Friday 30 September 2011

Guppy Multi-Moving Average (GMMA)

Found this interesting indicator today. Not too bad. Has the potential to make a great deal of profit, since it gets a person to hold longer. The idea is based on a swack load of EMA's: Fast EMA's (3,5,8,10,12 and 15) and Slow EMA's (30,35,40,45,50 and 60). Now when the Slow EMA's 'compress' (tangle together) there is no trend; when the Slow EMA's 'expand' the trend is back. When the Fast EMA's 'compress' the market is going sideways; when the Fast EMA's expand the trend is starting again.
Now, entry points occur when the Fast EMA's pass over the Slow EMA's, and both begin to 'expand' away from each other, as well as themselves (see example graph below).
Exiting comes when the Fast EMA's 'compress'.
And that's it!....

....But there is some drawbacks with this indicator:

  • The typical timeframe for this indicator is the daily charts, and from there the entry points are few in number (even less entry points if you wait for all the EMA's to cross the 200 Moving Average; described in video below).
  • When there is a possible entry point I found that the price will start to go in the other direction, and the trend will die.
  • The exit points are not great since it take too long for it to realize that the trend is over and should be going in the other direction (I tried a parabolic SAR to find my exit points: wasn't too bad, but it wasn't the adrenalin inducing excitement I was hoping for). 

Check out this 3 part video on YouTube that explains the indicator a lot better than I can:
http://www.youtube.com/watch?v=mfiYfWORHGI

All-in-all, this indicator could be very useful for a long term trade, but on an intray day basis I don't think it is very useful.
There is another indicator out there that is similar to the GMMA, and that the Rainbow MMA. Again, it has the same problems as the GMMA.
I wonder what this indicator would look and act like if a Hull Moving Average, or a JMA, or a Volume-Weighted Moving Average was used instead...?
Food for thought.


7 to 8am (Pacific) Reversal

Saw this pattern yesterday after my last post:
For Light, Sweet Crude Oil Futures at 7 to 8am Pacific Time, or EST 10 to 11am, there is a reversal. I went and crunched some numbers with excel and found that there is a 87% chance of getting a reversal during that time period (this would be within 4 months worth of data; might have to increase that look-back period to get more accurate results). So, in a 1 hour timeframe I would observe if the 7am bar is up or down at the close, and bet against that 7am price direction. Again, I would use a very small take-profit limit (i.e. 5 points + spread).
I also thought that this could be just a coincidence. So, I looked at the 5-6, 6-7, and 8-9 reversal possibility (4 month period). I got a 86% chance for 5-6, 86% chance for 6-7, and 75% chance for 8-9. It would seem I picked the most likely reversal period, but it is not far off from the rest. I'm curious if there is some kind of way to take advantage of the high reversal rate of all three hourly price actions...?
I looked up this phenomena, and found that there is a general rule saying that 10:30am (EST) is a likely time for a reversal because it is the likely time for the `big money` come in.

Here's a daily chart showing all the hourly reversals that did not work (Purple is 6-7, Orange is 7-8, Red is 5-6):

 

Thursday 29 September 2011

1-2-3 Detrend

I've been fooling around with a couple indicators, and one that stands out is the Detrending Price Oscillator (DPO). The default setting on this indicator work well in the long term aspect, but it doesn't catch them a lot of cycles. So, I changed the settings to its lowest possible period number (period = 1), and now it shows every single price move and down... naturally, it is wrong a lot of the time. !HOWEVER! if this fast DPO drops below/above zero and remains there for three price moves (three bars), then we have a trend, or at least the tail end of it. My exit is just a small take-profit limit (5 points + spread), and as always the stop-loss is set just below/above the low/high of the previous bar.

One of the best parts about this is that it works even when volatility is low (for light sweet crude this would be when the US market is not open). 

I'm using this at a 30 minute timeframe on Light, Sweet Crude Oil Futures. Here's an Example:

 

Monday 12 September 2011

....hmmm, its been a while

Looks like I haven't been publishing a lot of post lately, but that's only because I've been trying to figure out a better system the last 2 months. Sadly I'm still working on it, but I believe I'm getting closer to my holy grail here.
I've found a couple indicators that SEEM to be doing the trick, but I've only looked back to the past to determine the profitability of these indicators. I've been fooled before with indicators that re-paint themselves (those are very disappointing moments), so I can't say for sure that what I got is going to work the way I want it to.
I should be doing this in a demo account, but I tend to put little pressure on myself on finding the best entry point on demo accounts, and instead just jump in all willy-nilly. So I'm going to use a real account on very small volumes. I've done a few so far, but I'm still not completely satisfied (I go on daily charts). I have a strange suspicion that one of my main indicators changes its mind to look better. The indicator is an Adaptive Kaufman Moving Average (KAMA) set at a very fast setting:  fast = 1, slow = 11. I chose this setting because of another indicator that I use that has the same settings, and gives very reliable entry points. This other indicator is called EMA Crossover (this indicator, however, will change it's mind and remove its alert if the price moves in the opposite direction of the previous close; now use it only for the alert sound it gives off).
I also have another KAMA with standard settings: fast = 8, slow = 21. This give me an overall idea of market direction, as well as providing a couple good entry points.

I have a score of indicators on both of my computers right now. Each of which have a very responsive display (zero lag on a couple of them), and seem to provide very good entry points.But what seems perfect often ends up being less so, or not at all.
Dam delusions of grandeur!