The 5-day ARMS Index ($TRIN) closed at 0.6 today. This is the lowest 5-day reading since 2003. Ultra low readings (which suggest the market is way overbought) usually occur early in rallies coming out of deep bottoms. It's very rare to see such a low ARMS reading at a top -- in this case, a triple top vs. the April and November highs.
Subscribe to:
Post Comments (Atom)
![[Most Recent Quotes from www.kitco.com]](http://www.kitconet.com/images/live/s_gold.gif)
![[Most Recent USD from www.kitco.com]](http://www.weblinks247.com/indexes/idx24_usd_en_2.gif)
Posts
Parker, I think in one of Terry's previous Sunday updates (a few months ago) he mentioned that at some point, after the expiration of the last T, there would come a time where there would be an extermely low 5 day average on the trin. An average that would be below 1. I think he even said it would come after a sick/silly looking rally. I'm certain I remember him saying that. I believe he said that when that occured, a collapse would be immenent. It is possible that finally the bubble that we all know the S&P is might finally be ready to burst?
ReplyDeleteHunt ditto to that.
ReplyDeleteI notice that you're taking the average of the last five days for your reading of 0.6. If you take the raw reading, which is:
ReplyDeletesum of advancers
-------------
sum of decliners
divided by
sum of adv volume
---------------
sum of dec volume
= you get a reading of 0.72.
A reading below 0.75 has happened 90 times at swing highs (which I defined by looking at SPX at 20-day highs combined with the low TRIN readings) since 1990.
Most recent occurrences were:
11-Mar-10
27-Aug-09
07-Aug-09
18 & 19-Sep-07
21-May-07
I think if you look at those dates on your charts you'll notice some good places to sell, however the selling pressure across all 90 occurrences was primarily contained to the first 10 days, on average.
Cheers
Hunt
ReplyDeleteThanks for bringing that up. Maybe Terry will comment on Sunday.
Thanks High Plains. Interesting stuff.
ReplyDeleteYes - I look at a 5 day simple average. It's the path of least resistance, since it's quick and easy on Stockcharts.
I wonder what a weighted average of the raw data would show, weighting recent days more heavily than than earlier days.
Another version of the Arms Index I look at is the 5 day EMA.
I think Terry said he was looking for the 5day sma to come in at .70 or lower to help confirm the short term rally before the expected drop off--
ReplyDeleteanyone see this?
ReplyDeleteBen just want to QE the market even when it rallied hard..
http://finance.yahoo.com/news/Ben-Bernanke-on-60-Minutes-cnnm-1490415534.html?x=0&.v=1
High Plains, thank you for the analysis. Two quick questions for you: 1) In your analysis did you look at how long, on average, it took after the trin reading low for the correction to start? 2) Did your analysis include how deep the corrections typically go in either points or percentage? Thanks again!
ReplyDeleteHunt,
ReplyDeleteshort answer, 1) immediately, and 2) not very far.
The results of my study probably require a separate blog post to shed the proper light on the results, which I'm not going to do. I only wanted to complement Parker's look at the 5-day average with my posted comment.
But I will say this, since we are looking at a five-day average of buying pressure via the TRIN, a pullback from this buying pressure will only have an impact for the next few days as the pressure eases or reverses. Any selling pressure from there will most likely be due to longer timeframe issues.
As Terry Laundry has said in his weekly postings and Parker is highlighting with this post, the 5-day TRIN reading is just another tool to reinforce or confirm what you are seeing with the Ts. If an end date coincides with a blowoff reading in the TRIN, it merely enhances your confidence that your pegged end date is correct.
Cheers
Parker, I may not be applying the technique correctly but it looks like ther is a small Vol oscillator T with A at the Nov peak and a bottom in the recent 1170 area. Using B as the first low and C as the last low on the SPX (rising bottoms oscillator may be better treated as a V than W but the timing does not work) gives D on or about the 9th.
ReplyDeleteI see that too, Luv
ReplyDeleteA = Nov 5
B = Nov 16
C = Nov 30
D = Dec 9
interesting stuff guys
ReplyDeletequestion: what market action would make ttheory or the money-flow T miss on this next call?
ReplyDeleteIf there is no significat (>30+) pt drop by 12/17 or so?
Given the strong market action over the last few days, do you guys really think we will have a sell off next week? A small pull back? Sure. But we've seen this before - huge overbought readings coming out of a bottom that just speaks to more strength.
ReplyDeleteRobW
ReplyDeleteThe cycles indicate an interim top next week, followed by a correction, then more strength into 2011.
We'll see if we get a top, and if so, how deep the correction is.
Rob, I agree. Uptrends usually begin and end with large moves. I'm a shorter-frame trader (days to weeks), and I see after the last couple of days that the market needs to digest that move a little, but that the bigger selling appears to be over for the moment. We're still above the 56EMA and until we break that, I'm bullish and will be buying on pullbacks.
ReplyDeleteI looked at TRIN <= .70 since 1992. Here are the 14 cases.
ReplyDeletedate price&percent #days nh=newhigh soon
7/13/10 1099->1057=4% 5d nh
8/25/09 1038->992=4% 6d nh
8/7/09 1018->979=4% 6d nh
12/8/08 919->851=7% 5d nh
8/20/04 1100->1093=1% 2d nh kickoff
3/18/03 867->896->844=3% 9d nh kickoff
10/14/02 844->nh kickoff
3/3/00 1411->1347=5% 3d nh kickoff
3/16/00 1458->1477->1446=1% 3d nh kickoff
11/16/99 1420->1425->1389=2% 9d nh
1/11/99 1264->1207=5% 2d nh 1d after top
5/5/97 830->812=2% 3d nh kickoff
9/17/96 686->679=1% 2d nh kickoff?
5/19/93 448->445=1% 2d nh kickoff?
Friday's TRIN reading was one of the lowest of all time. As much as I hate to say it since I've been expecting a significant November top, all 14 cases had bullish outcomes with a max 7% pullback and average 3% pullback. As many as 7 of the 14 cases were part of what I'd loosely classify as kickoffs coming within a month or so of significant lows (after usually 10%+ and/or multi-month 5%+ corrections). That is not the case today considering an August or July low. 5 of the 7 kickoff cases did not pullback more than 2% with one of them never looking back. In 2 of the 14 cases, the actual day of the TRIN reading was a slight down day so I used the following small up day as the price high. Only 1 reading came after the top (by 1 day). Friday was not a down day or after a top, so 1225 is the price high to use. I should also mention that there were a few TRIN readings in the .75 area that marked intermediate-term tops but not <= .70. And, almost all of the 14 cases marked a price high that day or only slightly higher the next 1-2 days. Timing-wise, 2 cases took 9 days to reach a low while 7 cases took 2-3 days and the 4 most recent non-kickoff cases since 2008 took 5-6 days.
So, using this indicator alone based on a limited data set, we can anticipate a few things. SPX is unlikely to blastoff next week but could possibly test 1230-1240. If SPX approaches 1240, the pullback is likely to hold at 1210-1220. If SPX cannot break through 1230ish, it has a reasonable chance of holding at 1195-1205. However, given that only 1 non-kickoff case made a shallow 2% pullback while the other 6 fell 4-7%, the more likely SPX target would be 1140-1175. And, the pullback should end within 9 trading days with 2-6 days for almost all cases.
To summarize, that means SPX should pullback to 1210-1220 if it approaches 1240 next week OR 1195-1205 and most probably 1140-1175 if it cannot break 1230ish next week. And, that pullback is likely to end by December 13th or OPEX December 17th at the latest.
Prior to this week's rally, I had numerous reasons for 1145ish to be reached. Unless SPX can reach 1240 next week, that seems plausible to occur OPEX week if 1195 is broken. That might fit Terry's mid-December null echo low but it would not hit his lower band expected to be near 1120 at that time. So, if 1120 is somehow reached, that would obviously overlap the 1129 high and deviate from historical TRIN behavior likely signaling the next big bounce as the dead cat variety. If either 1195 or 1140 holds, SPX is likely to make a new high in January based on history. Good luck.
Thanks for the thorough analysis Stu!
ReplyDeleteTICK exceeded 1000 Friday. That does not happen very often and typically leads to a 2-4% pullback within days but sometimes occurs on a large up day within the first 3 days of a significant bottom suggesting a kickoff. The kickoff scenario is possible but unlikely in SPX, since Friday was not a very big up day and it was the 4th rally day after a mere 4% pullback. When combined with a small NYAD change near a NYAD high as we had Friday, the TICK extreme indicator works even better. However, this signal is not meant to catch intermediate tops, just short-term ones. Good luck.
ReplyDeleteISEE (options sentiment index at http://www.ise.com/WebForm/viewPage.aspx?categoryId=126) reached a bullish extreme Thursday of 183. Here's what that meant in the past.
ReplyDelete7/28/10=182 was part of a 1121-->1088=3% drop
4/15/10=185 led to a 1214-->1184=2.5% drop, then a slight new high, then the flash crash
11/24/09=184 led to a few pts higher the next day and then a 1111-->1084=2.5% drop and a 1084-1120 trading range for 4 weeks
10/14/09=182 led to a few pts higher for 3-5 days and then a 1101-->1029=6.5% drop
3/9/09=220 led to the massive rally from 667=kickoff move
12/29/08=206 led to a 5-day blastoff to the new year of 873-->944 but then a 2-month near-30% drop ensued=kickoff from the huge November lows followed by a collapse
Thus, ISEE currently suggests a short-term 2.5-6.5% pullback, since we are not likely in a kickoff as we were in Dec 08 and Mar 09 after massive lows. But, that's not even the most interesting part. ISEE Equities only, which is considered the dumber money, hit its 2nd and 3rd highest level in 3 years on Thursday and Friday. Here are the highest 8 values in the last 3 years.
12/1/10 and 12/2/10 (ISEE eq=327 and 296)-->???
4/15/10 (ISEE eq=348) led to an immediate 2.5% drop followed by a slight new high a couple days later and then a 15% drop with flash crash mixed in
4/5/10 (ISEE eq=276) led to a 0.3% 1-day rise followed by a 2-day 1.5% drop and then a 3.5% bounce into the April 2010 high
10/29/07 (ISEE eq=275) resulting in a 2-day 0.5% rise followed by a 17-day 10% fall
10/8/07 (ISEE eq=279) led to a 3-day 1% spike followed by a 9-day 6% fall
7/12/07 (ISEE eq=275) led to a 0.5% rise for 2 days followed by a 22-day 12% drop
6/15/07 (ISEE eq=280) led to an immediate 8-day 3.5-4% drop followed by a slight new high and 12% drop
So, ISEE equities-only paints an uglier picture. The dumb money is even more juiced up than the smart money at 3-year extremes. One of the next highest values occurred a few weeks before the fat finger mini-crash in early 2007. You'll notice 4 of the 6 previous highest readings led to a 0.3-1% rise for 1-3 days. From Thursday, we have a 1-day 0.4% rise thus far. You'll also notice 3 of the 6 previous highs fell back 1.5-4% before a making a slight new high and collapsing. That would fit a 1230ish-->1200ish type scenario followed by 1240-1250 and then 1100-1150. However, that would not fit Terry's null echo low in mid-December. 5 of the 6 readings ultimately led to 10-15% drops weeks later. Perhaps today's scenario is closest to the 9-day 6% drop at the October 2007 top. I think it would take a near-immediate drop of at least 6% and possibly 10%+ to satisfy Terry's null echo low and his lower band. It's not guaranteed but the ISEE conditions certainly support that happening. Good luck.
Good work, Stu! Thanks for sharing.
ReplyDeleteThe AAII sentiment survey bumped back up to 50% bullish, near the 58% extreme 3 weeks ago, with bearish sentiment at 24.68% last week essentially tying for the 2nd lowest reading of the year. The bull-bear spread is 23.4% currently, and that makes 10 of 12 weeks in the 19-30% range. You should know what that means in recent history. There were 2 of 3 weeks above a 20% spread leading into early January 2010 resulting in a 1-month 9%+ SPX pullback. There were 3 weeks of a 19%+ bull-bear spread leading into May 2008 resulting in a 2-month 17% pullback. There were 2 consecutive weeks over 20% leading into October 11, 2007 and you know about the 11% 6-week drop then. There were 2 of 3 weeks above 20% 4 weeks before the Fat Finger February 2007 episode resulting in a 7% drop and the real estate sector top. There were 3 of 4 weeks above 20% leading into early November 2006 resulting in a +/- 4% 5-month sideways stock market with the Fat Finger episode mixed in. There were 6 consecutive weeks of 20%+ leading into mid-December 2005 at which point the market went +/-4% sideways for 10 months. So, you can see the most recent cases with multiple weeks of 20%+ bull-bear sentiment have been very bearish while older cases led to many months of sideways action +/- 3-4%. However, none of my examples stayed so bullish for so long as today. So, let's step back to early August 2005 when 9 of 11 weeks (bingo) were above 20% resulting in a 6-7% SPX drop. And, leading into year-end 2004, bulls exceeded bears by 20%+ for 9 consecutive weeks resulting in a 2-3% rise for 2 months and then a 6-month 9% drop. All in all, during the more bullish 2003-2007 time period, sentiment like we have today at best led to months of sideways action with the possibility for 5-10% downside while, during the 2007-2010 time period, downside has been in the 7-17% range. Good luck.
ReplyDeleteI did not do all that research this weekend. I have presented some of that before at my blog at s2trading.blogspot.com and just added to it. I have been trading for about 5 years and am improved but still have a lot to learn. My system has performed much better than me but that's also a work in progress for about a year. I am not looking for money or members...just saying in case you wonder who the heck this guy is presenting all this research all the sudden in Parker's blog which I came across through Terry's site.
ReplyDeleteAnyway, as you can see from my previous posts here about ISEE, TRIN, TICK, NYAD and AAII, a short-term pullback to SPX 1200ish seems highly likely even if SPX goes a little higher Monday. And, a 7% pullback to 1145ish has very good odds if 1195 is cracked. A 9-10%+ pullback to 1090-1120 is less likely but possible. That analysis is based on multiple confirming indicators, not a crystal ball. I believe USD, VIX and my discretionary spending analysis support the same thing. However, the indicators are conflicting in terms of whether or not the 1227ish high is an intermediate-term high. If SPX breaks 1130, I'd heavily lean towards the next large bounce failing to make a new high. Otherwise, an SPX drop supported at 1145ish would probably allow for a new high or failed retest in Jan 2011 (like Nov 2007) depending on how the indicators and other markets shake out. Terry's theory seems to expect a bearish drop more in line with 1145ish or 1100ish. In any case, I expect some crazy action in the next 10 trading days. Good luck.