Let's build together a system which makes 1000%/month | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Bollinger Band: let's build a strategy together 35 replies

A Different Kind Of Challenge - From 1000 to >1000 0 replies

Let's build a trading system (EA) 8 replies

BELLA SYSTEM. Let's try it, let's improve it! 118 replies

Attachments: Let's build together a system which makes 1000%/month
Exit Attachments

Let's build together a system which makes 1000%/month

  • #61
  • Aug 18, 2014 5:26am Aug 18, 2014 5:26am
Ok guys anymore ideas about a methodology which can measure accurately the trends, and it is reliable in the long term picture.

It can be an indicator, a fundamental strategy, or whatever.Throw me all your ideas and we see then which one is the best.
"There's a sucker born every minute" - P.T. Barnum
  • #62
  • Edited Aug 19, 2014 5:15am Aug 18, 2014 3:07pm | Edited Aug 19, 2014 5:15am
Quoting Proximus
Ok guys anymore ideas about a methodology which can measure accurately the trends, and it is reliable in the long term picture. It can be an indicator, a fundamental strategy, or whatever.Throw me all your ideas and we see then which one is the best.

Ok just an idea for a moving average combo to identify trends:

Attachment

Purple = 100 SMA; Blue = 100 EMA; Black = 21 EMA; Red = 13 EMA

If Blue is over Purple, long term trend is up. If Purple is over Blue, long term trend is down.
If Red is over Black, short term trend is up. If Black is over Red, short term trend is down.

Here is a template if you want to have a look and play around with it:

Attached File(s)
File Type: tpl MA_template.tpl   1 KB | 441 downloads
  • #63
  • Aug 18, 2014 7:55pm Aug 18, 2014 7:55pm
Quoting Ethlao
{quote} Ok just an idea for a moving average combo to identify trends: {image} Purple = 100 SMA; Blue = 100 EMA; Black = 21 EMA; Red = 13 EMA If Blue is over Purple, long term trend is up. If Purple is over Blue, long term trend is down. If Red is over Black, short term trend is up. If Black is over Red, short term trend is down.
Well i thought more about the cause of the trend.Some method which can predict the trend in advance, or atleast detect it in its early stage, to make as much profit as possible.

The problem with the MA method is that, if the MA is looking at a too big period, while the trend is shorter than the MA period then the MA will be very lagging and show it too late.

So if we would use the MA method, we either need to look for longer term trends with a long period MA or always optimize it to our needs.

For example i have been experimenting with this gaussian filtered rainbow MA and it showed the trends pretty well, however it had its drawbacks too

Attached Image (click to enlarge)
Click to Enlarge Name: AUDUSDM30.png Size: 95 KB


Here is the indicator:
Attached File(s)
File Type: mq4 -wmmm-GaussianRainbow.mq4   8 KB | 566 downloads
"There's a sucker born every minute" - P.T. Barnum
  • #64
  • Aug 18, 2014 8:39pm Aug 18, 2014 8:39pm
Quoting Proximus
{quote} For example i have been experimenting with this gaussian filtered rainbow MA and it showed the trends pretty well, however it had its drawbacks too {image} Here is the indicator: {file}
There is no indicator, whether to show trend, momentum, volatility, cycles, S/R, etc, that does not have it's drawback! Point is to try and find the sweetspots, compensate for the drawbacks with one or more simple filters, find the sweet spots,....and ultimately, make some money.
  • #65
  • Aug 19, 2014 12:21am Aug 19, 2014 12:21am
True, well guys download that indicator and experiement with it, or show me some alternatives to detect the trend.I`m all ears.
"There's a sucker born every minute" - P.T. Barnum
  • #66
  • Aug 19, 2014 1:18am Aug 19, 2014 1:18am
Quoting Proximus
True, well guys download that indicator and experiement with it, or show me some alternatives to detect the trend.I`m all ears.
Proxy you can code, right?

Code an EA with any trigger you want that trades with the monthly trend only.

The monthly trend is simply defined as if iclose()[0] > iopen()[0] then double trend = 1 || iclose()[0] < iopen() double trend = -1;

Check out your equity curve, you might like what you see.
If you are good at something, never do it for free--Joker
  • #67
  • Aug 19, 2014 4:58am Aug 19, 2014 4:58am
Quoting Proximus
Well i thought more about the cause of the trend.Some method which can predict the trend in advance, or atleast detect it in its early stage, to make as much profit as possible.
You say want to use fundamental data (=the cause of the trend) to predict the trend. The problem I see is that when a news is released its effects are usually already in the chart. The big boys pay people to analyse/survey/investigate. They know the figures before and perhaps more precisely than the official sources themselves. When you get the data in hands the move has already started.

Per definition a trend is a continuation of a move. Nothing can continue if it hasn't already begun. What does "the beginning of the trend" can now be? Anyway. QE lasted several months. World wide crisis or high interest rates last for years this lets you all the time you need to start participating. If you can estimate the trend with a high enough acuracy and low enough latency you could just follow. But...

Problem: G/U has been up on the monthly chart for a year. In 12 months the price moved approximately 2100 pips upward. This is a quite strong trend yet the retracements inside are 350 pips. The amplitude of those oscillations is 16% of the size of the whole move. The up months are barely bigger than that and only 7 out of the 12 are up. In 2005-2006 there was a trend of similar slope. The amplitude was ~500 pips and 1000 pip retracements occurred... The width of the trend channel is as important as the direction. You are certainly not willing to hold losers 300 pips in the red nor you want to take hundreds of paper cuts just to be in the "right" direction.

Quoting jmn5611
The monthly trend is simply defined as if iclose()[0] > iopen()[0] then double trend = 1 || iclose()[0] < iopen() double trend = -1;
For the few I know of MQL this is simply the color of the last monthly candle. It classifies the trend either UP or DOWN. 40% of the time the trend is flat even on the monthly chart. So this is at best correct 60% of the time. A closer look at this formula reveals that it takes an artificial reference point: the 1st of the month. It could be smoothed by using a sliding window over the last month and this will just be the 25 period momentum of the daily chart which is no more than the slope of some moving average.

Quote
Code an EA with any trigger you want that trades with the monthly trend only.
I did this

Quote
Check out your equity curve, you might like what you see.
I did not!
No greed. No fear. Just maths.
  • #68
  • Aug 19, 2014 6:07am Aug 19, 2014 6:07am
Quoting jmn5611
{quote} Proxy you can code, right? Code an EA with any trigger you want that trades with the monthly trend only. The monthly trend is simply defined as if iclose()[0] > iopen()[0] then double trend = 1 || iclose()[0] < iopen() double trend = -1; Check out your equity curve, you might like what you see.
In the past 12 months i tested all sorts of automatizable methods and concepts, which included this aswell, with no succes.I wasted enough time of my life testing these to figure out that they cannot possibly work.

I dont say that this cannot work, but using only this as a guide will only give random returns and hardly any positive expectancy over the long run.Maybe with more trending pairs like GBP/NZD it could theoretically work, however i want a universal method.Because if it works on 1 pair but no the other then the methodology and its reliability can be questioned.

Quoting PipMeUp
{quote} You say want to use fundamental data (=the cause of the trend) to predict the trend. The problem I see is that when a news is released its effects are usually already in the chart. The big boys pay people to analyse/survey/investigate. They know the figures before and perhaps more precisely than the official sources themselves. When you get the data in hands the move has already started.
Sorry PipMeUp but thats just nonsense, first of all fundamental analysis is not just jumping to trade the news spikes,it is much broader, it is the whole economic analysis of the 2 countries of a pair.
Secondly the "priced in" concept is flawed because it implies an efficient market, which the market is clearly not,because the trend itself makes it not efficient.But thats another topic.
What i`m saying is that how i see news events is that, first the speculators come,which cause the spikes, but the later longer or mid term traders/investors might put their money on it.So the fact that a GDP report has a 300 pip spike upwards, that doesnt mean that the move wont continue,i believe a GDP report is pretty much relevant until the next one comes out.

Quote
Per definition a trend is a continuation of a move. Nothing can continue if it hasn't already begun. What does "the beginning of the trend" can now be? Anyway. QE lasted several months. World wide crisis or high interest rates last for years this lets you all the time you need to start participating. If you can estimate the trend with a high enough acuracy and low enough latency you could just follow. But...
Well it has to start somewhere isnt it?Something must trigger it, usually its a big order, depending on size which lasts long time.It may start with 1 trader placing a few hundred million in the market, and later the others join him pushing the price up, while the retracements are partial trakeprofits.
But we have to ask ourselves why did that trader put that order on in the first place? It must have a reason, and its usually a fundamental event or a geo-political news.
The Ukraine put a little downward pressure on the EUR/JPY for example on the Weekly perspective


Quote
Problem: G/U has been up on the monthly chart for a year. In 12 months the price moved approximately 2100 pips upward. This is a quite strong trend yet the retracements inside are 350 pips. The amplitude of those oscillations is 16% of the size of the whole move. The up months are barely bigger than that and only 7 out of the 12 are up. In 2005-2006 there was a trend of similar slope. The amplitude was ~500 pips and 1000 pip retracements occurred... The width of the trend channel is as important as the direction. You are certainly not willing to...

The GBP/USD is too volatile, either we look for smaller trends or we just take another pair.But we can work around this, by no trading when there is a retracement.We wont risk trading against the trend, but we could put a limit order when there is a retracement.

Theoretically we could fade the moves aswell, because a stronger move in the opposite direction hits the TP faster and our market exposure is smaller, while if we faded at the wrong time the trend builds up slower than the "crash" crashes, so we could have plenty of time to exit our position at a smaller loss, however if we were to fade the moves then we could also get whipsawed or spoofed by some long wicks.

So a trend trading needs big stoploss to survive retracements, but also needs frequent trail stop to the previous low/high.

Quote
{quote} For the few I know of MQL this is simply the color of the last monthly candle. It classifies the trend either UP or DOWN. 40% of the time the trend is flat even on the monthly chart. So this is at best correct 60% of the time. A closer look at this formula reveals that it takes an artificial reference point: the 1st of the month. It could be smoothed by using a sliding window over the last month and this will just be the 25 period momentum of the daily chart which is no more than the slope of some moving average. {quote} I did this {quote}...

I agree completely.
"There's a sucker born every minute" - P.T. Barnum
  • #69
  • Edited 8:49am Aug 19, 2014 6:18am | Edited 8:49am
For the trend direction you can try using ... Cycle Identifier on higher tf or on the one we are trading... depends on the way we will trade? Or whatever its' name is. For strength or entry points you can try using Raptor Explosion (not a very known indicator i think).

Support/Resistance levels or Pivots for targets or confirmation for an entry along with Raptor. There will be a lot of tweeking done with all indicators included in the strategy so just putting it together won't be it - that's half of the work we have to do :-)

That is of course if we use any kind of indicators, there are traders that don't.

Tell me what you think and have a nice one,

Emil
  • #70
  • Aug 19, 2014 8:52am Aug 19, 2014 8:52am
Quoting Proximus
The GBP/USD is too volatile
You wanted to trade RUB for its crazy super high volatility (sic!) and now G/U is too volatile...
No greed. No fear. Just maths.
  • #71
  • Aug 19, 2014 2:15pm Aug 19, 2014 2:15pm
Quoting PipMeUp
{quote} You wanted to trade RUB for its crazy super high volatility (sic!) and now G/U is too volatile...
Both are.The RUB has big daily and weekly candles, but its direction is more straight than not.The GBP/USD changes direction more frequently, so if we compare both currencies on a volatility basis, they are almost on the same level.

Usually but not always exotic currencies have a higher consecutive "candlestick color" count, while high liquidity pairs like EUR/USD have a smaller consecutive color count.

For EUR/USD 5 consecutive same color daily bars are rare, while on EUR/RUB 10 is the average.

But if we choose trend following then volatile currencies are really not an option because if the intra-trade DD we have to endure on retracements, while the trend movement wont be much more than its volatility.

We could measure each trend by a simple formula, the (avg trend movement) / (avg candlestick size in that trend), and see which pairs have the highest ratio, those will have the less risk.
"There's a sucker born every minute" - P.T. Barnum
  • #72
  • Aug 19, 2014 2:23pm Aug 19, 2014 2:23pm
If a pair moves in straight line it is not volatile. The volatility is the variance around the straight line.
No greed. No fear. Just maths.
  • #73
  • Aug 19, 2014 2:30pm Aug 19, 2014 2:30pm
Quoting PipMeUp
If a pair moves in straight line it is not volatile. The volatility is the variance around the straight line.
LoL but candlesticks carry also a volatility.The market is fractal, it is a straight line wrapped up in a zig-zag form, the trend is only the direction of the CLOSE price,
but the HIGH and LOW prices we also have to endure in a trade :d

Attached Image (click to enlarge)
Click to Enlarge Name: EURRUBDaily.png Size: 24 KB


Attached Image (click to enlarge)
Click to Enlarge Name: EURRUBDaily2.png Size: 31 KB
"There's a sucker born every minute" - P.T. Barnum
  • #74
  • Aug 19, 2014 5:11pm Aug 19, 2014 5:11pm
Quoting Proximus
{quote} Well i thought more about the cause of the trend.Some method which can predict the trend in advance, or atleast detect it in its early stage, to make as much profit as possible. The problem with the MA method is that, if the MA is looking at a too big period, while the trend is shorter than the MA period then the MA will be very lagging and show it too late. So if we would use the MA method, we either need to look for longer term trends with a long period MA or always optimize it to our needs. For example i have been experimenting with this...
You are right. Moving averages are lagging to one degree or another. If we want, we can adjust the settings on our MA configuration of course (e.g. using a 50EMA/50SMA cross to determine long term trend and 13EMA/8EMA cross for shorter term trend, rather than the more lagging settings I gave previously). This of course would be more responsive to changes in direction but to pay for that we have to accept more 'false' trend changes being generated.

Perhaps one good way to anticipate the end of a trend is when a Major Trendline Break occurs. There are many ways to draw trendlines and different trendlines to draw. To qualify as a major trendline I suggest drawing it from the high/low point of the recent trend to the last swing high/swing low, or between two or more recent swing high/swing lows. There is an element of subjectivity to this I admit.

Attached Image (click to enlarge)
Click to Enlarge Name: gbpusd-h1-pepperstone-financial-pty.png Size: 58 KB


It's not always a signal that a Major Trend Reversal is about to occur (this only happens about 40% of the time, usually after a retest of the prior trend's high/low). More often it will lead to ranging that can then break out in either direction. It's a good idea to pay attention to them though because reversal setups are one of the highest RR types of trade there is.
  • #75
  • Edited 7:04pm Aug 19, 2014 6:46pm | Edited 7:04pm
Quoting Proximus
{quote} So if we would use the MA method, we either need to look for longer term trends with a long period MA or always optimize it to our needs. For example i have been experimenting with this gaussian filtered rainbow MA and it showed the trends pretty well, however it had its drawbacks too {image} Here is the indicator: {file}
The gaussian filtered rainbow MA looks great, thank you!

If, like you say, we could use an indicator like this to accurately identify longer term trends (perhaps with a valid fundamental reason behind them), it should be possible to then trade in the direction of that trend intraday. The dream scenario is when you can enter with a smallish stoploss (and therefore larger position) at the beginning of a long term trend and then keep adding to your position on the way up, hopefully exiting somewhere near the top.

If we were to adopt such a strategy I think we would need to concentrate on as wide a basket of currency pairs as possible because there might only be a few out of say 15 or 20 that are in a strong multi-week or multi-month trend at any one time. For this we would need to look carefully at the choice of broker as some offer better swap rates than others.

Although I usually use it for entries, I've experimented with adding a moving average crossover alert to the chart along with your gaussian rainbow, as an added trend filter. My chart is for H4 GBPUSD and I chose settings for the alert that will print a green arrow whenever the 13EMA crosses the 21 EMA to the upside (bullish) and a red arrow when it crosses it to the downside (bearish) but you can put in any settings you like. I hope it's okay I share the indi

Attached Image (click to enlarge)
Click to Enlarge Name: gbpusd-h4-pepperstone-financial-pty.png Size: 59 KB


Here's how to change the settings:

Attached Image (click to enlarge)
Click to Enlarge Name: movingaveragecrossoveralertsettings.png Size: 346 KB
Attached File(s)
File Type: mq4 MA Crossover Alert.mq4   8 KB | 378 downloads
  • #76
  • Aug 20, 2014 1:48am Aug 20, 2014 1:48am
I assume that you do all realise that any system which could make 1000% per month will fail because it is taking on too much risk...
  • #77
  • Aug 20, 2014 4:08am Aug 20, 2014 4:08am
Quoting Ethlao
You are right. Moving averages are lagging to one degree or another. If we want, we can adjust the settings on our MA configuration of course (e.g. using a 50EMA/50SMA cross to determine long term trend and 13EMA/8EMA cross for shorter term trend, rather than the more lagging settings I gave previously). This of course would be more responsive to changes in direction but to pay for that we have to accept more 'false' trend changes being generated.
You can also let the market decide the period of your MA by adapting it to the duration of the recent swings. Below is E/U daily with an EMA which period changes based on the duration of the previous 4 swings.
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSD-D1-swing-modulated-ema.png Size: 33 KB
No greed. No fear. Just maths.
  • #78
  • Aug 20, 2014 6:03am Aug 20, 2014 6:03am
Quoting PipMeUp
{quote} You can also let the market decide the period of your MA by adapting it to the duration of the recent swings. Below is E/U daily with an EMA which period changes based on the duration of the previous 4 swings. {image}
Hey buddy, love the logic behind that ema. I'm getting tuned in to letting the market decide most aspects now/... mind sharing?
  • #79
  • Aug 20, 2014 6:47am Aug 20, 2014 6:47am
Quoting Ethlao
{quote} You are right. Moving averages are lagging to one degree or another. If we want, we can adjust the settings on our MA configuration of course (e.g. using a 50EMA/50SMA cross to determine long term trend and 13EMA/8EMA cross for shorter term trend, rather than the more lagging settings I gave previously). This of course would be more responsive to changes in direction but to pay for that we have to accept more 'false' trend changes being generated. Perhaps one good way to anticipate the end of a trend is when a Major Trendline Break occurs....
Not good, by the time you draw the trendline the trend is almost over.

And the break of the trendline doest not necessarly imply a trend reversal.

Also adjusting MA's and their crosses are a waste of time i`m certainly not doing that anymore.

Quoting Ethlao
{quote} The gaussian filtered rainbow MA looks great, thank you! If, like you say, we could use an indicator like this to accurately identify longer term trends (perhaps with a valid fundamental reason behind them), it should be possible to then trade in the direction of that trend intraday. The dream scenario is when you can enter with a smallish stoploss (and therefore larger position) at the beginning of a long term trend and then keep adding to your position on the way up, hopefully exiting somewhere near the top.
I`m waiting for that, i`m certainly more inclined into fundamentals than technicals, so i`d be happy to discuss that too.It seems like fundamental analysis is very discriminated in this forum, everyone uses technicals too much and its annoying.I would like to discuss fundamentals too , if someone would be interested in it.


Quote
If we were to adopt such a strategy I think we would need to concentrate on as wide a basket of currency pairs as possible because there might only be a few out of say 15 or 20 that are in a strong multi-week or multi-month trend at any one time. For this we would need to look carefully at the choice of broker as some offer better swap rates than others. Although I usually use it for entries, I've experimented with adding a moving average crossover alert to the chart along with your gaussian rainbow, as an added trend filter. My chart is for H4...

This is also optimization, every currency will have difference parameters for MA crosses, and its just annoying to set them all.Also they could change week-by-week so we need to constantly optimize it which is also annoying,
"There's a sucker born every minute" - P.T. Barnum
  • #80
  • Aug 20, 2014 6:49am Aug 20, 2014 6:49am
Quoting MaxDoom
I assume that you do all realise that any system which could make 1000% per month will fail because it is taking on too much risk...
Yeah maybe 1000% is just for getting the attention of traders, however that is not to say that i dont have that kind of expectation.We should do this and then we see how much we can get out of it.

Anyhow, we also have a risk cap, so the risk will never be too much.I focus on risk minimization even if it cuts our profits.
"There's a sucker born every minute" - P.T. Barnum
0 traders viewing now