Ok so I’ve completed the oscillator that I was aiming for, but unfortunately I didn’t quite get what I wanted. It is working as designed but it isn’t as useful as I thought it would be. However, it has definitely given me much food for thought and I have come up with some useful ideas from it.
Firstly, introducing the TrendZ oscillator. It effectively tries to capture the trend angle at each period using the Mid points calculated on the TrendF indicator. It supports up to 5 simultaneous configurations of TrendF, useful for comparing different periods.
Below is a long downward trend in AUDJPY.m15 in July 2013. TrendZ is running with two different periods, 40 and 120. I chose this trend because the trend itself changes into different types of trends as it continues, and this is what I am hoping to try to capture in the TrendZ indicator.
The rectangle zones on the price indicator the different sections of the trend where the trend characteristics are different, or so I believe.
From TrendZ I can see three trend characteristics
The highlighted areas where the green curve is pretty flat seems to be an indication of that the price was not fluctuating much during these periods.
The actual value of the oscillator represents the gradient of the trend, which can be said to be the strength of the trend. It is being captured in #pips per bar – pretty neat I think as we can even use this to try to predict profit limits.
The time since the line crossed zero indicates how long the trend has been running.
So from this TrendZ lesson, I have actually determined that there are three characteristics of a trend… and they are:
Trend Volatility: How smooth is the line representing the gradient. If the gradient is always the same value, then the price is not moving much.
Trend Strength: What is the actual value of the gradient. Is it half a pip per 15 minutes or a 2pips per 15 minutes. A value close to zero means a sideways/consolidation market.
Trend Maturity: How long has the trend been in play?
I think these three characteristics are key to an all-round successful trending strategy. If we can successfully identify these then we can dynamically change our strategy accordingly to customize each trend.
So, before I dived in on trying to detect each category (Does anyone know how to measure the ‘smoothness’ of a line – if so please let me know!)… for comparison’s sake I dropped a couple of MACDs on the chart too. The top one m15 and the bottom one H1.
From looking at the MACDs, it should be immediately obvious why I am not happy with the oscillator despite it allowing me to make these great discoveries? Essentially, having a couple of different ranged MACDs (or some other indicators I tested) can give us all the same information on each trend just as my TrendZ oscillator, maybe even better.
We can gauge the Trend Volatility by seeing how many times the lines cross, or how flat the line is (again how do you calculate flatness anyway?)
We can gauge Trend Strength by seeing the actual MACD values (though it might not be as useful as pips per bar).
Trend Maturity is captured in the time since the line crossed zero.
Well, either way that is good I guess. We still have made some progress… so next, how do we plug all this into a trending strategy.