One of the gem that I got by my recent reading of "The misbehavior of the Market" by Benoit Mandelbrot (mostly known for his fractal and chaos theory), is that looking for price movement pattern in market is at best an illusion. Market has capability of showing you any pattern you seek. That when added with human psychology create the real problem. It will reinforce within your brain of these patterns.
The problem is that there are 1000s of such patterns, and we have chosen one which one we find convenient. There are absolutely no fundamental reason why the pattern shouldn't break.
There is another problem with technical analysis too. This is what I call "signal noise problem", In long term, price follow earnings (provided you haven't paid a very high pe to begin with), the detail of which comes once in 3 month. So this is what I call the signal. For the rest of 89 days, stock price move here and there, without any solid reason behind them. This is what I call noise. One need to remove these noise from the price movement to get a feel of where the stock is going on. To a layman like me, technical analysis seems to be much more concentrated on the the noise, rather than the underlying fundamental signal.
Technical analysis on quarterly earning, sales, ebitda, margin might be a better option (not necessarily the correct option) than doing technical analysis of price movement. But here lies the problem, doing tech analysis on 1% of the data don't seem like a high effort, won't fetch us good money for the effort that we are doing, and won't give us specialization. Besides we can't do sophisticated point-and-figure charting, elliot wave analysis on them.
The only thing that I feel one need to look is the long term stock price movement/earning growth, movement of various ratios (whether stagnant, moving north or south). Besides we need to look for short term and long term cyclical for cyclical stocks and economy. Looking for delivery % has shown to help us little bit, and make logical to me.
The only reason we are safe doing tech analysis at valuepickr as we are doing the same for fundamentally solid stocks.
To me technical analysis which can't predict sudden downturn of Arshiya, sudden upmove of Ajanta is a "no use" for me, as these can make or break me. Fundamentally I safe in both case (won't touch Arshiya for super high DE ratio, and Ajanta, well as we know is a bright student in fundamental school of investing).