Why I have been using the Linear Regression Indicator for over 8 years

[Forexmospherian]

Hi F,

Are you still using "15 - 25 - 45 -95 - 115 - 180 - 285 - 500 - 900 " on cTrader?

Cheers
Karmit

Hi Karmet

Approximately - yes

I have just downloaded another updated ctrader platform on a new computer with window's 8.1 operating system - and changes again.

For now new settings at 10 - 16 - 27 - 60 - 80 - 105 - 240 - 300 - 450 - slighty quicker but only had 2 days to find out how they perform.

All brokers platforms will perform differently according to your computer set up etc etc - thats what i have found over last 7 years - but never heard anyone else mention it - so whether its the inaccuracy of indicators or what I dont know.

I think it best to test slightly different setting to fit your style.

I have adapted faster settings because I have been using then so long - if you are not after quick scalps - you might want to slow them down slightly to slower than original etc

You will know when they are right - check them over a busy few days and also on at least 3 different FX pairs. When you get them agreeing in time windows to change - then you are on the right track


GL

Regards


F

 

[frostbliss]

Hi Karmet

Approximately - yes

I have just downloaded another updated ctrader platform on a new computer with window's 8.1 operating system - and changes again.

For now new settings at 10 - 16 - 27 - 60 - 80 - 105 - 240 - 300 - 450 - slighty quicker but only had 2 days to find out how they perform.

All brokers platforms will perform differently according to your computer set up etc etc - thats what i have found over last 7 years - but never heard anyone else mention it - so whether its the inaccuracy of indicators or what I dont know.

I think it best to test slightly different setting to fit your style.

I have adapted faster settings because I have been using then so long - if you are not after quick scalps - you might want to slow them down slightly to slower than original etc

You will know when they are right - check them over a busy few days and also on at least 3 different FX pairs. When you get them agreeing in time windows to change - then you are on the right track


GL

Regards


F

Hi F,
What is this time window? Are they time durations marked out n fixed on the chart? What's the use of these time window?

My broker platform got only Least Square Moving Average LSMA and parameter is limited from 1 to 100. But it seems to match the lower LRs in your charts posted in another thread.

Regards,
Frost

 

[Forexmospherian]

Hi F,
What is this time window? Are they time durations marked out n fixed on the chart? What's the use of these time window?

My broker platform got only Least Square Moving Average LSMA and parameter is limited from 1 to 100. But it seems to match the lower LRs in your charts posted in another thread.

Regards,
Frost

Hi frostbliss

I have my own 2 TW's every hour .

They are 9 minutes either side the 30 and 60 minute time frame changes.

I then have 6 key times an hour - which are within the 2 Tw's that I take maybe over 80% of my scalp entries from.

I use PA - TW's - LR's - TL's - interim S & R levels and price session structure to then line up and then cherry pick the scalps that should have the highest probabilities of making 7 -25 pips - normally within 30 mins#

Hope that helps


Regards


F

 

[myattitude]

ok so i just finished reading this thread this morning...

you say the settings will differ between platforms etc, so what kind of action do we need to identify the sweet spot settings?

this chart is for 1min eur/usd with settings 10,20,35 for short ticks (the grey-white ones) at 2:52pm

looking good like yours? if not, what kind of accuracy adjustments am i looking for?

 

[Forexmospherian]

ok so i just finished reading this thread this morning...

you say the settings will differ between platforms etc, so what kind of action do we need to identify the sweet spot settings?

this chart is for 1min eur/usd with settings 10,20,35 for short ticks (the grey-white ones) at 2:52pm


looking good like yours? if not, what kind of accuracy adjustments am i looking for?

These charts are on GU this afternoon

2 charts from after 2 00 pm to after 5 00pm on GU and then one blown up from for 4 00 pm to 5 00 pm to show the zoomed in quick LRs with their crosses

I have 2 many on here atm - 9 is ideal 3 super quick - 3 med and 3 longer term giving more of a session bias over 8 -12+ hrs

The more over - price should fall - the more under price - then price should rise

3 quick LRs take first priority for scalps but 3 Longer term show direction of main move - ie we have waves within waves etc

Zoomed in

Hopefully they will help - longest 2 need to be over price on way down and then mid 3 adjust with waves and tights 3 give you in TW's the clues for price changes etc

Regards

F

 

[myattitude]

Okay I've spent the last 2 days pinpointing this and I think I almost have it. I assume the 2 small black lines located tightly to the bars are your short term LRs and the wider coloured dashes are your mid-term LRs, in which case what are the contiguous coloured dashes? Those are the only ones I haven't done, I'm not sure if those were meant for me, since you said you have too many LRs.

This is my long term comparison chart:

This is the shorter term:

 

[Forexmospherian]

Okay I've spent the last 2 days pinpointing this and I think I almost have it. I assume the 2 small black lines located tightly to the bars are your short term LRs and the wider coloured dashes are your mid-term LRs, in which case what are the contiguous coloured dashes? Those are the only ones I haven't done, I'm not sure if those were meant for me, since you said you have too many LRs.

HI MA

That looks better.

Yes I am always trying new settings and trying to delve more into understanding wave structures using LR's - but normally 6 to 9 LRs are fine - I might test up to 12 and then drop 3 or more depending on the pair and the sessions.

Now - get this set up on say 3 or 4 pairs - - not to trade them simultaneously - but looking for some type of correlation - ie for example if at 8 30 am on Monday morning you see on the EU the quick LRs going under price and PA giving a HH and HL on the one min- its a scalp buy - that will move up either breaking through R areas from LR's or interim levels or fizzle out after 3 to 7 pips.

You will need a few days and weeks to watch and understand how the set ups come about - and to understand how you might buy for a gain of 12 pips whilst in a very bearish set up that suggests over the next few hours the price might drop 50 pips.

Remember we will be working of initial stops of say 5 -7 pips max - so a 10 pip move is good. You will overtime sense if your entry is spot on - or to exit and try again from another price 6 pips away etc

If you are happy with say 25 pip stops and don't want to two way scalp etc - the 3 longer LRs will give you more of the direction bias - sometimes 25 pip stops are not big enough - sometimes you might need 40+ pips - and for me I would be thinking if this pair is possibly going 40 pips against the intraday bias - I want 25 -35 pips of it on the wrong way - and then same again when it joins the intraday bias again - ie 50 -70 pips there rather than just hoping you don't get taken out.

Its worth asking some of the other guys - other than Major M how they are getting on with their set ups and understanding how to use them in time windows for best effect - ie dearlife / Fx works / etc.

I think with MM its took him time to fully understand the effect of every LR setting and match it with TW's and PA and trendlines and interim levels - etc etc

But in the long run - as MM has convincingly shown - it does work - and no 2 ways about it - used correctly - it will make you nice green pips.

GL and no problem with more questions etc as you progress

Regards


F

 

[metrader]

It's not a leading indicator. Lrs can only help in guessing future moves like some other indicators in case of a divergence.

If one puts a shift (-2 or whatever negative) on settings afterward it appears as if lrs predicted moves, but that only looks good in hindsight – can't be used in trading (if one is to trade crosses etc.)

Regards

 

[itspossible]

 

[MajorMagnuM]

It's not a leading indicator. Lrs can only help in guessing future moves like some other indicators in case of a divergence.

If one puts a shift (-2 or whatever negative) on settings afterward it appears as if lrs predicted moves, but that only looks good in hindsight – can't be used in trading (if one is to trade crosses etc.)

Regards

http://www.guppytraders.com/gup329.shtml

 

[tar]

Just another indicator ...

 

[metrader]

@ itspossible

Why do you find it irrelevant? - it’s a thread about linear regression indicator

They were talking about it being a leading indicator which is not. Why not help them understand it better?

 

[itspossible]

@ itspossible

Why do you find it irrelevant? - it’s a thread about linear regression indicator

They were talking about it being a leading indicator which is not. Why not help them understand it better?

im sry i just couldnt resist

 

[IFeelFree]

The optimal regression fit for a linear system is the Kalman filter. The Kalman filter is dynamic, in the sense that it is an optimal linear algorithm that updates the expected value of a hidden variable based on the latest value of an observed variable (in this case, usually price or spread). The Kalman filter is commonly used in mean reversion trading strategies, so that the "observed variables" are the prices of the trading instruments in one's portfolio, and the "hidden variables" are the optimal hedge ratio or weighting coefficients. It is often used with a cointegrated portfolio of stocks, ETFs, etc.

I'm currently using such a strategy with a Kalman filter applied to a cointegrated triplet of leveraged ETFs. I been placing trades once a week for the past 6 months, and I'm on schedule for a 50% average yearly return.