Catching the Wind

vladv

Member
grood,can you post some screenshots with your trades,as well?Anyway,thanks for sharing the ATR indicator!
 

stovedude

Active member
This strategy seems to be so incredibly accurate in picking exact tops and bottoms, much like ZigZag, only without the lag and repaint. However, it must be done correctly and quickly, as price tends to move a little faster than the manual calculating process. I will try to finish an indicator and EA with the concepts involved. I have found that the available TF's for MT4 are not enough, because price can reverse on other TF's, so I have made some indicators that do not use other TF's. Instead, they can calculate any TF period on the chart. This was quite an improvement. The one indicator I do like the most seems to be more difficult for me to interpret and I may use it later when I figure it out. The other indy is perfect in every aspect except one: it needs a starting reference point for other TF's. For example: all TF's are calculated from the same starting time as the Daily (0:00). However, if we have something like a TF of 23, it doesn't divide into anything very well. I am able to calculate it anyway the same as the rest of the TF's, but the channel lines and calculations can become a little offset compared to a TF of 24. Therefore, one must be careful in analyzing the next TF to make sure it conforms. I will try to explain this a little more in my next post, along with the indicators.
 

stovedude

Active member
Here are 2 indicators I'm working with. The first one is a mod of the original from before to include fibonacci. However, it is TF locked to only the ones available in MT4. The second indicator can transcend that problem and allow any TF. The basics of the strategy is to study and analyze lower TF's (this is the Length setting), starting at any major reversal and working you way to the next reversal. On different chart TF's, this will produce different reversal points of course. I use it on E/U M1 (a noisy choice of course) and it seems to work good. You can start with a Length of 1 and work your way up. What you are looking for is the single dotted line (which is called the Prev Median line if you hover your mouse over it). This will be your most important line, and when it crosses the double dotted line (which is called the Curr Median), it means that TF has changed trend. If the Prev Median is above, the trend is down (because the current faster median has passed below the previous slower median) and vice versa. Now if the trend is up, price will hit the upper channel and, if down, it will hit the lower channel line (by hitting I mean it can come within 1 pip on M1 and still be valid), and I have yet to see it waiver from this rule as long as you pay attention. What I mean is: if price has already hit it when you were not paying attention, you would be in error expecting it to hit again.
But, it isn't just that easy, because the next higher Length can signal a trend reverse somewhere along the way, so you will need to monitor this one as well. Now if it still hasn't changed and still is indicating the opposite trend (if the 2 median values are exactly the same, it is valid for the change also), price will hit that channel anyway, due to one steadfast rule: if price crosses that Length's Prev Median dotted line, it will very possible reverse on that TF channel. And here is another steadfast rule: until that Length changes its trend to match the lower ones, you will not check any more higher Length channels. Just because it did cross another higher Length dotted line, it doesn't mean anything, because it can indeed cross many different medians of all sorts of Lengths in its course. The only one you need to be concerned with is only the highest one that hasn't changed trend yet, as long as price has crossed its previous median line.
This presents a manual problem of course. Usually price moves fast near these reversal points, and the Length can increase dramatically by sometimes 1000's as price reverses back into a previous trend. If you look at E/U, it is working its way up to last year's high, which, on an M1 chart, can be something around 200,000 minutes back of market time, which presents another problem with automation. If we go that far back, we will need to make sure we have that many bars in our chart and enough CPU and memory to manage everything, so a higher TF chart might be needed.
Anyway, I believe this could very well be the answer to the old trend vs. range problem. This strategy treats all movements as trends and catches even the smallest reversals (beware of spread of course).
SL is extremely low (somewhere around 5-10 pips above/below the swing high/low on M1), TP will depend on your chart TF as well, and can just be the final reversal signal (you can safely use Length 5 trend change or even Length 1 for more aggressive, or use your fav CCI or stoch). TS will of course be difficult to determine on a noisy chart. You may find that final trend can continue for hours and days and weeks, so you might want to limit yourself to a certain number of pips, or keep watch on the lower Lengths to exit your trade. You can very easily enter again, starting the process over on a smaller reversal. You can also increase the Length by 15 increments on a M1 chart without too much trouble I believe.
Overall, I think this might be too overwhelming for manual, but it might help for some people to try it on demo and study it for awhile to see if there's anything to improve. Like I said before, the first indy has limited TF abilities, and sometimes price will reverse without hitting the next TF, just because it chose something in between. I like to put several instances of the 2nd indy on a M1 chart with all different colors. That way, I can see how they all interact with each other. It can give you a headache though.
 

Attachments

stovedude

Active member
Just as an update. I have the EA working now. I'm just fine-tuning the logic some more and killing a couple bugs. One thing I noticed over the weekend is that some extremely simple calculations inside the EA were completely wrong, especially near the reversal points (by simple I mean getting the high and low of the current candle). This has me wondering about our lovely brokers and how they manipulate the data. Sometimes I believe there are two different sets of data being sent to our platforms, one we can see and one we can't (at least not through automation). Very strange indeed. However, today it's working fine, so go figure....
 

grood

Member
I heard rumors that price data received by an EA is different than what you see on the chart/price feed. At least they can control the ticks sent to us, like go markets who smothes the price feed (they even admit to doing that).
 

vit007

Member
Just as an update. I have the EA working now. I'm just fine-tuning the logic some more and killing a couple bugs. One thing I noticed over the weekend is that some extremely simple calculations inside the EA were completely wrong, especially near the reversal points (by simple I mean getting the high and low of the current candle). This has me wondering about our lovely brokers and how they manipulate the data. Sometimes I believe there are two different sets of data being sent to our platforms, one we can see and one we can't (at least not through automation). Very strange indeed. However, today it's working fine, so go figure....
How the trading going with you new EA Stovedude???
Vit
 

stovedude

Active member
I have found that it is even more accurate on renko or range bars. I'm currently constructing it for that system. It's very accurate in finding the exact tops and bottoms by calculating dynamic price channels based on past history behavior. In other words, it calculates trend first at a small level, and then increases the period as trend confirms. The turning point is calculated when trend no longer agrees, and the top/bottom is calculated from present and past highs/lows using fibonacci ratios. It's quite complicated to say the least and has taken an extreme amount of work. Renko seems to work better to calm the noise. It still worked with regular charts, but sometimes you have to suffer very small moves where spread is a problem.
 

Joy22

Member
I have found that it is even more accurate on renko or range bars. I'm currently constructing it for that system. It's very accurate in finding the exact tops and bottoms by calculating dynamic price channels based on past history behavior. In other words, it calculates trend first at a small level, and then increases the period as trend confirms. The turning point is calculated when trend no longer agrees, and the top/bottom is calculated from present and past highs/lows using fibonacci ratios. It's quite complicated to say the least and has taken an extreme amount of work. Renko seems to work better to calm the noise. It still worked with regular charts, but sometimes you have to suffer very small moves where spread is a problem.
The best set for lenght MTF Fibo channel is 3,4, 5,I Thing this is an great method.
I am studying very Fibochannel method.
Very thanks joy22
 

Tinojazz

Active member
VIP Member
Ive made some channels using simple MA 200 , i think this can be usefull for you all. You can play with settings how can you find the best S and R levels and channels per every pair and every TF. This settings are for H1 and H4.

MA-fibo_buy = buy area
MA_fibo_sell = sell area

Regards

You must be registered for see images
 

stovedude

Active member
I've made considerable progress with the MTF MA channel indy. I have reduced the process of calculating the highs and lows of the channel to just one simple formula, with only 2 values reversed depending on if you're needing the high or low. It is actually looking much more profitable than I originally expected. Now it's calculating highs and lows of each individual candle, until the period increases. It loses a little sensitivity at higher periods, but not enough to miss the major reversals. I should have the final indicator finished soon that will do all the dirty work. It's basically finding the trends within the ranges, as well as the ranges within the trends (a strange concept I know)...
 
C

cfanfx

pls help

pls help

i need OSMA INDICATOR with red color wen below zero line and blue color wen above zero



stove pls help with this ..


thanks in advance
 
Top