There does seem to have been a sea-change since the March lows. Note to self: look into whether such statistics can indicate the underlying bull/bear market direction?

And secondly introduce a new chart where I’ve normalize the above by a straight line fit (note this is semi-log space):

This illustrates what I was referring to in my last two posts; we are clearly into the upper half of normal topping territory. We could still go along way up but forming a top now and from here on would be quite normal.

There’s also a pretty well formed rising wedge in the SPX over the last seven months and the SPX is about to push up against the upper trend of this. I rising wedge is a reversal pattern but I don’t know if it applies on this time-scale?

We are currently up 3.4% from the minor bottom last week. This is within the channel I showed in the Sept 2 post but I’m noticing that chart had a lot of flutter in August which seems to try to pull down drop the lower trend line of the channel. For this reason I’m now looking at the 6-day chart (which looks better than the 1-week/5-day because of some other fluctuations in December). This chart has a fairly clean supporting trend line for the last 12 months – ever since the crash of last September. Today’s high is just beginning to push into the normal range. Thus, a topping action now would not be unusual but neither is it likely yet. Note that the majority of the dots near the trend line are red (drops) and that all but one of the black dots have been in the upper reaches of this trending channel. I thus think we have some way to go upward, at least to 5-6% before there is a moderate risk of topping. In terms of odds, over the last twelve months, only two black dots out of twelve were within 1% of this lower channel line.

Other thoughts: the utility of this down-trending channel line is ultimately limited. Once it reaches the 1-2% level it will be useless as a threshold. Could that be a point of regime transition to a different channel?

Today we had a 1.25% rise on bad unemployment news that appears to have been factored in by the beginning of the week. That factoring in led to the bottom I called on the 2^nd. We are now up 2.5% from the bottom which is already above the lower channel for amplitudes (see the second chart of last post which, by the way, has not changed). Since we are now on a new up leg but not yet four days from the bottom we could technically erase this bottom if we dropped lower quickly but I doubt that will happen. Recall, however, the intermediate goal is a leg down from the late August high.

I’ll be watching the chart to see when the up leg crosses above the 50% waterline in the channel; it may not before topping but if it does then I’ll make a call…

Here’s the four-day zigzag. The last leg was made in just two days (mostly yesterday).

The corresponding amplitude chart

With what I think is the main channel for the last twelve months. The latest move is of 5% down has only 1% left before reaching the top of that channel. Now, based on the graph, it seems uncommon, and thus unlikely, to push through the top of the channel BUT we do know that this week is wrought with bad news. Could that be the nature of the pushes through the channel boundaries?

Here, by the way, is the ratio of the legs that is what is labeled in the zigzag retracement diagrams:

It’s hard to see any pattern or tendancy to match the Fibonacci levels – to my eye at least.

intermediate term pictures? Here’s the five-day amplitude chart which gives a different story to the four day. Eyeballing the zigzag chart I think this five-day time scale corresponds to one-three week long legs. (The four-day corresponds to around one week. I think the relationship between filter timescale and behavior is very non-linear – I should look into that…)

The last leg down here is middle of the road – it could be a bottom but just as likely to become bigger.

Finally, here’s the ratio chart for 14 years using a four week time-scale. (I showed the amplitude chart in a post last week – and it hasn’t changed much). On this time-scale, we are on an up leg and it is in the vicinity of 200% where quite a lot of turning points have happened before. It’s worth noting that on this chart, big spikes are not necessarily big moves because it may be a big ratio simply because the preceding leg was small. I’m really not sure what to make of these ratios…

Using the 4-day time scale is pretty similar although the extreme channel is higher (15-20%) but using the 5-day time scale places this 4% drop near the lower end of the range BUT not abnormal for a bottom.

Bottom line: could be bottoming already for this week with a further 2% drop most likely if any. Keep in mind the high level of news coming out this week. This will prove how fallible this zigzag analysis is…

One possible scenario would be to whip back up tomorrow and Wed by 3-6% and then plummet again to fulfill the longer time scale prediction.

Note that each of the Fibonacci levels for the two month climb does have some significance over the last three months. I don’t know if that adds strength to their meaning for the next turning point but if so then the low of mid-August (976) is both a retracement and a level of support.

We seem to be testing the upper trend line of a (maybe) bull horn pattern and formed a shooting star today. I imagine the candlestick charters will start gambling on a reversal tomorrow. Futures are currently down.

As for the zigzag statistical system, I decided to not preclude the last day from being in the zigzag (I think I initially did that for technical reasons). The charts from yesterday have changed little but I did try a four day period (instead of five) and (naturally) more zigzags appeared with smaller amplitude AND today now looks like a moderately extreme top!

Here’s the four day zigzag:

And the corresponding amplitude chart:

The trend lines are parallel. The extreme events (in this short time series) appear to happen in bursts and it looks like we aren’t due for one right now. My new take – we definitely should be topping now on the week time-scale.

A few weeks ago Atilla was asking for predictions using, and explanations of, Eliott wave analysis for the current market. I got thinking about what underlies this style of analysis and here’s what I came up with – not as an explanation but a related new analysis, derived from analyzing zigzags (which at least appear in the EW plots!).

Plotted in red is a zigzag using lowest lows or highest highs for a five day period (with objective removal of cascading highs/lows, and the last day is excluded from the zigzag analysis).

Note that today’s high would natural subsume the last high of the zigzag. Next plotted is the percentage change for each leg of the zigzag. At the end, I’ve plotted the equivalent percentage change for today’s high-low from the last (green), penultimate (blue) and third to last (red) points on the zigzag. The green is near zero since today should really be the last point on the zigzag. The blue is where the last zigzag would be if we moved it to today, ignoring the last zigzag point. The red is where we would be if we similarly ignoring the last two zigzag points. True zigzag peaks rarely fall within a range of +/- n% (for this plot n appears to be 5%). We can thus eliminate red and green possibilities. If today were a new peak (blue) then it would not be abnormal but nor would it be extreme.

Indeed, based on recent extremes, the blue could move up to 10 or 15% and still be within recent norms. This suggests we could potentially reach (1.10 or 1.15)x978.51=1076 or 1125 and still appear normal on this graph. The next plot shows the same magnitude of the changes (red are drops, blacks are ups) for a broader period and with some parallel trend lines for the range of amplitudes.

The last high (i.e. today’s) is, again, not unreasonable for a top but we could go further. 10% would take us two thirds up the channel and 15% would be near the top of the channel.

Now we do the same using a four week time scale over 3 ½ years:

which looks like a good fit to the eye, and the magnitudes

.

With some hand drawn guides to highlight a pattern (deliberately hand drawn since CAD lines would reveal significant imprecision). See the pattern? (Of course – I highlighted them). Naturally, it is pure coincidence that the low of March 2009 appears to be in a line between the biggest climb and a previous weaker climb, and that today’s high is in line with the same pivot and the November 2008 plummet. Nevertheless if today, or the next 1050-1200 range does form a near term top, then the pattern would suggest a general decline in the amplitude of the swings.

Now that doesn’t fit with Atilla’s notion that the big drop of all drops is coming but it would be consistent with gradual decline, much more like the mid-2000′s.

More evidence for a top: insider trading is indicating a sale (not sure why they would precede the big funds though).

My take and the bottom line: we are quite likely to be at a top on the monthly time-scale and any indication that we have formed such, namely by a decline over a few days, would point to at least a 4% drop for the subsequent period. Note that if the current top is pushed higher over the coming week(s), the odds of the top forming increase significantly. Kind of common sense!

Finally, for the record: the same four week time-scale since 1980 shows some modulation of the amplitude of percentage changes. I wonder what drives this behavior?

Weekly S&P500 over last three years: The crash last summer dropped out of a channel (blue) and we recently broke back into it. Or did we? The orange line is an alternative channel bottom that ignores the one week false break downs out of the channel. In the case of the orange trend it looks like the index is now testing of resistance.

The weekly one year chart: There’s a congestion of support/resistance lines around 1000 emanating from the October-November bounce. These don’t look strong or significant to me. The orange dashed line is the orange trend from the three year chart.

The daily six month: We clearly broke out of the blue channel, bouncing sideways off of the lower blue line (I believe I posted to this effect on May 22). We’re in a “megaphone” pattern which has been mentioned in other blogs. A megaphone top is bearish but needs to break out on the down side to complete. The gap ups in May and July seem to fall in line with the upper trend lines for those months. The latter became a sustained move up.

More interesting on this chart is the gap behavior; a gap down or up seems to be a reliable predictor for the subsequent few days. A strategy to consider would be to watch for futures to indicate a gap and buy first thing on the day. Moving stops by the end of the day and the next would capture the sustained moves and avoid loss for the one day reversals.

I note that my two holdings right now, left over from before my extended break, are not good performers. My TXT has a cost basis of 11.33 (if I sell share for share, although in reality this is 100% profit since I previously sold enough shares to cover the cost). Before I left, I was considering holding this for 12 months to reduce taxes but note how after the big gap up, TXT consolidated while the markets climbed. I don’t think TXT is necessarily going to improve over the coming year, even if the recession is weakening.

MOST foolishly, I didn’t sell the bear ETF FAZ before vacationing and it’s now down 26% (it had a reverse split whilst away). If we really are at a local high, will the financial sector follow the market down? I bought FAZ May 12 at the last high since then. It has been tantalizingly close and seems to oscillate on fortnightly timescale. I’ll gamble it will be up again within the next month and then cut my losses.

All in all, I essentially hedged right now, loosing on one hand and gaining on the other. If the market continues up in the next few days then I’ll cash out FAZ otherwise aim to cash out when it’s at least higher than now.

Sentiment is mixed – bears have taken a beating recently but the bulls are gaining on weakening volume…

John sold his TXT at 11.99 today. Historically, that suggests TXT will really rally now. J

There’s no real indication of a direction here. At best, it looks like FAZ is in a loose channel down. I’ll hold out for a definite move up or pull out if price drops below some level I need to decide on.

On this chart I’ve added more drawn Fib. Retracement levels and support. It looks like the next support below 880 is at 865 so if Monday shows any downward movement we will have further to go but then hit real support. Given the coincidence of a support and lower channel line on Friday, I think the best odds or for a move up in the channel.

The weekly chart over the same time scale tells a similar story though here the week as a whole will determine where things are going for the next month or more.

Next stop 944. Then 1008?

I’ve been waiting for the market to end this amazing bear market rally. It’s seems like every day it’s about to dry up and then finds new energy. What gives?

On that movement, John suggests that this deal is unlikely to live past the politics of an Arab owned company building helicoptors for the military. On the other hand, this is a different economy to that in which a similar situation evaporated for port security take over. It’s possible that even the politicians are desperate enough for foreign investment in US companies that they’ll even support it. I expect that the price will hop around but creep towards that $21 target for a week or more but fail to get there when realization of these conflicts sets in. Holding on to 2/5 of the shares for a target of around 17-18.

Today was an example of stops at  work. Once I saw the price mid-morning, I moved the stop to about 3% below the point it was at at that time and sure enough it was triggered while at lunch. The volume trailed off after the morning and I could have sold ever so slightly higher but the stop was the safer way to manage it so I have no regrets.

One thing to bear in mind (pun intended); if we do see a bear trend shortly, how far down might it go. The intra-month swings seem to be of order 100-250 points, possibly putting us back to early March which would surely be a good time to start buying again.

I don’t know what this means about my being part of the class action lawsuit brought against the STLD by stock holders. I presume holding the stock at the time of the deception is sufficient to benefit, whether or not I still hold shares.

TXT on Feb 28

http://finance.yahoo.com/news/Textron-slashes-dividend-91-apf-14472054.html

The S&P500 didn’t quite make it down to the November lows so there’s room for it to drop lower.

Based on horizontal support lines we might be seeing 750 soon but the diagonal support sugars 775. We’ll know soon enough. If we see a bounce up we should buy some bargains…

Still a long way to go (closed at 6.76) and the blobs are talking of revisiting the November lows soon!

Consider a stock price, p, to be made up of just two triangular wave signals; a slow signal, s, with peak to peak amplitude S and period 2T, and a fast signal, f(t), with peak to peak amplitude F and period 2t.

Let us consider two traders who can only buy low and sell high (i.e. no short selling), who have perfect timing and, for simplicity, let us assume that they can only buy one unit of the stock for a trade (this lets us avoid the issue of compounding profits). One trader, Sam, only buys on the slow signal. In an interval 2T, he can realize a gain of S. Our other trader, Fiona, likes to trade on the fast signal. During an uptrend of signal s, she realizes a gain of F+S*t/T for each period 2t. During a downtrend she realizes a gain of F-S*t/T for period 2t. On average, she realizes a gain of F for each period of 2t but during the period 2T there are T/t such gains. That is, Fiona realizes F*T/t gains in the same period that Sam realized a gain of S. The ratio of Fiona’s gains over Sam’s gain, call it “r”, is r=F*T/(S*t). Generally, F<S but T>t so the size of r is not a given. However, if F/t<S/T then there is no signal for Fiona to trade on! The short term wiggles are not strong enough to create local highs and lows. Typically, the fast signal is very much apparent in many stocks, meaning F/t>S/T or r>1. i.e. Fiona makes more money than Sam.

One way to look at this ratio is to consider the amplitude of intraday signals which can be of order 1%. With 250 trading days in a year, the stock would need to gain 250% over six months for a Sam to make the same kind of money than a Fiona makes. How many stocks go up 250% in six months, let alone a year?

Bottom line: if there is an observable fast signal (i.e. local highs and lows on a short time scale) then even in the non-compounded case, there is more to be gained by trading these short term swings than taking the long term view. The effect of compounding makes the case even more clear cut. Hence day traders (intraday) and swing traders (days to weeks) wonder why anyone would want to be an investor.

Commissions work against very frequent trading. I chose a self-directed brokerage that gives me a fixed number of free trades (30) per month; this basically allows me to trade at the multi-day to week time scale with out “slippage”. To date, I’ve made only one or two trades a month.

But what about if 90% is the bottom? Here’s the irrational thought process kicking in. Well, what if it drops to 45%, like TXT did? The challenge is to find the right margin of safety; a stop at 99% would almost always be triggered because prices vary by a couple of points during the day. To this, I don’t yet have the answer and I need to do some analysis fairly soon to change that.

Lesson: Do not be lazy or nonchalant; choose a stop limit at the same time as a good buy price.

My limit buy order at 12 placed on the 23rd stayed in place and activated at 11.33 during that precipitous plummet. I’ve been waiting for the bottom to form and I think it formed yesterday. I said as much to JD but evidently I wasn’t confident enough to place an order over night. I missed the first hour of trading where the price rose steadily from 6.25 to shy of 7 so I just bought another x1.5 shares at market (6.76). This brings my cost basis down to 8.73 (green arrow).

I’d really liked to have put even more money down but now I have 25% of my account in the one stock which is getting rather risky.

Go TXT, GO! [Please]

Solid companies: In the likely event I make bad calls on timing (i.e. buy a high instead of a low), I only want to buy stocks that are also good long term investments. i.e. are stocks that should grow over the next year.

Buy low, sell high: Not sell high, buy low. I intend to only buy long, rather than sell short. I know the drops are faster and there’s more money to be madeas a bear but selling short has no bound on the risk (there’s no upper limit on how higher a price can go when you are gambling the price will drop). Buying a stock only puts at risk 100% of the cash used to buy the stock (not more than >900% for short selling).

Sell at 20% profit: Cash in when I’ve made the 20% profit. I doubt I’ll ever pick that stock that will double it’s value in a month. So I plan to cash in when the stock gains 20% and look for new oppurtunities.

Be rational, not emmotional: This, I’ve read, is the hardest part. Not to panic. Not to send good money after bad (there’s the pun for the blog name). Not to let emmotions disctate bad decisions. If a stock is under valued would I buy it? Why would that change if I already own the stock at a loss? Should I sell at a profit now or wait for more profit?

I will not chase: Following from the above, if I miss an oppurtunity (i.e. a bottom), I will try to avoid chasing and just wait for the next oppurtunity.

Placed a limit order, good until cancelled, at 12, on the chance I might catch an intraday dip. I was feeling like I’d missed the train on the real dip earlier in the week (around the 23rd) [don't chase!!!!] and all the news seemed positive and bullish.

Perfect! Sold at 16 (3.5/12.5 ~28% profit) on Jan 6th. Almost timed it perfectly on the day too. Beginners luck?

Watched the drop during the 23rd, saw upward buying during the day and bought at 12.5. I was feeling proud of finding this wedge/triangle in the charts but wasn’t really sure what it meant.

Sold STLD on 12/9 for 10.5 (1.2/9.3, ~13% profit) for no rational reason other than wanting to cash out after the severe roller coaster ride down to below 6.

Lesson: should have watched for another day before buying; would have later been able to buy at 6.

I put in a limit buy order at 9.3 on the morning of Nov 10 based on this chart. The blue line was my anticipated support line. News from politics suggested massive infrastructure investment which should be good for steel.

For the record, I started this blog on Feb 5th, 2009 but started investing/trading in November, 2008. I have created posts for all my past trades and modified the date for those posts. Hereafter, posts will be honestly date stamped.