Let's build together a system which makes 1000%/month | Page 12 | 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

  • #221
  • Aug 26, 2014 1:50pm Aug 26, 2014 1:50pm
Quoting PipMeUp
{quote} Absolutely not! Read again.
I will respond in PM to this
"There's a sucker born every minute" - P.T. Barnum
  • #222
  • Aug 26, 2014 2:00pm Aug 26, 2014 2:00pm
Quoting Proximus
{quote}... So only 5% of the membership (i dont know for sure but lets just assume that its 5 for a though) is uncommon, and has very innovative ideas. Now also innovative can be 2 sided, one being (2.5% of members ) absolutely BS, meaning they invent new BS methods, while the other are the Holy Grail inventors, they invent very very useful trading ideas and tools. So that being said 2.5% of the members are completely useless, while the other 2.5% are golden and very precious, and the other 95% being a nuance of them...
Hi Prox. I've said it before ... you are you are too optimistic ...

You almost go it, for your endeavour (95% + 2.5%) 97.5% of the members are completely useless, only 2.5% may be very precious.

One suggestion on (intraweek vs intraday durations), "ride the trend daily or weekly, it doesn't matter, until it's over" and why "close the trade at the retracement, and buy back at a lower price" the principle is good but for RR-Ratio itīs better to open various smaller trades with the trend and close them all in higher points along the "ride" ( maybe with some trailing-stop strategy ) and buy back at a lower price ( only if the trend is still up with confirmation ), best to use indicators to confirm open and close prices.

And yes you're right " the duration of the trade could depend very well on the strength of the trend".

You are in the right path.
Don't swim with the sharks ...
  • #223
  • Aug 26, 2014 3:42pm Aug 26, 2014 3:42pm
Quoting Proximus
Ok lets talk about topic. So i think we proved beyond doubt that the trend following method is better for our cause, i`m not saying that counter-trend is useless, but i`m saying that trend following is more efficient for us,used in this system. So if anybody wants to add something to the trend vs anti-trend debate,then do it now, because its time to talk about the duration of trades. So i want to hear opinions about the intraweek vs intraday durations, which one is better with our already established trend following method. So lets debate this please....
Attached Image


I love these magic screenshots! So, for the first buy, how did you decide to buy there? At that point in time the price had just dropped a huge amount and you had no purple trend line so what would have made you buy at that point?

For the exit, again what was the trigger to exit? You have magically selected the last high before the price took a massive drop, but at that point in time you did not know it was going to drop, so how would you have validated that exit?

I know that you had no triggers for entry and exit but were just marking a chart, but the marks are ideal and the real world is not. So, back onto the trend following discussion. What are you using to define a trend and do you have any evidence that it works?
  • #224
  • Aug 27, 2014 5:33am Aug 27, 2014 5:33am
Quoting beachloup
{quote} {image} I love these magic screenshots! So, for the first buy, how did you decide to buy there? At that point in time the price had just dropped a huge amount and you had no purple trend line so what would have made you buy at that point? For the exit, again what was the trigger to exit? You have magically selected the last high before the price took a massive drop, but at that point in time you did not know it was going to drop, so how would you have validated that exit? I know that you had no triggers for entry and exit but were just marking...
If i knew the aswer to that, do you think i would still be here discussing this?

The whole point of this thread is to determine some basic principles about the markets, and to build a system , and for that we need a cooperation, i am not all knowing to know the answers to everything, i share and already shared many things on this forum that i know.Now its the other guys turn to share what they know.

I mean thats the whole point of a forex forum to help eachother out isnt it? So i`m open to all ideas, lets discuss it, and if you feel you have something to contribute to this thread and discussion , please do it, thanks
"There's a sucker born every minute" - P.T. Barnum
  • #225
  • Aug 27, 2014 5:44am Aug 27, 2014 5:44am
Quoting beachloup
What are you using to define a trend and do you have any evidence that it works?
The trend can have many causes, as i pointed out in previous posts, but lets not confuse the cause with the effect.

And the trend is an effect, of a cause.

The trend has many definitions, but i think the most correct definition of a trend is: "an underlying and consistent bias in the price movement"
That is that the price goes (relative to a starting point) more in 1 direction (up or down) than it hovers around the start point.

So we define a yearly trend as the price moving more up or down compared to the year's start, than it stayed at the start.So basically a yearly trend for example is the strongest if the price went as far from the starting point as possilble.

Also the trend has to be more or less consistant, so if we want to be scientific then, it would also have a constant velocity, on a price*time chart, that is a constant speed of going in 1 direction, without acceleration or much change in velocity, in that process (no breakouts,whipsaws,etc).

So thats why trend lines work somewhat, but PipMeUp 's velocity indicator is a more quantifiable measurement tool than a trendline which is completely subjective.

Because the trend is visible to everyone, it must have an objective essence, and i think that can be very well measured with that type of indicator.
"There's a sucker born every minute" - P.T. Barnum
  • #226
  • Aug 27, 2014 8:16am Aug 27, 2014 8:16am
Quoting Proximus
Ok lets talk about topic. So i think we proved beyond doubt that the trend following method is better for our cause, i`m not saying that counter-trend is useless, but i`m saying that trend following is more efficient for us,used in this system. So if anybody wants to add something to the trend vs anti-trend debate,then do it now, because its time to talk about the duration of trades. So i want to hear opinions about the intraweek vs intraday durations, which one is better with our already established trend following method. So lets debate this please....
If using a trend-following strategy, the question is a trend on which timeframe? If, as most people on the poll seem to have voted, we are going to use higher timeframes to determine long-term bias, then it seems to me a sensible approach would be to use multiple timeframe analysis to determine the trend on an upper timeframe and then look for entries and manage trades on a lower timeframe to maximize gains (e.g. an H4-H1 or D1-H4 combo).

This requires at least two levels of analysis so, unless we are using a different system to determine bias (bull/bear/stand aside) on an upper timeframe and to execute trades on a lower timeframe, it seems logical that the entire system too would in theory be scalable for both intra-day and intra-week trading. That would be good as it would enable people with different time-availability and trading personality/style to use the same system on their preferred timeframe.
  • #227
  • Aug 27, 2014 9:25am Aug 27, 2014 9:25am
Proximus are you going to program it sir ?
  • #228
  • Aug 27, 2014 11:03am Aug 27, 2014 11:03am
Quoting Ethlao
{quote} If using a trend-following strategy, the question is a trend on which timeframe? If, as most people on the poll seem to have voted, we are going to use higher timeframes to determine long-term bias, then it seems to me a sensible approach would be to use multiple timeframe analysis to determine the trend on an upper timeframe and then look for entries and manage trades on a lower timeframe to maximize gains (e.g. an H4-H1 or D1-H4 combo). This requires at least two levels of analysis so, unless we are using a different system to determine...
Please dont be fooled by timeframes, there are no timeframes in reality, the entire market is a constant stream of bid/ask prices, tick by tick.

The timeframe serves only as a grouping mechanism as humans like to group stuff.I think if we were to use an intra-week/ day strategy, the H1 timeframe should be our basis (because it filters the least, and usually trends are visible here).We could use the M1 too, but then the price action would not fit in our monitor, so H1 is a pretty good timeframe for our objectives.

Also M30 is ok aswell, but we need to determine the power of the trend, for which we need a tool which measures it, like PipMeUp's velocity indicator is a very good tool for it.Also the indicator should work on M1 data, because the indicator can easily multitask, while human traders cant do that , especially not in the heat of the action.So was thinking of making an indicator which measures the speed& direction of the market.

After that you can take the trade wherever you want, because the indicator would give the same results on all timeframes, because the strength of the trend does not have a timeframe, its on every timeframe.

The fact that most of M5 trends dont make it till the weekly chart, is only due to a lack of strength, so you see the strength of trend is an absolute not a relative attribute of the market
"There's a sucker born every minute" - P.T. Barnum
  • #229
  • Aug 27, 2014 11:04am Aug 27, 2014 11:04am
Quoting KroneMad
Proximus are you going to program it sir ?
Probably, but we dont even have the basics of the system yet, so we need to work that out.
"There's a sucker born every minute" - P.T. Barnum
  • #230
  • Aug 27, 2014 4:36pm Aug 27, 2014 4:36pm
I agree that you need to be able to quantify the trend. You might find this interesting http://www.forexfactory.com/showthread.php?t=227948
  • #231
  • Aug 28, 2014 5:07am Aug 28, 2014 5:07am
Quoting beachloup
I agree that you need to be able to quantify the trend. You might find this interesting http://www.forexfactory.com/showthread.php?t=227948
Interesting method, however its a Voodoo BS.Sure it can be quantified and its objective, however counting bars in nonsense, as i said there are no bars in reality, the entire FX market is a stream of BID/ASK data, there are no candlesticks!!!

What you call candlesticks are only imaginary groups of BID/ASK prices taken by a period of time,so it doesnt make sense to count bars like that because in any different timeframe you will get different results.

What it would make sense is to count the ticks not the bars because the ticks are the basic price unit.
"There's a sucker born every minute" - P.T. Barnum
  • #232
  • Aug 28, 2014 9:10am Aug 28, 2014 9:10am
Quoting Proximus
{quote} ... so it doesnt make sense to count bars like that because in any different timeframe you will get different results. What it would make sense is to count the ticks not the bars because the ticks are the basic price unit.
Sorry Prox, but it doesn't make sense to count ticks either, the are values and countless transactions behind every tick, and if you don't take does into account, it is useless to count them.
Don't swim with the sharks ...
  • #233
  • Aug 28, 2014 9:48am Aug 28, 2014 9:48am
Quoting lcast
{quote} Sorry Prox, but it doesn't make sense to count ticks either, the are values and countless transactions behind every tick, and if you don't take does into account, it is useless to count them.
Maybe i didnt explained it well, but what i was trying to say is to count ticks by their direction, upticks and downticks, and see how they distribute.

Of course ticks have many transaction behind them, and they vary from broker to broker, from LP to LP, but its still the basic price unit you got, and if you can measure the distribution on the basic scale, then you can scale up that to any "timeframe" you want, but it should always start from the basics.

It would be just like this: http://forum.mql4.com/30911

Unfortunately the link to that guys website is not working, so we need to rebuild this indicator from scratch.But it would be an interesting concept, a "speedometer" which measures the speed (velocity) of the market, and if we find a stable constant velocity, then we can define that as a trend, regardless of timeframe and such.
"There's a sucker born every minute" - P.T. Barnum
  • #234
  • Aug 28, 2014 1:45pm Aug 28, 2014 1:45pm
Ok so we count the ticks up and the ticks down. Shall we count them this since the beginning of the market? In such a case the value is... the current price! Or do we count the ticks up and down say in the last minute, the last hour or the last day? Oh a timeframe!
Same question for the speed-o-meter. velocity = movement / time. If you measure the speed by (tick[t] - tick[t-dt]) / dt with dt the time elapsed since the last tick you'll get a funny and absolutely useless result.
Due to the non-linear (fractal-like) nature of the market you need a time scale to define a trend.
No greed. No fear. Just maths.
  • #235
  • Aug 28, 2014 1:52pm Aug 28, 2014 1:52pm
Quoting PipMeUp
... Same question for the speed-o-meter. velocity = movement / time. If you measure the speed by (tick[t] - tick[t-dt]) / dt with dt the time elapsed since the last tick you'll get a funny and absolutely useless result. Due to the non-linear (fractal-like) nature of the market you need a time scale to define a trend.
Hehehehehehe ... right on target

I know someone who is counting ticks for almost 6 months ( dreaming of HFT ), counting counting, while the fish is passing by, wasting time and money.
Don't swim with the sharks ...
  • #236
  • Aug 28, 2014 2:10pm Aug 28, 2014 2:10pm
Quoting PipMeUp
Ok so we count the ticks up and the ticks down. Shall we count them this since the beginning of the market? In such a case the value is... the current price! Or do we count the ticks up and down say in the last minute, the last hour or the last day? Oh a timeframe! Same question for the speed-o-meter. velocity = movement / time. If you measure the speed by (tick[t] - tick[t-dt]) / dt with dt the time elapsed since the last tick you'll get a funny and absolutely useless result. Due to the non-linear (fractal-like) nature of the market you need a...
Shit you got me there

However 1 thing is still true, the fact that we will base all indicators on tick data, or atleast M1 if we want to analyse the past, as tick data is not available in MT4.

We should base all the indicators of MT4 on M1 data, to be the most accurate, and we will filter it as we see fit, on whatever timeframe we wish to work with.
It offers atleast flexibility.

Anyhow, i dont know if its a great realization or not, but what i observed is that the ATR indicator itself is actually measuring the average speed of the market (excluding the rounding errors which is has in the MT4 build in function) as the ATR formula is average (High-Low) over X period, which is on a PRICE*TIME chart is the speed, and not velocity because it doesnt show the direction, only the net movement.

Now we could enhance the ATR by measuring the velocity instead the speed, and of course correcting the rounding error.Let me see if i can code this
"There's a sucker born every minute" - P.T. Barnum
  • #237
  • Edited Aug 29, 2014 12:02pm Aug 28, 2014 2:33pm | Edited Aug 29, 2014 12:02pm
Ok here is the quick indicator which i made, a corrected ATR, without rounding errors, and with maximum bar settings:

It will measure the average speed of the market, that is the rate of change of price, without including direction, /time (in this case due to timeframes / X nr bars which can be set by the parameter PERIOD)

Parameters:

PERIOD = As already said the number of bars for which to average the price difference
MAXBARS= Max nr. of bars painted on the chart , this does not affect the formula of the indicator in any way, it just saves PC resources to not paint more bars than needed

Attached Image (click to enlarge)
Click to Enlarge Name: EURUSDDaily.png Size: 55 KB


As you can see its almost identical to the built-in MT4 ATR indicator, however:

Attached Image


The built-in ATR has some rounding issues, so mine is not rounded at all.

Hope it can help you guys, however tomorrow i will try to enhance this by looking at the velocity instead and including the direction in it, instead of measuring the speed only

Attached File(s)
File Type: ex4 PROXIMUS CORRECTED ATR.ex4   6 KB | 458 downloads
"There's a sucker born every minute" - P.T. Barnum
  • #238
  • Aug 28, 2014 3:17pm Aug 28, 2014 3:17pm
Quoting Proximus
tomorrow i will try to enhance this by looking at the velocity instead and including the direction in it, instead of measuring the speed only
http://en.wikipedia.org/wiki/Average...movement_index
I'm almost sure it is available on MT4
No greed. No fear. Just maths.
  • #239
  • Aug 28, 2014 3:36pm Aug 28, 2014 3:36pm
Quoting PipMeUp
{quote} http://en.wikipedia.org/wiki/Average...movement_index I'm almost sure it is available on MT4
Almost, but not quite, the ADX is filtered trough an exponential MA which is a logical fallacy called: Recency Bias

Since i learned this, i dont use anymore exponential or any other weighting only linear (Simple Moving Average).

So we need a linear tool which measures the velocity.I will work on it tomorrow
"There's a sucker born every minute" - P.T. Barnum
  • #240
  • Aug 28, 2014 4:00pm Aug 28, 2014 4:00pm
Quote
Please dont be fooled by timeframes, there are no timeframes in reality, the entire market is a constant stream of bid/ask prices, tick by tick.
​


Huge mistake. Time is a fundamental concept of the auction process.
0 traders viewing now