Unfounded Claims -- Amid Interesting Results

From the current coherence project, Invincible America, PR machine:

"Since the Invincible America Assembly began 94 days ago, the Dow has climbed 1100 points to set an all-time high at 12,116, energy prices have plummeted 28%, and inflation fears have dissipated. At the same time, the White House has abandoned its combative, militaristic approach and has adopted a more effective, diplomatic, consensus-based solution to the conflicts in Lebanon, Iran, and North Korea.

WHY THE WALL STREET BOOM SINCE JULY 23?
New Research Documents Positive Market Impact Generated by 1200 Advanced Meditation Experts* Graphs Highlight Dramatic Rise in the Dow, S&P 500, and the Nasdaq Since the Start of the Invincible America Assembly on July 23, 2006

Wall Street started its record-setting bull drive on July 23—the first day of the publlcly announced Invincible America Assembly at Maharishi University of Management in Fairfield, Iowa—the largest scientific demnstration project ever to monitor the positive effects created by 1200 advanced Transcendental Meditation experts on economic, social, and even climatic trends in the nation.


*The probability of observing such a large increase in the rate of growth purely by chance is less than 3 in 10,000 for the S&P 500, 3 in 10,000 for the Nasdaq, and 7 in 10,000 for the Dow.
back to top




Two points:

1) the claims about S&P500 "advanced econometric model results" etc. simply don't meet the most basic of eyeball tests -- which is an essential part of all econometric analysis.

A researcher needs to carefully look at the data -- and all the supporting slices and dices of it that econometric and statistical software provides. Since Oct 2002 when SP500 hit its low after
internet bubble slide, the that been 2-4 upward reversals of the approximate magnitude of the current one. (2 perhaps larger, 2 a bit smaller). The claims of the current reversal being a very rare event -- something like 7/10,000. Thats not at all credible. Simply look at the data -- it happens about once a year.

Current reversal within green bars, other prior reversals within other colored bars.

And note, the current price level is simply a bit above the regression-based trend line. Util it breaks the upper bound (one standard deviation above trend), the current "fluctuation" is within normal bounds and is not a rare event. Its an intersting event and result, worth monitoring. However it is not the phenomenon that MUM researchers claim. Their biases are showing -- and its not dignifed nor attractive.

2) Tweaking sophisticated models allows one to produce analysis results that allow highly qualified (aka limited, "spun" ) statements, with wiggle room, that imply something (the point a researcher wants to make) but do not, in reality demonstrate such. The above four similar reversals in four years is a good example. One can set up and specify the analysis to show the current event to be unlike any other -- but on parameters that are essentially unimportant or meaningless to the point being made.

As a former analysis and researcher in a large corporation using sophisticated statistical, econometric and simulation software, with such analysis having a lot depending on it (being used by top management, and as part of supporting the corporate case in regulatory proceedings and customer proposals), I know from experience that when one has 100 variations or results to provide, it is simply part of (any and all) corporate cultures to provide the best "defensible"
results. One one simply does not provide all results and drone on and on about "on the other hand" -- (which would give a clearer view of the true picture and associated uncertainties and limits of the analysis.) (Partly because the further complex analysis goes up the food chain, the more "cartoon" like it must become.)

My sense from my TMO experience and observations is that the (often quite subtle, and often self-based) pressure to provide such "best defensible" results is even stronger. The analyst behind this "research" (KC) I am sure is getting lots of pats on the head, a few minutes of glory, for his results. And would be looked at with "that look" and displeasure if he produced a comprehensive and unbiased set of conclusions. Or simply looked at the data and said "these rare events findings are clearly anomolies and due to unjustified model tweaks, because 'look' -- anyone eyeballing the raw data can see it does not support these model results."

On the other hand, the IA course does coincide with the exact low point of a major reversal -- something that only occurs about once a year. That is an "interesting", "hmmmm, chin scratching result".

And if the SP500 goes up another 10-15% in the next few months, it will be a "rare event" -- and something to crow about. Or, if SP500 goes into a long sustained 2-5 year aggresive rally -- more likely the results of ME -- IMO -- since reasonable theroy would indicate some lag period between intervention and full effect.