If you read the article without preconceptions & bias you will find it has got nothing to do with using MA crossovers as trend indicators; it's only when you
extrapolate an existing average to arrive at a
projection for the future - (with the underlying assumption that every variable involved will remain the same and/or the resultant outcome will be the same despite changes in the variables) - that you are likely to go wrong.
Like if you try to predict a 5th year top for the mkt by projecting the four year average or the 5th year return on your portfolio based on the avg returns of the past 4 yrs you may go wrong ('coz the 5th yr may actually be a down year due to changes in economic fundamentals and/or overvaluation, liquidity crunch etc.).
Take Bank stks for e.g. Just because this year has been good for these stks if you assume the same avg returns for them next year, you'll be doing so at your own peril. Rising int rates may result in a complete change in scenario for these stks. But then again int rates is only one variable (or only one input for a simulation) There are others like credit growth, deposit mobilisation, sectoral cosolidation etc. which need to be considered for a complete, exhaustive simulation of all probabilities (& then off course bet on the best, most likely scenario).
In the end, beware of quacks who assume a disease where there is none (misrepresenting facts, presenting them out of context either deliberately or out of ignorance) and provide blatantly false &/or ineffective cures where a malady does exist. Their sole objective is to draw 'patients' for their quackery and will keep mouthing one theory or the other (actually knowing or caring little about any of them) to lend themselves some credibility. Needless to say, when grilled by people with real medical expertise they will be found to be grievously lacking in the very theories they vociferously propound.
Regards,
Kalyan.
