This content explores the construction and comparative performance of trend-following trading systems, focusing on the impact of different time frames and trend identification indicators (EMA vs. Super Trend) using Nifty as a case study.
Mind Map
Clique para expandir
Clique para explorar o mapa mental interativo completo
Welcome to the long and the short a show
where you can expect an honest take on
trading something you won't hear
elsewhere. I'm your host Sepra. In the
last episode we broke down what trend
following really is and how Richard
Donain often called the father of trend
following helped shape systematic trend
following as we know it today. We even
tested some of his ideas right here on
Nifty. Today we are taking that a
[music] step further. I'll walk you
through how to build a simple trend
following system that actually works.
[music] And to do that, we'll focus on
two key concepts. Trading time frame and
trend identification. Time frames are
actually the price intervals that you
trade. And trend identification
approaches are about methods and
indicators that can be used [music] to
identify a trend within a given time
frame. All of this once again using
Nifty as a playground. All right, then
Like I always do, a bit of a side story
from the world of astrophysics.
Incidentally, I'm reading the book
Astrophysics for People in a Hurry by
Neil Degra Tyson. In it, he describes
how supernova explosions happen. Most
stars spend their whole lives just
burning quietly like a candle that never
goes out. Inside them, tiny atoms crash
together and make light. And that's what
keeps them shining for millions, even
billions of years. That's what the sun
is doing right now, just glowing
steadily. A slow and steady burn of
sorts while keeping everything on planet
Earth alive. But some stars are much,
much bigger than our sun. So big that
they live fast and burn out young. When
these giant stars run out of fuel,
something amazing happens. These stars
can no longer hold themselves and they
collapse, exploding into what scientists
call a supernova. It's like a massive
boom. That explosion to me feels a lot
like a breakout. Sudden, powerful, and
impossible to ignore. That steady burn
that happens before the supernova, well,
that's like a moving average. Slow,
persistent, quietly building momentum.
Markets tend to work the same way.
There's a long stretch of gradual
process, then boom, a breakout. [music]
That gives us two ways to catch a trend.
If you want to ride the slow buildup,
moving averages are your friends. If
you'd rather wait for the big moment,
that's where breakout tools like Super
Trend come in. Of course, there are many
ways to capture breakouts. Any channel
or rangebased indicator will do. I just
prefer Super Trend because it reacts
faster than most and catches the move
early. Even if you look at the research
and literature around trend following,
almost all of it comes back to these two
fundamental methods. [music] Take the
book following the trend by Andreas. I'd
call it the manual of systematic trend
following. He dedicates an entire
chapter to comparing exactly these two
approaches, breakouts versus moving
averages. Similarly, Ken Tropins's
Graham Capital, one of world's leading
CTA firms, has published several
insightful papers exploring how
professional managers identify and
capture trends. Again, their work also
highlights the same two pillars, the
steady signal-driven momentum captured
by moving averages and the explosive
threshold-based shifts captured by
breakouts. So, in a way, I'm not really
reinventing the wheel here. These
methods are foundational to trend
following. Time-t tested, datadriven,
and battleproven across decades of
market cycle. That's why we chose them.
That brings us to the next and perhaps
the most important question. Do time
frames of analysis really matter? By
time frames, I mean daily, 4hour, or
even 1 hour bars. Some people call them
time intervals or data intervals, but
all they refer to is the same thing. the
frequency at which you look at market
data. Now, in most books on trend
following, time frame selection is
treated as a matter of personal
preference, something you pick for
convenience, not as a parameter that can
truly change outcomes. But I found that
approach unsatisfying
except for the book technical analysis
using multiple time frames by Brian
Shannon. It also barely scratches the
surface, but it plants an important
idea. the power of looking at markets
through multiple time frames. Shannon
talks about aligning higher and lower
time frames so that your trades move
with the broader tide [music]
not against it. Later while reading the
book inside the black box by Rishi
Narang that concept really clicked. He
makes a simple but profound point. What
looks like trend on one time frame might
appear completely memereting on another.
That's when it hit me. Time frames
aren't just about convenience. They
define perspective. So when we build or
test a trading system, it's not enough
to pick one time frame and run with it.
We need to test a system across time
frames, compare outcomes, and understand
where the edge truly lies. Later when I
present the data for three different
time frames, daily, 4hour, and 1 hour,
you will notice the difference for
yourself. So we've spoken about two
fundamental building blocks of creating
a trend following strategy. Trend
identification method and time frame
analysis. Now let's get to the
interesting parts that is the back test.
A fair disclaimer, this is not a
recommendation to trade any of the back
tests or simulations that I share. These
examples use spot prices and are
discussed only for illustrative
purposes. Over to the back test. The
instrument or index I'm testing is Nifty
Spot. I will be testing all of this on
three different time frames. Daily,
4hour, and 1 hour bars. Across these
time frames, we will apply EMAs and
super trend to separately compare their
performance. The data here is from 2015
Jan to 2025 October. Let me give you a
quick example of how these strategies
are defined. For the EMA, we go long
when the price closes above the EMA and
continue holding the position as long as
the price stays above it. We go short
when the price closes below the EMA.
This setup is known as an S or stop and
reverse system. That means the exit for
a long trade is also the entry for a
short and vice versa. The system is
always in the market switching direction
whenever the signal flips. [music] With
super trend, the logic remains similar.
Only the reference point changes.
Instead of the EMA line, we use the
super trend bands. If the price closes
above the upper band, we go long. If it
closes below the lower band, we go
short. We will use a similar logic
across all time frames. The look back
period for EMAs would change. Of course,
let's start with the daily time frame.
For daily bars, I use a 8 period EMA.
And in super trend, I use the default
103 parameters. Since super trend is a
volatility based indicator, the look
back input [music] in it need not match
the EMA look back. Hence, I use the
default 103 across for this
illustration. In general, the 8 period
EMA on daily does a pretty bad job
compared to a 103 super trend. Of
course, by now, if you have been
watching the episodes of the long and
the short, you would know that it makes
little sense to look at the short side,
especially on daily time frames. So for
this one, let's just compare the long
only parts. Look at the table from the
top. Total P&L speaks for itself. And
now look at the average P&L that's 35
versus 270. That's a massive difference.
What that means is on an average per
trade, EMA system makes just 35 points
whereas super trend makes 270 points.
This happens because EMA takes a lot of
trades 256 versus 44 in super trend. And
that many trades is not always a good
idea as the system keeps chopping around
with many loss-making trades. The only
thing good about the EMA system is the
return to max draw down ratio. So it's
quite evident that super trend is
outperforming on almost all parameters
on a daily time frame. But what would be
the reason for such an outperformance? I
think it's because Super Trend adapts to
volatility rather than reacting to every
price fluctuation like the EMA does.
Since it uses the average true range to
set its bands, the super trend naturally
widens during volatile phases and
tightens when markets are calm, allowing
it to filter out noise and capture only
sustained directional market moves. On
an index like Nifty which has a strong
long-term upward bias, this volatility
adjusted approach keeps the system
invested during the large trending
phases and avoids frequent whips saws in
the faster indicators. In essence, Super
Trend strength lies in its built-in
noise filter, its ability to stay
committed to a direction until a genuine
reversal occurs, resulting in longer
profitable holds, higher average trade
returns, and overall better risk
adjusted performance, but with a higher
max draw down. Now, if you look at the
equity curves of both EMA and super
trend, it becomes quite clear. Let me
show you. The blue line trending up is a
line representing the long only on
daily. Almost no action till 2020 and
then it picks up. Now compare this with
the equity curve the super trend on
daily. Looks smooth, isn't it? More
importantly, it picked up the trend much
earlier from 2017 onwards with a minor
dip in 2020. So all in all on a daily
time frame, super trend wins on
consistency and risk adjusted
performance while EMA wins a bit on
speed and efficiency. Nothing much
otherwise. With the daily time frame
done, let's [music] now go one level
deeper. The 4hour time frame. Here the
look back period [music] for the EMA is
set to 21. Why 21? Because it maps to
the same 1.5 week duration similar to
the 8 period look back on daily. and the
super trend parameters remain the same
at 103. Here I want you to see both the
long only and the short only sections of
the table. On the 4hour chart, the
absolute P&L for longs may appear higher
at first glance. But what truly matters
is the average P&L per trade. 48 for EMA
and 153 for super trend. This is where
super trend clearly leads mainly because
it generates few trades. On the long
side alone, EMA triggers nearly four
times as many trades as super trend.
Once you factor in transaction costs,
most of the EMA's apparent edge may
disappear. Interestingly, as we shift to
this lower time frame, the short side
starts turning profitable. Particularly
for EMA, look at the table. While the
short trades on super trends are in
negative of close to 4,000 points, EMA
based shorts are slightly positive at
700. That's because emas react faster
and in nifty short trends tend to be
sharp and swift giving emas a natural
advantage there. When it comes to
drawdowns, super trend still shows
higher max draw down compared to EMA but
the gap between the two narrows
significantly. On the daily chart, Super
Trend's draw down was about 70% higher
whereas on the 4hour chart it's around
20% higher compared to the EMA. Also
note how the total points captured
increase as we move to a more granular
time frame. A sign that shorter time
intervals can pick up more market
movement though often at a cost of
higher trade frequency. Now let's look
at the equity and the draw down charts
for EMA. It is slightly more smoother
than the super trend. Look at the period
post 2024 especially. This smoothness
comes from the faster reaction time of
EMAs. As you can see, the draw downs in
the super trend on 4hour period are more
deeper. All in all, if you are exploring
the 4hour time frame, you may have to
drill down deeper to analyze the year
and monthwise returns and the average
P&Ls and then find the sweet spot. So,
we're now done with the [music] daily
and the 4hour bars. Now, let's move to
the 1 hour time frame. Among the three,
this is the fastest time frame that we
are going to look at. Here the EMA uses
a 50 period look back exactly what I
track in our weekly market metric series
while the super trend parameters remain
unchanged at 103. Let's start with the
tables on this time frame. We will
compare all the three long plus short
long only and short only. Now the key
difference [music] between the two on
this time frame lies in the gross P&L
points that is the total P&L. On that
metric [music] EMA definitely takes the
lead. However, when we focus on the
average P&L for long only trades,
[music] Super Trend still outperformed
but not as much as it did on daily. Same
on win rate. Super trend offers a way
better win rate for a trend following
system compared to the 50 period EMA.
42% versus 29%. Makes a difference on
the margins. Where things get really
interesting is on the short side here.
The EMA clearly wins with 4,700 points
versus 1,400 points on super trend
because it reacts faster and on a 1 hour
time frame, short trends in Nifty tend
to develop and fade quickly. The EMA's
responsiveness helps it to capture those
swift downward moves far better than
super trend does. If you want to see
that difference, look at the equity and
the draw down curve chart. You'll notice
that the gap between the long short and
the long only results is much wider for
EMA. That tells you something simple but
powerful. The value added by the short
trades in super trend is almost
negligible. Whereas for EMA, the shorts
contribute meaningfully to overall
profitability. You can look at the short
only equity curve as well. [music] The
green line in the super trend version.
It tends to give up on the gains more
often than not. Now that we've compared
EMA and [music] super trend across three
time frames, daily, 4hour, and 1 hour,
you might be wondering what exactly is
the conclusion here. Well, there's no
one answer. But what I can offer is a
framework, a way to think about it
systematically so that you can decide
for yourself which model fits your
trading style best. [music] Now, let's
go back to the key ideas. We started
with the time frame of trading and the
method of capturing trends. Talking
about time frames, you'll notice that as
we move to lower time frames, the gross
P&L increases. Now, that's an important
insight. One possible reason is that
maybe a large part of the trends often
unfolds intraday in what is called an
impulse leg, the first strong move of a
trend. Another factor is that market
corrections and crashes tend to start
during the day and continue over to
subsequent days. An hourly system can
therefore help exit long trades faster,
protecting profits during sudden drops
and that I think makes a huge
difference. Slower time frames like the
daily chart often miss this early
momentum. Of course, this behavior can
vary from one asset to another depending
on how its intraday price action tends
to be, but in our case with Nifty,
shorter time frames clearly seem to
capture more of the trend. Now, let's
turn to the two indicators we used, EMA
and super trend. Here's what stands out.
In certain contexts, like capturing the
short side trends, EMA performs better
because it reacts faster. But for
longside trends, super trend often comes
out ahead. It takes fewer trades and
gives each position more room to breathe
thanks to its volatility adjustment. So
maybe the final choice depends on what
you want to capture. If you want to play
the long only game, super trend might
suit you better. But if you want to
include short trades, EMA gives you that
extra edge. Or perhaps the best approach
is a blend of both. combining [music]
the stability of super trend with the
agility of EMA. But here's the thing,
don't take any of what I shared as
gospel. I just shared an illustration
with you all. On top of that, if you
tweak the lookbacks, add buffers to
reduce the trades, especially for EMA,
you may be able to get better results
than what I shared. Same with super
trend as well. I'll leave that work to
you. And that brings me to the end of
this part two episode on trend
following. I hope you found these ideas
useful. If you have any questions, drop
them in the comments and I'll do my best
to answer them. In the next episode,
I'll focus on the idea of trend
following across two uncorrelated
assets, Nifty and gold. And yes, don't
forget to subscribe to this channel. See
you in the next episode of the Long and
the Short Show. Till then, trade safe
Clique em qualquer texto ou marcação de tempo para ir diretamente àquele momento do vídeo
Compartilhar:
A maioria das transcrições fica pronta em menos de 5 segundos
Cópia com um cliqueMais de 125 idiomasPesquisar no conteúdoIr para marcações de tempo
Cole a URL do YouTube
Insira o link de qualquer vídeo do YouTube para obter a transcrição completa
Formulário de extração de transcrição
A maioria das transcrições fica pronta em menos de 5 segundos
Instale nossa extensão para Chrome
Obtenha transcrições na hora sem sair do YouTube. Instale nossa extensão para Chrome e acesse com um clique a transcrição de qualquer vídeo direto na página de reprodução.
