10-19-2013 Comments

[Mike]

I am detecting super-exponential movement in the markets. This can lead to bad news as it means more than just an overbought situation is before us. The chart below depicts three times that the price curve bent upwards on a log scale chart. A straight line on a log scale chart depicts exponential growth, thus these are super exponential. So some cyclical bull markets end in a blow-off phase, the prior two did and the jury is out for today.



The chart also depicts what the first year in a new cyclical bull market looks like (powerful). Some people believe that because we had a 20% correction in 2011 that that was a bear market and we began a new cycle in late 2011. I think the preponderance of evidence suggests that that was just a big correction in on ongoing and now 4 1/2 year cyclical bull market. This is old and the super-exponential phase we are witnessing may mean that we are approaching an endpoint.

The only work that I am aware of that attempts to define an endpoint in a super-exponentially growing market is Didier Sornette's Log Periodic Power Law technique. For what its worth here is an attempt at the analysis.
The critical time parameter (tc) that produces the best fit suggests that we are now arriving at a regime change (actually Monday is the critical date).



I will say that October 31, 2007 had a day that is reminiscent of what Friday looked like. That was the end of a super-exponential growth phase at the end of a cyclical bull market.

I am uploading a new watch list in the stock selection area. Because of the overbought situation most stocks I would want to own are extended beyond safe buy points. This makes the list very short. I am listing here all of the stocks that I do want to own if a buying region develops. I own some of these already. These stocks have good earnings and sales, most show accelerating earnings, they show institutional buying and have been behaving well. They are just mostly not at a safe region to make a new purchase. We need some kind of a pull back. This will be tricky because the implications of the top of this post is that we just may get a pullback that doesn't stop for a while.

ACT
BLOX
CBI
CLR
CRZO
FB
INVN
KORS
NQ
NSM
NUS
OAS
OCIP
OII
PCLN
PRAA
PRLB
QIHU
SFUN
TSLA
VEEV
VRX
YNDX

[Chern]

The last time you made this prediction we certainly got a pullback. Based on this work why not short? The risks seems low a few percentage points the reward huge if the expected large pullback develops.

[MTman]

Mike - It seems that parabolic moves end with stocks having climax top type action as defined by O'Neil in HTMMIS. The only stocks that seem to be hinting at this might be some of the oil and gas plays. Until we get 1-3 weeks of huge moves in the leading stocks (which would indicate climax tops), it seems these quick pullbacks are simply shakeouts. The Put/Call ratio seems to indicate we are due for another shakeout. Your thoughts?

Mark

[Mike]

Mike - It seems that parabolic moves end with stocks having climax top type action as defined by O'Neil in HTMMIS. The only stocks that seem to be hinting at this might be some of the oil and gas plays. Until we get 1-3 weeks of huge moves in the leading stocks (which would indicate climax tops), it seems these quick pullbacks are simply shakeouts. The Put/Call ratio seems to indicate we are due for another shakeout. Your thoughts?

Mark

You could be correct however almost every stock that I would like to own is quite extended. The last two times the Log Periodic Power Law suggested a regime change led to short term pull backs which indeed was a regime change. There just seems to be too much federal reserve pumping up the markets to get too excited about a long term decline at the moment. So we might pause here and go on to new highs, or perhaps real weakness will develop. As to the shorting comment, most good short positions develop 5 to 7 months after a stock tops. If we are topping out here there will be time.

[Chern]

Mike thank you for your comments. I am still somewhat confused. I thought Sornettes work was designed to detect a "crash", hence my question, why not short the index here with fairly tight stop, take a small risk on the market going higher(granted likely with this FED liquidity) but actually bet on the crash which would be huge though low probability upside given how the curve looks. Thank you again for sharing your work.

[Mike]

Mike thank you for your comments. I am still somewhat confused. I thought Sornettes work was designed to detect a "crash", hence my question, why not short the index here with fairly tight stop, take a small risk on the market going higher(granted likely with this FED liquidity) but actually bet on the crash which would be huge though low probability upside given how the curve looks. Thank you again for sharing your work.

Sornette originally used the word crash but has since moved onto another description of a regime change. The two past times I detected super exponential growth led to short pull backs. Perhaps that is just the nature of what is going on with the FED buying programs. I noticed that quite a few stocks made new highs today, not much of a regime change so far. TSLA however is below its ten-week moving average. Sornette did indeed short an index on the 1987 crash when he predicted it. I am just more of a stock picker and QE is still here. RAX is more of the kind of set up for a short that I prefer.

[Pascal]

Mike, would you have time to see how this Didier Sornette exponential pattern matches the current environment?

We had a pull-back in March and I'd like to see if this pull-back was fitting the Sornette curve and then mostly what possibly could happen down the road. Is it a break and then race to a new high or a break and then bounce and then fail again?

Also, do you think that we are in the process to an exponential move up in the S&P500 and not just the Momo stocks or the Nasdaq?

As you can see in the attached Figure, the pink curved trend line corresponds to a QE infinity that is slowing down somehow. That would indicate a choppy market for the next months. Of course, we could also see an "end of QE" type of move such as we saw with the green and blue curved lines.

The other aspect that we might consider is a "buy all and everything" with free recycled money out of China. The hypothesis is that Chinese oligarchy, Chinese funds and Chinese companies with international activities want to park funds out of China because they know that their real-estate market is imploding, which means deflation and more Yuan printing. Buying anything in US$ is better than keeping funds in Yuan.

My hypothesis is that China is exporting mal-investment to the US, which is bidding everything US$ based, disregarding value (to the continuous chagrin of John Hussman)

http://www.hussman.net/wmc/wmc140602.htm

Rumors of a European hidden QE next week is also good fuel to the fire.


Pascal

I am detecting super-exponential movement in the markets. This can lead to bad news as it means more than just an overbought situation is before us. The chart below depicts three times that the price curve bent upwards on a log scale chart. A straight line on a log scale chart depicts exponential growth, thus these are super exponential. So some cyclical bull markets end in a blow-off phase, the prior two did and the jury is out for today.

Attachment 20376

The chart also depicts what the first year in a new cyclical bull market looks like (powerful). Some people believe that because we had a 20% correction in 2011 that that was a bear market and we began a new cycle in late 2011. I think the preponderance of evidence suggests that that was just a big correction in on ongoing and now 4 1/2 year cyclical bull market. This is old and the super-exponential phase we are witnessing may mean that we are approaching an endpoint.

The only work that I am aware of that attempts to define an endpoint in a super-exponentially growing market is Didier Sornette's Log Periodic Power Law technique. For what its worth here is an attempt at the analysis.
The critical time parameter (tc) that produces the best fit suggests that we are now arriving at a regime change (actually Monday is the critical date).

Attachment 20377

I will say that October 31, 2007 had a day that is reminiscent of what Friday looked like. That was the end of a super-exponential growth phase at the end of a cyclical bull market.

I am uploading a new watch list in the stock selection area. Because of the overbought situation most stocks I would want to own are extended beyond safe buy points. This makes the list very short. I am listing here all of the stocks that I do want to own if a buying region develops. I own some of these already. These stocks have good earnings and sales, most show accelerating earnings, they show institutional buying and have been behaving well. They are just mostly not at a safe region to make a new purchase. We need some kind of a pull back. This will be tricky because the implications of the top of this post is that we just may get a pullback that doesn't stop for a while.

ACT
BLOX
CBI
CLR
CRZO
FB
INVN
KORS
NQ
NSM
NUS
OAS
OCIP
OII
PCLN
PRAA
PRLB
QIHU
SFUN
TSLA
VEEV
VRX
YNDX

[Mike]

Pascal,
I visited Didier Sornette's Financial Crisis Laboratory this morning at: http://risikopedia.ethz.ch:2375/#http:// which has the following data:



Shown is a warning on S&P500. When you click on the red S&P500 warning the following image pops up:



This shows the confidence level over time that the S&P500 price data fits a Log Periodic Power Law signature. When I used to run the LPPL fit myself in excel, lacking the mathematical skills I never knew how good a fit I had achieved. Note the highest confidence on this chart was end of 2013 when I was reporting a possible fit. Didier Sornette is trying to address the confidence issue with the analysis he shows on this site. I don't really know how confident 0.11 level is. So I look at history that is available on the site for previous bubbles. The following is from 1987.



The middle historical chart is the confidence level chart for comparison with today's chart. What is common amongst the various historical bubbles is that the confidence level show multiple peaks on the way up (as we show now). From looking at other examples 0.11 really isn't all that remarkable, the historical examples showed higher confidence.

May is now in the rear view mirror, a historically weak period. The one chart that has me the most worried is the Russell 2000, shown here:


The small caps have formed a possible H&S pattern which so far has found support at the neckline. Last week the index gapped up on light volume and now sees the 50-day as resistance. Friday was a small cap distribution day on a day that the S&P saw accumulation. The S&P visually looks like stalling to me. It made smaller upward progress on substantially higher volume than the day before. It doesn't quite meet the MarketSchool definition of stalling however, missing by a small factor (up 0.18% vs. 0.10%). In 2007 the small caps were exhibiting a similar characteristic by lagging the general market from July onward to the October NASDAQ top. So my guess is that either the Russell is going to have to rally to new highs or it will eventually pull the rest of the market down. The Russell for me is the Canary in the coal mine.

I also note that the MarketSchool Exposure Model did not agree with the IBD FTD call and is at zero exposure level with the buy switch off. The exposure model however does not look at the S&P500 which has rallied to new highs which by itself would turn the buy switch on. Momentum investors however are better off following what the NASDAQ has to say. We seem to be in a narrow large cap rally which has me sleeping lightly at the exposure level I have in the market today.

[Pascal]

Thank you Mike.

I see on the Russell 200 that the RS index is almost back to its lowest level. This shows that the money is looking for a combination of yield and safety by bidding the large caps up.


Pascal