Beware Extrapolating from One Data Point

CNBC had a robust session on the housing data Monday morning, concluding that housing had bottomed and the bounce was on! Their single data point: a 23% bounce in June over May. No mention that:

• this was the worst June on record (back to 1963)
• May got revised downward 11%, a pretty steep change
• the 23% gain in June was elevated due to the May reduction
• April got revised downwards, too, by 15%, an even steeper change
• June was 17% below a year ago, June 2009
• June was the second worst month on record
• May was the worst, the all-time record low (back to 1963) for any month 

To the contrary, a commentator asserted that "housing bottomed in the first quarter of 2009." He must have failed math. Maybe now he can make the claim, after the two worst months on record. Some discussion that it follows a really poor May, which plummeted after the homeowners credit expired for new homes in April. The perils of government intervention. A reversion to the mean snapback was to be expected in any event. More data at Calculated Risk, a great source on housing.

CNBC was not alone in this irrational extrapolation. Bloomberg reported:

Sales of U.S. new homes rose in June more than forecast following an unprecedented collapse the prior month, a signal the worst of the slump triggered by the end of a government tax credit is over

A different but equally misleading error shows up at the normally reliable Pierre du Plessis blog. I think it inadvertent, since he is exploring an interesting issue. The ECRI indicator has dipped below -10%, which has always signaled a recession. The inputs to ECRI are a black box, but Pierre uses correlation with the S&P to argue that ECRI really tracks the S&P more than a cluster of leading indicators. Or maybe it is the S&P which tracks ECRI's indicators? It makes a huge difference, since the hoary old saw is that "the market predicted six out of the last three recessions."  ECRI may be more accurate than the S&P.

A third data point getting a lot of press is the Chicago Fed's Index, which shows a decline in June. In my most recent Double Dip Countdown post, I noted how the Empire State (NY) and Philly reports had shown a slowing of growth, but not a decline. The Chicago survey is national, and the decline is a bad sign - yet still just one datapoint, and one that needs to be understood: a decline means growth below the historical trend, not an outright contraction. The three-month moving average has gone to -0.05, but it needs to get below -0.72 to reliably signal a double-dip. 

 

A final pair of datapoints come from the regional Feds. The Dallas Fed Manufacturing Index, which was suppose to improve from -4 in June to -2.5 in July, instead fell to -21, the lowest level since July 2009. The Richmond Fed also reported a drop from 23 in June to 16 in July. Again, these drops do not signal contraction, just a slowdown in growth in manufacturing.

Individually, these single factoids can be used to create a bullish or bearish picture. Taken as a whole, they show consistent evidence of a second half slowdown, if not a double dip. 

Comments

[Roger D.]

Hi Guys still here.

MCD,put a fork in it. There's a crash in our future.

http://www.screencast.com/users/parisgnome/folders/Default/media/14851b23-1b29-435e-a4e1-dd901448734d

Roger D.

Wait in the 2nd qtr GDP blows big Friday. Consumer confidence is heading down the toilet.

[Rogers Dad]

"MCD,put a fork in it. There's a crash in our future." - Roger

My son, please stop playing on the Internet during Summer School. I really wish that you would attend to your classes so you won't be put on academic probation again. Your mother and I are paying good money for your education, and yet you continue to cry out for attention on Internet blogs ( like this one ) making all sorts of absurd claims and wasting everyone's time with your constant drivel.

[Chanug Tzu]

Is everyone ready for July 30 - August 1 ?

http://www.youtube.com/watch?v=_XqQYjrUTiw

[twitter.com/DrBubb]

"MCD,put a fork in it. There's a crash in our future."

If so, that must be one of the most significant $1 moves ever!

[Chanug Tzu]

Mars Uranus cycle is no joke! (It hits next week)

Check out the charts

http://www.houseofcharts.com/images/stories/Crawford20A.20-20Mars-Uranus20Cycle20in20US-Stock20Prices.pdf

[Chanug Tzu]

Quote from PDF file "EVERY ONE OF THE MAJOR VOLATILE BEAR MARKETS SINCE 1920 TOOK PLACE IN THE LATTER HALF (RECEEDING PORTION) OF THE MARS/URANUS CYCLE!"

[da bear]

Chanug Tzu,

looks like Mars Uranus cycle hit in 1937 AND 1987.
Next week could be peak week for this corrective rally.
The high of the day was about 10,580, so 86 points from my DOW 10,666 target.
In my estimation only about 300 points upside from that level. No real reason to be long here.

gold is weak today. I checked the 1987 stock chart yesterday, then compared it with a recent gold chart. Looked similar. A steep run-up then a sharp sell-off, then another rally that sort of runs out of steam then -- BOOM!!!!!!!!!!

da bear

[da bear]

great post yelnick!
This housing mess is about to take us down ... AGAIN.

da bear

[Mamma Boom Boom]

Remember, around the Porch of July weekend, I mentioned that an air pocket had formed under gold?

Voilà!

[Waver]

Tomorrow is the 9th trading day off the July 2nd low, according to EWI and the S&P has been going through 9 day cycles.

Stay tuned!

[Astro Bob]

As an astrologer, I'm kind of surprised and disappointed that Crawford would make such a glib offhand statement about the Mayan calendar indicating the end of civilization as we know it, as there is very little evidence to support such an assertion, either astrologically or archeologically.

[-Anikitos]

Astro Bob, there are a lot of dipsh1ts out there that take this seriously. Enough people can create a critical mass and become a self fulfilling prophecy. I don't believe that we are governed by stars, just by our beliefs that we are.

[Michael]

"This housing mess is about to take us down ... AGAIN." - da bear

If you were to assume that your claim is true, what is so significant about housing at this point and time that will be the catalyst for taking the markets down... AGAIN???

I would suggest that falling off of a 2-foot step "ladder" is nothing compared to the 12-foot step "ladder" that housing USED TO BE.

[Hank Wernicki]

Great Post Yelnick ...

Beware Extrapolating from One Data Point << " THIS INCLUDES THE MARKETS TOO ! "

Hank

[da bear]

Michael,

The fall should be measured in percentages. lol
To carry this analogy further than it should be, if you own a paid off house and prices decline another 20% (our two foot step ladder) then you are down 20% if you had to sell, BUT if you had only 20% equity then you would be wiped out aka you fell 100% off a 2 foot high step ladder. And that hurts.
Final analogy, suppose a 150 ft. step ladder (aka a ten story building) shows up representing "the American commercial real estate market" and suppose Jack and Jill, who already fell down the hill (Florida condo collapse in 2005) and wish not to get hurt again, decide to step down from the proverbial mid-rise step ladder and turn the keys into the bank.

That is our current situation. And yes, the banks are now Humpty Dumpty.

da bear

[DG]

Hi Guys still here.

MCD,put a fork in it. There's a crash in our future.

http://www.screencast.com/users/parisgnome/folders/Default/media/14851b23-1b29-435e-a4e1-dd901448734d

Did you seriously just label two waves as "5a" and "5b"?

You don't have to belong to the NeoWave school of wave theory to know that's not proper Elliott Wave. Why must you be a parasite on a theory you obviously don't make any effort to understand? Anyone with even a little understanding of Elliott will know you don't know it and those who don't understand it will (rightfully so) look at your market calls and run as quickly as possible from anything called Elliott Wave.

Maybe that's your game, trying to discredit Elliott Wave by putting the crappiest possible counts out there. If so, bra-fucking-vo, dude, you are doing a bang-up job.

[Chris]

5a and 5b is, indeed, proper Elliott Wave might it might not be Elliott Wave as you know it, DG. One of the beautiful things about Elliott is it's open to a variety of modifications and interpretations. No single school has developed a very reliable method. Roger, however, has come about as close as any and has been generous enough to share his views here. Rather than risk scaring him off with inflammatory language and naughty words, would you kindly muster a "thank you" or at least stay silent?

Some of us appreciate his ideas and have actually made a pretty penny off them.

[DG]

One of the beautiful things about Elliott is it's open to a variety of modifications and interpretations.

Yeah, right.

Roger, however, has come about as close as any and has been generous enough to share his views here.

Yeah, right.

Some of us appreciate his ideas and have actually made a pretty penny off them.

Yeah, right.

[Wave Rust]

waver

you might want to restate or re-count the 9th day thing

16?

wave rust

p.s. i'm short for trade ,,,, tomorrow should be down ,,,, maybe cover some gaps ,,,, looks like an a-b with the c tomorrow and maybe thursday morning ,,, spx 1057 gap, dow 10121 ,,,, monster reversal up from whereever it stops ,,,, dont play touch and go with a low flying cruise missile

[Wave Rust]

da bear
consider this possibility,

gold has been running the same simple corrective down pattern for a long time.

off a top is a deep spike then a gentler roll south which finishes the 'a'. then the bounce for 'b'. then the 'a of 'c' begins with a deeper spike and gets it well oversold which sets up the divergence in 'c of 'c'.

looks like 'a of 'c' is done now, so bounce then the 'c of 'c' lower low but diverging momentum buy ,,,, late this week or early next maybe.

then off to 20 hundred again

it's like gold pong :)

wave rust

[Chris]

Very mature, DG. Try trading off Roger's calls. Your mood will improve.

[da bear]

Waverust,

ah. Yeah the Tuesday session had a steep sell-off, down $20 so now I think gold needs a small retrace of that loss then a retest of the lows. I guess you can count that as a small iii, iv, v or as a small a-b-c move.

The move after that will determine the next big move in gold. A break below $1,150 tips the scales towards the big 'C' of 4 down. A move up next week would allow for a new nominal high. I am not sure that gold has put in a full five waves off its Fall 2008 low near $690.

da bear

[Wave Rust]

da bear

its all overlaps from 690. the 2008 low at 690 was the end of the 2 wave ,,,, 38% retrace ,,,, my guess is it's still in 1 of 3 and overlapping ,,,, if so, a long long way to go ,,,, 2300 may be too conservative.

oh, and it isn't anything like '87 ,,,, at all.

gold has become an economic teddy bear, tightly clutched by all who realize America can and does elect idjits to run a country into the ground.

it happens about every 24 years ,,,, then we finally right the ship of state, and elect a reagan or an eisenhower

2300 is my minimum for the 3 wave. it's the 4th wave that kills ya and it should take a long sideways ride in 500-600 trading range from 2300 (if that is where 3 stops).

i could be wrong though, but it sure looks incredibly bullish to me. buy-the-dips bullish.

wave rust

[MARK HOLSCHER]

DG DG DG YOU ARE TALKING FROM A WEAK POSITION.....
PROOF: YOUR BEHAVIOR!
WITHIN THE BOUNDRIES OF THE MARKET PLACE ARE BUYERS AND SELLERS, EXPECTATIONS OF PRICES GOING UP OR DOWN. THE PLAYERS ARE SAYING THIS OR THAT IS THE MAJORITY. THE MINORITY SPENDS THERE TIME CRITICAL OF OTHERS(BACK BITING)THIS IS A WASTE OF TIME AND UNPROFITABLE FOR THE EDIFICATION OF ONES BROTHER. BUT BY FREEWILL EVERYONE HAS THE SAME RIGHTS TO EXPRESS THEMSELVES.

THE KEY: THE RATIO OF THE TRUTH, TO THE LIE WITHIN ONES OWN MIND IS THE DIFFERENCE BETWEEN PROFIT AND LOSS, THE LOVE OF ONES BROTHER IS PROFITABLE, THE LIE OR THE HATRED OF ONES BROTHER IS UNPROFITABLE OR THE DELUSION. YOU ARE THE CRITICAL MINORITY, OR ONE WHO BELIEVES THE DELUSION IS REAL, OR BELIEVES THE LIE MORE THAN THE TRUTH, OR BELIEVES THE CREATURE MORE THAN THE CREATOR, OR BELIEVES IN THEMSELVES MORE THAN ANYONE ELSE, OR BELIEVES THEMSELVES TO BE THERE OWN GOD. THERE IS A PLACE FOR DG AND HIS FRIENDS IN THE MARKET PLACE.
http://knol.google.com/k/mark-holscher/the-enemy-is-on-the-horizon/13nmdnwfdknux/1#

[DG]

Very mature, DG. Try trading off Roger's calls. Your mood will improve.

Yeah, right.

[da bear]

I have gold's wave 3 as starting from the double bottom in 2000 or 2001 up to the Spring 2008 high near $1,000. Since then we have been in a wave 4 and gold is currently somewhere in wave B or at the beginning of wave C.

This is very similar to the wave 4 correction spoken about by Prechter and Frost in THE ELLIOTT WAVE PRINCIPLE. that wave 4 was an a-b-c correction and ended at $103 gold and led to the final wave 5 rise that took gold to $850.

That was the big wave 4 of the asset markets in the seventies.

In the eighties the major wave 4 of importance was the 1987 crash in stocks. After that low stocks ran up another 13 years into the year 2000 orthodox top.
Gold is at a similar point (within a fairly large correction) but with more upside to go.

da bear

[Account Deleted]

"Wait in the 2nd qtr GDP blows big Friday. Consumer confidence is heading down the toilet."

Posted by: Roger D.

Dont know about the consumer confidence but your confidence in analysing the mkt movements certainly seems headed in that direction.