While I do understand the basic concept around the use of statistics in financial markets, I do not consider myself as a mathematically talented individual. I possibly account for this lack with an ability of striving to be sensitive to my surroundings and forecast accordingly by observing striking patterns. My minimum statistical ability results in dependence on market statistical input from a few other.
A commonly used tool in statistics is Correlation. There are plenty of examples on the internet where you will find market pundits plotting charts to show correlation between markets.
The problem which I see with a pure statistics based approach is well expressed in this phrase
Statistics is like a bikini,it shows you a lot but hides the vital components
In other words the historical back test period, assumptions and more importantly the economic/investing concept that is being promoted can all influence its making.
Don’t get me wrong, correlation can be an indispensable tool in the hands of a broad minded investor/trader.But rather than trying to use it mechanically to develop an investing model/strategy, it is better left to observe periods when markets that normally correlate to an extent- no longer do. These changes in correlation in combination with economic policy could signify turning points to enter a position.
Various current market correlations are in an ever so susceptible economic policy, which could indicate turning points. All said and done, choosing the right instruments to study market correlation is a whole different story!