The Global Correlation Model

In one of his alarmist posts Will the Elite Wake Up in Time & Confront The Left Climate Fear Campaigns?, forecaster Martin Armstrong praises his computer as usual for being able to sort out the correlations:

To survive this total complete mess, it will take a computer to monitor everything and sort out the correlations.

In one of his recent reports that carry a very high price tag, he made use of such a correlation as demonstrated here: The Martin Armstrong Gold / Silver Ratio Fallacy which indicates that such correlation may not hold for the future for long, only in hindsight.

And his results show that there is no value in correlating markets in the way that he does it.

And there is no evidence ANYWHERE in his writings that he used the correlation between markets successfully to provide better forecasting.

Still he makes these claims. In Socrates – Artificial Intelligence (Real) he writes amongst other hyperbole:

He also monitors the entire world news feeds and looks for fundamental news events that correlate behind the global trends.

I know with 100% certainty that his programs do NOT use news feeds as input at all. I have analyzed his systems, and according to my analysis, Socrates applies the same single-dimensional logic based exclusively, without exception, on the price history to every market in isolation. For some background, see:

The Reversal System - Engineering Perspective

For those who are interested in correlations between markets, this has been studied for years, see Stock correlation network.

Correlation queries have been available as various online computer services like here at aistockcharts.com and here.

Correlation studies are standard part of broker services today.

Until this day, Martin Armstrong's so called correlation models have been unavailable. At best, single studies have been manual executions not reaching the level of complexity as seen elsewhere.

Socrates still presents a flat earth model where each market is analyzed in isolation, merely based on its own time series, not correlated with anything.

Martin Armstrong planned in 2013 to build the Global Correlation Model. Here in 2014, it becomes obvious that this, while still not available, aims to be no more than expressing technical analysis of markets in terms of various currencies.