Multi-Timeframe Trend Analysis in Pine Script

in today's video we are going to be
dressing another question from someone
in the Mastery course and this someone
wanted to know how to adapt the regime
filter to work on multiple time frames
so if you're not aware of what a regime
filter is it's covered in great detail
in Andreas clanow's fantastic book
stocks on the Move highly recommend that
book if you haven't read it already it's
one of my favorite trading books same
with Unholy Grails by Nick Raj he uses a
market regime filter in his trading
systems as well this approach is best
suited to stocks although it does seem
to show promise in crypto and certain
other markets if you're creative with
how you apply it but the general concept
is that the market needs to be trading
above a higher time frame moving average
before you start taking trades on a
lower time frame so in this case we're
on spy we're on the S P 500 market index
we're on the one hour time frame and
this moving average here is a higher
time frame EMA right now it's set to
four hours but we could set it all the
way up to one week and now you can see
that it's turning green when price is
trading above this moving average and
turns red when it is not the whole
concept here originally is for long only
stock market strategies to only trade
when the wind is behind our back so when
the market is trending upwards on those
higher time frames and this script you
see down here
is the regime filter script adapted to
show three different time frames so
we've got the 240 minute time frame or
the um four hour chart we have the daily
time frame and the weekly and if I hover
over these labels on the end we can see
the actual EMA value there so what is
more effective than using one market
regime filter not necessarily three but
if one indicator is good then several
must be even better now I'm joking a
little bit there because a lot of
Traders do like to throw on a lot of
indicators on their charts and it's not
helpful but in this particular case this
Market regime Builder could be useful
especially on Lower time frames so if we
go down to like the 15 minute chart here
and you would adjust these higher time
frame moving averages you could really
make a simple rules-based objective
method for determining your directional
bias for day trading you know sometimes
it's difficult to know which way to
trade a market when you're using just
discretion but with an indicator like
this it's pretty pretty clear where the
momentum is and which direction you
should be trading in and as a little
bonus at the end of this lesson I've
implemented this new Pine logging
feature we open up the pine logs will
actually tell us how often the market
trades above all of these higher time
frame moving averages so if we go to the
one hour chart again you can see we have
a 54 chance that a bar will trade on
this time frame this one hour time frame
we have a 54 chance that the regime
filters will all be green what that
tells us is that the market likes to
Trend upwards at least on this one hour
time frame and at least since 2009 until
today which is a decent sample size by
the way we have 20 000 one hour bars
that we're analyzing here to get these
numbers so yeah interesting script here
let's break down the source code uh the
source code is not particularly complex
and hopefully you guys find it
interesting so let's open up the source
code here and break down everything so
I've already written out the code
obviously to save time the first thing
we're doing is getting our user inputs
so we have three time frames here we
have a moving average length which is 20
by default now this isn't trading advice
but obviously depending on the time
frame you're trading these numbers will
matter a lot so on a lower time frame
you would want these higher time frame
moving averages to be a lot closer to
your trading time frame so if you're
trading a five minute chart you might
want the 30 minute hourly and and four
hour time frames for example you want
the regime filter to be more sensitive
to short-term momentum if you're trading
a short-term time frame and if you're
trading a higher time frame like the
four hour or daily you might want this
to be the daily one week in one month or
something like that same with the moving
average length the lower the time frame
typically the lower you probably want
this number to be the higher the time
frame you might want to extend it to 50
100 or even 200 not 1200 that's probably
a bit too much the next input is our
market so this is the market we're
referencing to get all of these EMA
values you could get rid of this input
and simply pass in the Sim info dot
ticker ID into all of these security
function calls and then you would just
retrieve all of the market information
for the market that you've loaded onto
your chart but the whole point of this
regime filter in its original conception
is to trade it on things like individual
stocks if I go to S P 500 here
and we jump on to what looks interesting
Netflix now we are plotting the s p 500s
regime filters over the Netflix stock
chart and so what this will tell us is
when the broader Market is moving
upwards
um you can see that the regime filter
Now does not match the price action I
have on my chart because this
information down here is referencing the
S P 500 while all of this including this
higher time frame moving average is
referencing Netflix's price action and
how you would typically use this regime
filter is to only trade these members of
the regime filter universe so in this
case members of the S P 500 when the S P
500 is trending upwards if you were to
use this on crypto you'd probably want
to reference Bitcoin on altcoins and if
you're using it on something like Forex
maybe you want to reference the dollar
Index when you're trading dollar
denominated pairs like euro dollar
dollar Yen Etc but anyway let's go back
to the S P 500 for now and finish
breaking down the source code so before
we get any of this information we're
creating a custom security function that
does not repaint so I've gone over
repainting many times in the course and
on my YouTube channel so go back and
check those videos out if you are still
uncomfortable with repainting but
basically including this code here these
little historical operator checks this
check is checking the real time price
information the real-time bar if the
real-time bar is currently a real-time
bar not a historical bar we reference
that bar otherwise if it is a historical
bar we reference the previous bar now
don't ask me why this expression
eliminates repainting it's just
Blackmagic that the training view
developers have decided to implement on
this particular function and it works so
that's all I know and so when you're
using the security function make sure to
include this code expression if you do
not want your script to act differently
on real-time price movements compared to
historical price movements that's a big
problem with uh strategy scripts in
particular if your script does not
perform the same on historical data as
it does in real time then you're
obviously getting dodgy back testing
results which can lead to the
development of systems that aren't
profitable but look profitable on paper
anyway you don't need to have this
security function this Custom Security
function I just implemented this into
the script because it makes these lines
of code a lot easier to read otherwise
we would have six of these really long
lines of code on these six variables and
so to make the code more readable I've
just implemented this this custom
function which by the way is easy to
just copy and paste into your own
scripts and then you have your own
non-repainting security function anyway
moving on we get our EMA values for the
higher time frame so to do that we just
get the current time frames EMA this is
what is called an expression a code
expression and the security function
requires us to pass in an expression
that's why this is called exp short for
expression and that's what these three
are and these three are so for our
higher time frame EMAs we are passing in
the EMA expression and this will
retrieve this value from this market and
this time frame so EMA value one will be
our first time frames EMA value
ema2 will be the second and so on and
same with our price values we're getting
our first time frames closing price
second closing price and third closing
price and then to validate our regime
filters for each time frame we simply
check if each time frame's closing price
is above its moving average and that's
it that's all we need for a market
regime filter now if we just wanted to
have one color down here we could just
simply add in a new bull variable here
called regime filter equals regime
filter 1 and 2 and 3 Etc passing those
but for today's lesson the student who
asked this question wanted to see all
three separately so that's what we are
doing here which obviously complicates
the script a little bit not a lot
there's a bunch of code here that looks
scary if you're new to paint script but
it's all just copy and pasted blocks of
code so I'll just go over one of them
and explain what's going on here I
simply wrote this code out once copied
it and pasted it three times and change
the variables so that they don't
conflict with each other so for the
first regime filter I create two H lines
which are horizontal lines that's these
black dotted lines on my chart here and
once we have two H lines if we assign
them to a variable we can then use the
fill function to fill between these two
H lines and we can specify the color
that we want to fill between those two
lines so to get the color of our region
filter to make it either green or red we
simply check is regime filter 1 true in
other words is the closing price of our
first time frame above its moving
average if so we want to fill the two H
lines as green otherwise we want to fill
it as red and then just to make it a
little bit clearer what each ribbon here
represents I've added in some labels
these labels here tell us which time
frame each ribbon represents that just
makes it a lot easier to understand
what's happening just at a glance
without having to remember which ribbon
is which time frame so to do this I
create a persistent label VAR means it
does not reset on every new bar on our
chart and in fact because we delete the
label on each previous historical bar
you don't need these persistent variable
labels so I'll get rid of them for now
just to make the code a little bit
easier to read let's save that this
should change nothing there we go so
we're creating a label
it's set to n a and then we check if we
are on the last bar loaded onto our
chart then delete the previous bars
label and draw a new label on the
current bar the reason we need to do
this is because if we were to run this
script on a real-time price chart and we
had a few bars print these labels would
start to shift to the left and be drawn
over the top or underneath these ribbon
colors which obviously is not what we
want we want these to stay on the far
right so we delete old labels and create
new labels
the new label
just has our time frame as its text so
240 is this time frame it has a tool tip
so if we hover our Mouse over it will
tell us what the EMA value for that time
frame is it has a transparent background
color white text and the label style is
to the left which just means that it the
label looks like this
however I've set the background color of
the label to transparent so we don't
actually see that but the text
is to the right of the current bar and
that's it that's the script
um I've simply copy and pasted this
three times I've changed the price value
that we're drawing each label so this
label is drawn at 0.5 which is here this
label is drawn at 1.5 which is here and
this label is drawn at 2.5 which is here
and the H lines simply draw between 0
and 1 1 and 2 2 and 3 and that gives us
these even
um this even scale for this particular
indicator and that's it for this script
uh just as a little bit of a bonus outro
or conclusion here is
some code using the new Pine script logs
functionality so what we're doing here
is we're counting how many bars past all
three regime filters so if all three
regime filters are true then total bars
past is incremented by one we add up how
many bars had all of these filters met
if we are on the very first bar in our
chart we use the log dot info function
to plot a bit of text here that just
says our analysis starts here so we can
see at a glance what the very first bar
that our script ran on so that was back
in January of 2009 and then we check if
this is the last bar on our chart we
calculate the percentage of bars on this
time frame where all of our filters were
met so in this case that's four 54 of
the time so we have an obvious bullish
bias on this one hour time frame from
2009 all the way till today so that
tells me that if I was to be day trading
the S P 500 100 on the one hour time
frame I should probably favor long
trades over short trades objectively the
market likes to Trend upwards or the
momentum likes to be to the upside at
least with the settings we've inputted
into this current script and then we log
how many bars passed out of how many
bars have drawn on our chart so bar
index on the last bar will simply be the
number of total bars on our chart and we
add in the percent here so that we get
all of that information we had 11
173 bars out of 20 000 bars that passed
all three filters and then I also drew a
table here table Z or Z I don't know why
I put a z on the end it's just it can't
be called table because that's a
reserved keyword so we can call this
whatever you want but basically this
table has one column with three rows so
one column three rows it's set in the
middle Center and we set each row to
each time frame so so that we can see
which time frame which ribbon applies to
which time frame even when we go back on
historical data now the table isn't 100
accurate if you extend this too much it
uh gets out of whack but you get the
idea at least At a Glance we can tell
the top one is the weekly middle one is
daily bottom one is four hour anyway I
hope you found this lesson interesting
it's a pretty cool little application of
the security function and higher time
frame information I'll leave this video
here thanks for watching good luck with
your trading and I'll speak with you in
the next lesson take care