Multiple Timeframes

S

slavduja

I want to know if some intelligent folks on this forum can help me determine the best way to rank time frames.

There has been a lot of buzz nowdays about mutliple time frames trading. Normally you hear traders say I use a Entry Time frame, Intermediate and Longer time frame to gage my entries and exits etc.
Thats fine but its not clearly state why they use 3 and not 5? Why do they use 4min on entry and daily on long term??

How does one design a metric to rank the time frames ?
How do you know if you you should have time frames be in multiples of 2,3,4,5,6....n?
Do you calculate the correlation between the time series and trade ones that are least correlated??Logically the correlation would increase as the difference in time interval decreases. 1 min tim series price chart would be more correlated to 2min than 30 min.

To conclude say I wanted to use a SMA cross strategy on multiple times frames. Executing it on each time frame exclusively without referencing to others.Which would mean If i am long in on time frame X I may be short on time frame Y. How do I select the # of time frames? And why?

I want to see if others have looked into this. Prof Logic said he uses constant bars and he uses squares. Something like 7^7.. So each time frame is 7 times greater. I want to know why??How does he come to this specific calculation and on what doe she rate it on?


Thanks

Slav
 
What did you base your decision for 3x on?
Why 3 over 5 or 6 or 8 etc?
What I am trying to get to is , if you tested your system on various time frames and selected the 3x due to better performance.
Can you quantify why you like the multiples of 3 best?
 
Quote from slavduja:

I want to know if some intelligent folks on this forum can help me determine the best way to rank time frames.

There has been a lot of buzz nowdays about mutliple time frames trading. Normally you hear traders say I use a Entry Time frame, Intermediate and Longer time frame to gage my entries and exits etc.
Thats fine but its not clearly state why they use 3 and not 5? Why do they use 4min on entry and daily on long term??

How does one design a metric to rank the time frames ?
How do you know if you you should have time frames be in multiples of 2,3,4,5,6....n?
Do you calculate the correlation between the time series and trade ones that are least correlated??Logically the correlation would increase as the difference in time interval decreases. 1 min tim series price chart would be more correlated to 2min than 30 min.

To conclude say I wanted to use a SMA cross strategy on multiple times frames. Executing it on each time frame exclusively without referencing to others.Which would mean If i am long in on time frame X I may be short on time frame Y. How do I select the # of time frames? And why?

I want to see if others have looked into this. Prof Logic said he uses constant bars and he uses squares. Something like 7^7.. So each time frame is 7 times greater. I want to know why??How does he come to this specific calculation and on what doe she rate it on?


Thanks

Slav
More...

I suggest trading only one timeframe, whatever you feel comfortable with.

In my case, scalping or swing, there is nothing like the good old daily chart, no matter what instrument I trade.

Good luck.

ESD
 
There is no difference between the time frames. We are just arbitrarily breaking price into different time segments. When you get a good feel for PA you can see how it flows through time frames. You have to be able to see past individual bars and realize that price is continuous, it doesn't really matter how we cut it up. You just want a good view of prices past, thus you need multiple tf's higher than the one you get entries on.
 
it's all about trend isn't it

the two components of price movement are price and time
two problems are noise and sideways

I use what I refer to as a CheckList on which I analyze All timeframes at the start of
each new trading day - Month down to 15 min, M and W are carried forward
each tf is analyzed with the five components of my system plus Direction - ie which
way the price is moving is noted - the trend
I find using the CL very useful, tunes me in to the market, reminds me what I was
thinking yesterday, notes price and time targets that are being completed, and of
course sets me up for what I might expect the price movement to be doing during
my trading session - the trend

trading the euro, I'll trade the 6E Euro FX contract using a low volume tick chart, but
use a MetaTrader 4 demo to analyze
the MT4 has the 5 and 1 min charts displayed Tile Vertically and 15 30 60 4H and D
charts open, minimized, M and W closed
one of my system components is applied to all the charts and the 5 , 1 min and tick
charts analysis is ongoing and updated often

MT4 and some other charting programs have the means to use smas that are multi
timeframe, display a 60 or 15 min sma on a 5 or 1 min chart, further, there are many
MT4 signal indicators that display several tfs and indicators in a single display, so one
can see what other tfs are doing in terms of those other indicators while looking
at one chart - see attached


however
by including the longer timeframes in my CL / analysis, I now ask myself
'why bother trying to analyze all the noise and sideways
why not just Buy&Hold / Day-to-Days trade and
save yourself all that intraday trading trouble' . . .
 

Attachments

Thanks for the reply Wallace. Nice chart, i saw something similar done by The Rumpled One.
Anyways, how does one measure noise??
How would you quantify noise?

Thanks
 
slavduja:
"Anyways, how does one measure noise ??"
why would you want to ?
"How would you quantify noise ?"
why would you want to ?
each higher tf removes the lower level noise and sideways, but has its own amount
of noise and sideways

since the price movement is IRREGULAR, it cannot be systemized
price targeting is fairly easy and accurate, but time targeting which produces the
noise and sideways, is more difficult to achieve, but easier the longer the tf used
however, individual tfs - when not being related to other tfs, are discrete, and can
then be analyzed, price - time targeted, and traded within their own tf

the attached is a NinjaTrader chart with 15 and 5 and 1 min charts overlayed
bars are Box bars, 15 are grey, 5 green and red, 1 white
think something similar can be done in MT4
 

Attachments

Quote from slavduja:

I want to know if some intelligent folks on this forum can help me determine the best way to rank time frames.

There has been a lot of buzz nowdays about mutliple time frames trading. Normally you hear traders say I use a Entry Time frame, Intermediate and Longer time frame to gage my entries and exits etc.
Thats fine but its not clearly state why they use 3 and not 5? Why do they use 4min on entry and daily on long term??

How does one design a metric to rank the time frames ?
How do you know if you you should have time frames be in multiples of 2,3,4,5,6....n?
Do you calculate the correlation between the time series and trade ones that are least correlated??Logically the correlation would increase as the difference in time interval decreases. 1 min tim series price chart would be more correlated to 2min than 30 min.

To conclude say I wanted to use a SMA cross strategy on multiple times frames. Executing it on each time frame exclusively without referencing to others.Which would mean If i am long in on time frame X I may be short on time frame Y. How do I select the # of time frames? And why?

I want to see if others have looked into this. Prof Logic said he uses constant bars and he uses squares. Something like 7^7.. So each time frame is 7 times greater. I want to know why??How does he come to this specific calculation and on what doe she rate it on?


Thanks

Slav
More...

The longer the time frame, the more significant the waveform you are in. Start with monthly, then weekly, daily and for intraday it will depend on the volatility of the instrument. Profs algo are best explained by him. I will contact to determine if he want to do some teaching, as he has left ET for the time being.

When u use multiple times look for a cascading effect. Ur signal of choice should first appear on the monthly, then the weekly, etc. This requires patience, which is something most traders lack, it is the prime reason they fail.
 
Quote from bkveen3:

There is no difference between the time frames.
More...

I could not disagree more. Depending upon the volatility and average true range of the product in question, there is a TREMENDOUS difference in the choice of timeframes used for rules and studies.

In fact, for me at least, it is the source of significant edge.
 
You must log in or register to reply here.
Back
Top