Yes, thanks Fred. I hearby vote this the #1 discussion thread on IU in 2011.
While we're at it, let's not just blame government officials. Let's talk sell-side analysts. I would guess if you are a data cruncher it might be better to take the other side of their recommendations on a trade, especially given that at any time they rate less than 5% of stocks a sell.
In fact, after the whole 2000 bubble implosion/Sarbanes Oxley/Regulation FD nonsense, it appears that things have gotten worse... earnings projections and results have gotten MORE volatile, due to the poor forecasting ability of sell-side analysts and "Chief Strategists."
I wrote about it here, take a look at the charts and the data. Earnings estimates/market reactions have become more extreme peak/trough:
http://www.investoruprising.com/author.asp?section_id=1287&doc_id=232420&image_number=2
After looking at that chart, it should give you doubt when the top sell-side strategists claim that S&P earnings of $100 per share can be maintained throughout 2012.
You are ABSOLUTELY right, @Fred! 56 out of 58 is NOT an aberration! They didn't know that they were dealing with the Data Whisperer when it came to YOU looking at the numbers, that's for sure.
Final score: Fred 1; DOL 0
I say this early and often but it bears repeating: We are SO LUCKY to have access to your insights!
Oh, sorry I misunderstood your point Street Smart.
I certainly understand the difficulties in sampling data, but there is no way to make an error in the same direction 56 out of 58 times if you are honest. An honest data collector would at the very least subtract the average systematic error from the data and present that as his/her estimate. Simply doing that would present a number much closer to the final revised figure.
Oh, no @Fred, I totally got your point and I thought it was really an interesting illustration of the "make news with the numbers now, revise them later" phenomenon.
I was making the additional point that the DOL's numbers were suspect to begin with. When you're dealing with an original number that is 1% of the total and then revise it to .5%, I'm just questioning the statistical accuracy of the original 4,000 number.
In other words, the DOL is saying that not only does it know exactly where the needle in the haystack is; it then knows when it becomes half a needle.
Impressive if true, but I'm dining at @Noreen's salt lick!
I guess I didn't make my point clear Street Smart. The error is nominally very small as you point out and as I wrote in my post, but the announcement emphasized the change between weeks not the total amount. The improvement last week was originally given as 4000 and then revised to half that amount when nobody was looking.
They announced twice the improvement that actually occurred and they have done that in 56 out of the last 58 weekly reports!
To me that is a significant fabrication that is probably designed to mislead.
I found the level of competency among government workers ranged from exceptionally high to abysmally low. The high performing workers often went out of their way to correct errors and omissions in data provided by some of the less committed workers. But when you consider the amount of data being gathered -- and the fact that it is often gathered by one person, working without much oversight -- you can see the potential issues.
If one person, working alone for as low as $12 or $13 an hour -- intentionally or inadvertently -- surveys the wrong household, enters data incorrectly, fails to explain a question correctly or misunderstands the question she is asking -- then the results could be compromised.
Yes, there are certain quality checks in place. But they don't catch every error, or every lie.
Noreen, is that because we can expect a government agency to massage the data? Or because they are not capable of collecting all the right data? Or both?
As a former federal employee, I have no hesitation suggesting the following: Take all government data with a grain of salt.
Love it, @Fred! Plus, what is the sampling error on these numbers? Gotta be more than 2,000, right?
Two thousand as a percentage of either 364,000 or 366,000 is roughly half of 1%, which might as well be ZERO in my book!
Re: Markets and Media--Chicken or Egg?
Street Smart
12/29/2011 5:05:50 PM
EXACTLY, @Scott! Or in this day and age of high-frequency program trading and increased market volatility , maybe the computers just had to sell or buy! Financial journalism really reached its point of total absurdity when it tried to explain the Flash Crash!
The blogs and comments posted on Investor Uprising do not reflect the views of Investor Uprising, PRNewswire, or its sponsors. Investor Uprising, PRNewswire, and its sponsors do not assume responsibility for any comments, claims, or opinions made by authors and bloggers. They are no substitute for your own research and should not be relied upon for trading or any other purpose.
|
 |
Latest Blogs
Telecom-equipment maker Ciena is a stock trader’s dream, as long as the timing is correct.
The FTC is offering a $50,000 cash prize to the person or group that can come up with a solution to those annoying robocalls.
Akamai is in the middle of four significant tech trends.
John Malone of Liberty Media will be taking over Sirius XM satellite radio when the existing CEO Mel Karmazin steps down. What's it mean?
Demand for students of the humanities exists, despite widespread aspersions on the discipline.
IU Education
Resources to help you become a better investor
Investor Uprising on Twitter
25 market-moving companies we're tracking
|
Tweet This