Thursday, August 14, 2014

TA lesson: it's all in how you look at it

Here's the $VIX. Check out the Bollingers:

Not bad. According to the Bolls, we're at -0.8 sigma, which is not much to get worked up about.

But check this out:

Ermagerd! With a Bollinger period of 10, all of a sudden $VIX is rather oversold, at -1.7 sigma.

The point of this is that all these period-based indicators will give you radically different info depending on what period you're using.*

That's why I like to vary my EMAs to properly fit the price movements, instead of just blindly taking whatever stupid number people always use.

If this was a slow-moving dataset, a longer Bollinger period is totally okay to use. But when it starts moving quickly, you should really change your indicator's period to reflect the speed of the moves. Otherwise you're getting info that isn't properly reflective of the data.

* - The only reason to care about EMAs and Bollingers is that they are a mathematical approximation of people's present attitude versus their emotional memory of previous price performance. When a stock pops above its EMA, that's significant because the viewers are seeing a price that has exceeded their expectations based on its prior trend; when a stock hits the Bollinger rails, that's significant because people can instinctively feel the sigma of the overbought/oversold condition and might be encouraged to take the contrary position.

No comments:

Post a Comment