beathespread

Here the auther makes a very clear observation about the fact that on different time frames the market states are different. And this should be taken into account.

The Anchor: The bigger frame is the anchor: This is the market state of the bigger market state

Trigger: The actual market state is called the trigger market state: 

This shot is from the following blog. And those are the comments of the author. What is really important is to see that there are combinations and some of the combinations make the trading easier and some of the combinations make it very, very hard.   

I think those concepts are really useful. This is because the model of the classical technical analysis presume that on the bigger time frame we have a Trend and on the lower time frame we have some kind of oscillatory activity forming price patterns. And we time the entry of the lower time frime in the direction of the bigger.  However this looks to me an oversimplification. In fact both  the classical technical analysis from the Dow Theory and the Elliot wave theory both are trying to put the market in narrow "frames". However for currencies this approach of the market looks like a better model. That is because one currency cannot go forever in one direction against the other. Or on the other hand it could be funny to see an Elliott wave picture of the Euro/Usd from 2001 - 2011. It will be full of XYW waves for sure. This model is trying to model the reality and not to reduce the reality into a narrow frame. Of course there are moments that both the Dow model and the Elliott wave model do fit very well. But both models cannot model properly the currency market.

Well this is a strong personnal and subjective opinion, consider it like that. However as I take the Dow model and the elliott wave model seriously for takinf into account samples for training the trading systems it is a big issue for me. A trading model cannot perform forever, it should be constantly adapted or autoadapted to the current market conditions. That is way the ability to model the market and set the beginning of the sampling period for training/testing is important, and last but not least to know when to stop the model when the market state clearly goes beyond its design parameters.

Fore example for the Anchor:

There we can have two models: one directionnal model and another model that could try to model the best entries.

Forexample for the Trigger: 

There also we can have two models: one model for example would model the market direction and another the best moments for the entries.

The approach of two levels Anchor and Trigger is interesting, that is because beyond the Anchor the risk level is so big that we could not consider it. I just liked it at lot. Thanks Momo.

От 04 октомври 2011