Sorry for so many posts but I heard people don't like too many edits. Anyways I think possibly as a trade filter using divergence. I know we are trying to avoid extra indicators but I have seen this posted in pa threads before by good traders. Basically any oscillator will work but you only need one lol. Don't think 5 of them is somehow going to make it better. Divergence is fairly reliable because it is giving you a reading of the flow of the market based on how many bars you ask it to look back. This can be helpful because we are looking for the trend to continue and this could alert us to a possible end to the trend and for us to stay out. Just go back and look at the losing trades and you will see that the previous leg usually has divergence and that the leg before that usually is not in divergence. Tradeprice swears by stochastic divergence and he is a professional forex trader and a level 2 nasdaq trader. I myself prefer rsi(14) or standard macd but I think just about any oscillator will work for this. Also I think most people after a while can spot divergence without an indicator but this makes it a little better on the psychology since it's not subjective. In other words either there is divergence or there isn't.
- #362
- Edited 6:40am Mar 11, 2012 6:29am | Edited 6:40am
- | Commercial User | Joined Feb 2012 | 3,770 Posts
I am not that successful but I am very logical and math orientated as well as a programmer. I hate programming mt4 though because the editor is absolutely terrible and that's being nice. Anyways, I just happened to notice that it seems that most of the winning trades are winning more than just 1:1. I am not trading this as of yet but tomorrow I think I will give it a try. While I am also against a hard set of rules for entering, I think a hard set of rules for exiting may benefit us (or at least myself) because it takes the emotion out of it....
will check it out on my mac, running windows...In the steve primo webinars he talks about 70% rule based allowing 30% for discretion... so if a signal fits all the rules you then have some discretionary rules you use to make the final decision.
I will look at how far the pair has already moved compared to its daily range, the slope of the ma and and really strong s/r levels.
GU daily is a good example now, valid set up but right on that s/r and just breaking back over the ma... I will be more inclined in that case to let it make another move then get in on the first retrace on H1 for instance.
tradewith60
In the steve primo webinars he talks about 70% rule based allowing 30% for discretion... so if a signal fits all the rules you then have some discretionary rules you use to make the final decision.
I will look at how far the pair has already moved compared to its daily range, the slope of the ma and and really strong s/r levels.
GU daily is a good example now, valid set up but right on that s/r and just breaking back over the ma... I will be more inclined in that case to let it...
It takes many trades to determine the profitability of a system. After 100 trades taken 30 times in an unprofitable system of 45% win rate and a 1:1 r/r approximately 4 out of the 30 times you will still be profitable. It seems that the sweet spot is about 250 trades. After 250 trades you have about a 1/30 or 3.3% chance of being profitable. That's valuable information because you might think that after 100 trades you would know but statistically you would still have a 13% chance of being profitable even with a losing system. This would tend to lend credence to those big backtest fans since it would take about 6 months to reach 250 trades at an average of 2 trades a day.
- #364
- Mar 11, 2012 7:04am Mar 11, 2012 7:04am
- | Commercial User | Joined Feb 2012 | 3,770 Posts
I might add that I have learned a lot from this program.
It takes many trades to determine the profitability of a system. After 100 trades taken 30 times in an unprofitable system of 45% win rate and a 1:1 r/r approximately 4 out of the 30 times you will still be profitable. It seems that the sweet spot is about 250 trades. After 250 trades you have about a 1/30 or 3.3% chance of being profitable. That's valuable information because you might think that after 100 trades you would know but statistically you would still have a 13% chance of...
are you saying that you need at least 250 instances to test before you have accurate data?
tradewith60
That went a bit over my head
are you saying that you need at least 250 instances to test before you have accurate data?
It's kind of hard to express but maybe a picture or two will do justice. This is based on a starting balance of 100 and a fictitious losing system that has a r/r of 1:1 and wins only 45% of the time. In the first picture you can see your balance after 100 trades. Each balance is a separate instance of 100 trades. You can see that many times your balance was above 100 even though it's a losing system. Now look at the second picture which is after 250 trades. Each number is your balance after a separate instance of 250 trades, and you can see that almost never does your balance stay above 100. I think 2 times out of 150 (5 tests of 30) you are profitable which is less than 1%.
- #366
- Mar 11, 2012 7:44am Mar 11, 2012 7:44am
- | Commercial User | Joined Feb 2012 | 3,770 Posts
tradewith60
- #369
- Mar 11, 2012 8:15am Mar 11, 2012 8:15am
- | Commercial User | Joined Feb 2012 | 3,770 Posts
tradewith60
- #370
- Mar 11, 2012 8:20am Mar 11, 2012 8:20am
- | Commercial User | Joined Feb 2012 | 3,770 Posts
tradewith60
with regards adding indicators, I personally have never seen any instances where it has improved a system, in my experience all you do is get more winners that were loser before, and more losers that were winners before... just my opinion.
In 3 years the only thing i have seen improve a system is the trader himself...
experienced traders will on the whole have a better feel for the market and make better decisions...
I am in a skype room of 42 traders all trading a similar system, all looking at the same charts, and the most talented/experienced...
well, in fact, i got "42" confirmations after all
keep up posting, pal.
- #379
- Mar 11, 2012 8:15pm Mar 11, 2012 8:15pm
- | Commercial User | Joined Feb 2012 | 3,770 Posts
tradewith60
- #380
- Mar 11, 2012 8:28pm Mar 11, 2012 8:28pm
- | Commercial User | Joined Feb 2012 | 3,770 Posts
tradewith60