Technical Analysis : sorting out the useful...

Investors and investment advisers tend to have strong opinions about this subject. There are virtual camps of believers in technical analysis and fundamental analysis and they have infinite arguments against the other side. My viewpoint is that there are useful pieces in both and neither should be written off entirely. I will ponder over the TA side in this post.

Technical analysis was my starting point in finance, when I was searching for a parallel computing problem. Without any background in economics and finance, I accepted things at the face value. However things started to change, when we started to observe the recommendations generated by these techniques. Over a period of time, I learned to differentiate between the useful and not so useful stuff.

Certain pieces / techniques lack any convincing logic. Many arguments have been given (by scholars) subsequently, but they still look unconvincing to me. I would like to enumerate a few...

• Elliott Waves and Fibonacci Relationships.
  
• Trend lines.
• Support and resistance.
  

I hope the life was so simple. It also contradicts commonsense. Any easily recognizable pattern will be exploited by the masses, immediately resulting into trend reversal.

However there are few which do give meaningful information, even though none of them will be useful in a market crash situation (the unexpected event):

• Bollinger Bands is a reasonably good indicator to visualize volatility.
  
• Upside / Downside Ratios and many similar indicators do indicate the market sentiment, fairly well.
  
• Stochastic Oscillator and Willams%R seem to work fairly well when the market makes a sideways movement and specially if it has periodic characteristics.
  

However, if you are planning to use any of these, my recommendation is to test your hypothesis over a reasonably large historical data which includes major rises and market crash situations...