What technical analysis cannot do

Every method in this chapter is computed from price and volume. That means none of them knows anything the price series did not already contain.

• No indicator predicts news. A policy change or an earnings surprise is not in the chart until after it happens.
• Patterns have failure rates, and the failure rate matters more than the success rate, because failure is what you have to survive.
• Thin markets break the assumptions. Indicators built for continuously traded instruments behave badly on stocks that sit at circuit limits and trade in gaps.
• Any method tuned enough to fit the past perfectly predicts nothing.

Chapter 8 is about the part that survives regardless of which method you use: not losing so much on any one idea that you cannot continue.