Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : 2026 TeoTwawKi ... 2032 Darkest Interregnum -- Ignore unavailable to you. Want to Upgrade?

To: carranza2 who wrote (44381)	12/27/2008 11:07:50 AM
From: maceng2	Respond to of 217546

Jesses work looks good as said. "over modelling" can be a problem and models are not written in stone or anything. Jesse has looked at some factor candidates that should model the pog. There are lots of variables that can also be looked at including currencies like jpy, cad, stock market indexes, you name it. some variables can be dependent on others and even act as proxies. an example would be a persons height is proprtional to a persons weight. Obviously care needs to be taken with such relationships. I used several types of software packages over the years, and have had some success with them in the real world. There are lots of unanswered questions to study.

To: carranza2 who wrote (44381)	12/27/2008 4:11:16 PM
From: GPS Info	8 Recommendations Read Replies (2) \| Respond to of 217546

Jesse's 89% positive correlation looks good to me so I keep an eye on MZM. In this case, the growth of MZM suggests the debasing of currency and the ascent of gold. After 30 years of professional statistical analysis in avionics, RF signal analysis and GPS systems, I would like to say that correlations are meant to be broken. The whole point of statistics is to find consistencies and correlations, but these usually occur under more benign conditions. Good engineers seek to have their models degrade in an orderly fashion, so as not to have missiles rollover toward the ground control bunkers. Earlier I asked TJ when the price of gold might break his model, but he seemed to suggest that he would simply adjust his holdings to somewhere between 0 and 100% depending on the current price. His exact price model is still a mystery to me. The MZM correlation to the POG is as good as any, so I hope that provides some profit for you in the future. However, I would caution anyone against trusting that “a deviation from their model would return to the mean.” LTCM mistakenly thought that they could recover from deviations in their options models, but failed to allow for the price disruptions caused by Asian currency fluctuations and Russian bond failures. The very smart people running LTCM had too much confidence in their correlation analysis; they apparently couldn’t conceive of the idea that their models could breakdown over the long run. The last year has seen a significant disruption to stocks, and any correlation that one might have created has likely moved into to the 3rd or 4th sigma of deviation. Charts for the BMW model for the DOW stocks are here: invest.kleinnet.com One can see that Bank of America has moved into the 5-sigma range, and GM is well into the 6-sigma range. It should be apparent that GM has broken the time correlation developed over the last 40 years. One could not reasonably expect GM to return to its mean value in the next few years. The simple conclusion is that the GM business model has broken. As a disclosure, I own some recently purchased AA and INTC, but I’m prepared to dump them should the world economy slip into a free fall. My simple point is to sell when your models breakdown, and with the currently volatility, that is much more likely than any time in the last decades. We are all quite fallible.