YT YouTube2Text

YouTube2Text → Transcripts → MASTER AI Trading in 70 Minutes (Full Claude + TradingView Course)

MASTER AI Trading in 70 Minutes (Full Claude + TradingView Course) — Transcript

by Mind Math Money · 12,171 words · 1,638 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00In this free video course, you will
0:02learn everything you need to know about
0:04AI trading for stocks, crypto, forex, or
0:09any other financial market. By the end
0:11of this course, you will not only be
0:13able to analyze any chart using AI
0:17directly in Trading View, but you will
0:19also be able to build, back test, and
0:21automate your very own trading strategy
0:24step by step.
0:28All right. So, hello everyone and
0:30welcome to this full course on AI
0:33trading. And before we begin here, I
0:36want to say something. This right here
0:38is probably the most important course I
0:42have ever made on this channel. And the
0:44reason is simple. AI or artificial
0:48intelligence is becoming a bigger and
0:50bigger part not only of trading but of
0:54basically our lives in general. And in
0:57the coming few years, the difference
0:58between a trader that really knows AI
1:01and a trader that is bad at using AI
1:05will be massive. But the good news is
1:08that right now as I'm recording this
1:10video, we are still so early. And my
1:13goal here is that by the end of this
1:15course, you will be on the right side.
1:18You will be one of these traders that
1:19are very good at using AI and as a
1:22result, you will make more money. But
1:25okay, so now let's begin here by taking
1:27a look at what you can expect to learn
1:31from this course. And before we can
1:33start using AI, connecting our AI to
1:36Trading View and start developing
1:38trading strategies with it, we of course
1:40first need to know what is AI trading in
1:44the first place. You probably already by
1:46clicking on this video have a little bit
1:49of understanding what we mean by AI
1:50trading. But after chapter 1, you will
1:53probably have a very clear
1:55understanding. And it's so important to
1:57have that understanding before we move
2:00on to take a look at how we can actually
2:02connect Claude. Claude is an AI. It's a
2:06LLM uh used as Shatbt. Uh but in my
2:10opinion, right now at least, it's
2:12actually the best uh AI out there,
2:15especially its latest model called Fable
2:175 is, you know, mindblowing in my
2:19opinion. And we will actually take a
2:21look at how we can connect claude
2:23directly to trading view. Trading View
2:26is the trading platform. If you have
2:28been following my channel, you know that
2:29that is the trading platform I use for
2:31all my analysis. And it is, I think, at
2:34least in the west the largest trading
2:36platform. So, we will be using Trading
2:38View throughout this whole course. And
2:40if you don't have Trading View yet, I do
2:42have a special link. Uh you will get,
2:44you know, a $15 bonus and you can try
2:46out the premium version for 30 days. So
2:49check that one out if you don't have
2:50trading view because I do think a good
2:52way to use this course is to have you
2:54know to basically follow along. So to
2:56have your trading view up and you will
2:58also need that as I said to connect to
3:00cloud. But after we have connected our
3:02AI to trading view we're going to start
3:04taking a look at some practical AI
3:08workflows. And with workflows, I
3:10basically means mean you know specific
3:13step-by-step ways in which we can
3:16actually actually use our AI. So I will
3:18try to show examples of how we can use
3:21the AI to find market structure. So
3:23things like support and resistance. Uh
3:25but we will also use the AI to you know
3:28use Trading View indicators. We will
3:30even try to do some multi-time frame
3:33analysis. And after chapter three, we're
3:37actually going to start taking a look at
3:38some, you know, more complicated things.
3:40We'll take a look at both building and
3:43back testing, you know, a rulesbased
3:46trading strategy with AI. And it's kind
3:49of magic because the AI can actually do
3:52all this on your own. Uh, and last, but
3:55definitely not least, we're going to
3:57take a look at AI risk management and
4:00how to set alerts. And we will also take
4:02a look at an example of a sort of daily
4:04routine uh that you can start uh that
4:07you can start using. Uh but here it's
4:10very important especially for this
4:12course that you watch the chapters in
4:15order. So in order to get you know the
4:17most value out of chapter 3, chapter
4:19four and chapter 5, you of course first
4:21need to install uh you know claude and
4:24trading view and you need to know how to
4:26connect them. Uh, and before you do
4:28that, you know, it's super important to
4:30know about what AI trading is. So, I
4:32highly recommend to watch the chapters
4:34in order, but if you need to, you know,
4:36rewatch a part or jump around in the
4:38video a bit, don't worry because I will
4:41include YouTube timestamps uh so you can
4:45jump to a specific part of the video.
4:47But now guys, this is a very important
4:48video. So, make sure to really lock in,
4:51remove all distractions, you know, go
4:53grab a coffee if you need one. uh you
4:56know we do have if you didn't know we do
4:58have these coffee mugs with you know
5:00candlestick patterns and chart patterns
5:02that you can check out in the mymath
5:04money store. Another thing that you know
5:07really helps me when I try to learn
5:08things is to actually use a physical uh
5:11pen and paper. This is the mymath money
5:13trading journal that you can check out
5:15if you want to. So yeah guys grab a pen
5:17and paper if you want to maybe grab a
5:19coffee and without further ado let's
5:21jump right into it. But right so now
5:24then what is AI trading? Well in very
5:30very simple terms AI trading just means
5:33that we're using artificial intelligence
5:36in your trading process and using AI in
5:42your trading in trading processes has
5:44actually been around for a very long
5:46time. In fact, in today's financial
5:48markets, most of the buying and selling
5:51usually happens by an AI. And this was
5:54true even before the Shatbt era and the
5:58era of LLMs. But what is very very
6:02exciting now is that the power of AI is
6:07starting to become accessible to you
6:10know normal traders like for example me
6:12and you. Right now, as I'm recording
6:13this video, you and me have the same
6:17access. We have access to the same
6:19powerful models as you know the top
6:22traders uh on Wall Street. It has never
6:26been more equal. You know the access we
6:28have to models. Of course, the
6:31institutional traders and smart money
6:33traders have their own advantages. They
6:36have more capital. They usually have
6:38more advanced data and so on and so on.
6:40But both you and me can actually access
6:43the most powerful large language models.
6:46And in this video uh later on in this
6:48course, we will take a look at how you
6:50can connect, you know, the best model
6:53directly to trading view. So I'm super
6:54excited about that. But here when we
6:57talk about AI trading, we're talking
6:59about everything. It's basically a range
7:02from AI assisted analysis to fully
7:06AIdriven systems. So on the one hand you
7:10can be a trader that do maybe you know
7:1390% of the work on your own and you use
7:16maybe AI you know for the 10% that you
7:19need help with. So maybe you need help
7:22with a certain aspect of your trading
7:24maybe it's risk management maybe it's
7:26back testing testing your strategies and
7:28so on and so on. So you can be almost uh
7:31you know fully human but using a little
7:33bit of AI but it can be all the way to
7:37you know 100% of your trading being
7:41fully done with AI. And in this course
7:44we're going to do a little bit of both.
7:46So so we are going to use AI a lot but
7:49at the end of the day in this course
7:52we're going to focus on that you in
7:54front of the screen will take the final
7:56decisions. So you will use AI to sort of
7:59make your analysis but it's you who
8:03decides you know whether you actually
8:05press the buy button or whether it
8:07whether you actually press the sell
8:09button. But here just to repeat
8:12institutions or in other words trader
8:14with with lots of money. Think about
8:16things like hedge funds or big banks and
8:19so on and so on. They have have used AI
8:22for decades. This is nothing new.
8:25But now as I said that edge is now open
8:29to everyone. AI the same AI that is used
8:32in Wall Street can now uh we retail
8:34traders can now use. And once again I
8:38can re-emphasize that in this course we
8:40will mainly focus on using AI as your
8:43trading assistant where you take the
8:46final decisions. And if we quickly look
8:48at this uh image right here, you can see
8:51the spectrum here from AI assisted
8:54analysis all the way to 100% AIdriven
8:58systems. And this right here is so
9:01important that the edge uh from AI is
9:05now accessible to everyone. And that is
9:07why this course is so so important. But
9:12right. So now then now I want to give
9:14everyone a little bit of a reality
9:16shake. We need to understand what an AI
9:20can do and what an AI can't do. Because
9:23if you have made it to this part in the
9:24course, you are probably already locked
9:26in. And as you can see on me, I'm super
9:29excited. But I want to give you guys
9:31realistic expectations. My channel here
9:33on YouTube is not the one to watch if
9:35you want to get, you know, become a
9:36millionaire overnight. There are already
9:39so many channels claiming that out there
9:41on YouTube. I really want to set
9:43realistic expectations before we
9:45continue here. So, let's actually start
9:47here by taking a look at what it can do.
9:50What are the upsides of AI and and it
9:53already can do a lot. So, it can do
9:56things like analyzing live charts. Yes,
10:00AI can now read, you know, every single
10:02candle on your chart. It can also apply
10:05uh technical analysis concepts like uh
10:08you know support and resistance. It can
10:10use indicators and so on and so on. It
10:12can do any math you need to do on the
10:15chart and it can help you stay
10:17objective. But what I want to say right
10:20away here and this is very important. It
10:22can't predict the future or guarantee
10:25profits. And when I mean that it can't
10:28predict the future, I mean that it can't
10:30predict the future with 100% accuracy.
10:33In fact, no AI system has ever been able
10:37to do that. But what an AI can do is to
10:41help us with our analysis and make the
10:44odds in our favor. So it can improve the
10:47odds of us taking successful trades and
10:50as a result make more money. What is
10:54also very important to notice here is
10:56that AI is improving every single you
11:00know week. If you are following the AI
11:03space, you might be confused about all
11:05the new models. There are basically
11:06coming uh coming new frontier models
11:09every single month now and they are
11:11improving and improving. And this is why
11:13it's so important to learn AI as soon as
11:18possible because if you don't learn it
11:20uh right now, there is a big risk that
11:22you will get left behind. But what is
11:25very important about this course is that
11:27what you will learn here will be useful
11:30in the future as well. So when new
11:32models inevitably come, you can still
11:34use this workflow and all of the things
11:37you learn here can make you stay ahead
11:39of the curve while AI is improving. So
11:43my goal here is that this course will
11:44actually be be useful years and years
11:47from now. But right so now then now now
11:50the time has come to take a look at how
11:52we can connect claude which is our AI to
11:58trading view which is our charting
12:01program. And this part is very very
12:03important. So make sure to follow along
12:05step by step. And first of all to make
12:08this connection we are going to use the
12:10trading view MCP.
12:13And what this basically is is that it's
12:15a free open-source bridge or connection
12:20between Claude and Trading View. And
12:22this allows Claude that is our AI to
12:25directly see everything that happens on
12:28the screen. So it can re read the real
12:30time live uh live price data and it can
12:34also control the chart automatically. So
12:37what Claude actually can do is that it
12:39can for example draw support and
12:40resistance lines without you even
12:42touching the key the the keyboard. It's
12:45actually amazing. And here the
12:46connection consists of four pieces. This
12:49is not really something you need to
12:51memorize or something you need to learn
12:52but it consists of basically a node.js
12:56the MCP which is the uh connector and
12:59we're actually going to use a debug
13:01mode. But as I said uh don't really
13:03worry about that because I have created
13:05a a guide and a prompt that will guide
13:07you through how to connect trading view
13:09to cloud step by step. But what is very
13:12important that is that for this to work
13:16you need the trading view desktop app.
13:19So you should not use trading view in
13:21browser. You need to download the
13:22trading view desktop app app. And I
13:25actually think you need a play paid plan
13:28for this. And here, if you don't have a
13:30paid plan, don't worry because I do have
13:32a special link to Trading View. You can
13:33use that link to get Trading View
13:35Premium for free for 30 days. And if you
13:37later on decide that you want to
13:38upgrade, you can also get a $15 bonus.
13:42So, you need Trading View and you also
13:45need to get the Claude desktop app. But
13:48now, I'm going to guide you through how
13:50to connect Claude to Trading View step
13:52by step. And to do this, I have actually
13:55written a guide here on
13:57mindmathmoney.com. I will make sure to
13:59link this guide in the description of
14:02the video or you can just go to
14:04mindmathmoney.com. You go to free
14:06educational guides right here and you
14:08scroll down and press right here. How to
14:10connect Claude to Trading View. And
14:13here, if you're interested more deeply
14:15on how this works, I actually recommend
14:18to read through this guide. But if you
14:22scroll down a bit here,
14:24I have actually created a prompt that
14:28you can copy. So what you want to do is
14:30that you want to copy this prompt right
14:31here. And then you actually want to
14:33place the prompt within Claude because
14:36here is the magical thing about AI. The
14:40easiest way to install this connection
14:42is to actually use Claude itself. So by
14:46using this prompt, Claude will guide you
14:48guide you through the installation
14:51process step by step. This was exactly
14:54how I did to connect a trading view to
14:57Claude. And I have made this prompt to
14:59be beginner friendly and step by step.
15:03But here real quick before you make this
15:05connection, I also recommend that you
15:07actually check out the official uh
15:09GitHub page for this Trading View MCP
15:14or in other words for this connection
15:17between Trading View and Claude. And I
15:20also want to uh really emphasize right
15:22now that this uh this uh connection
15:25between Trading View and Claude is not
15:28an official tool made by Trading View.
15:30This is not made by Trading View. As you
15:33can see here on the GitHub page, this
15:35tool is not affiliated with or endorsed
15:38by Trading View. And this GitHub page uh
15:42explains everything. It explains, you
15:44know, how it works and why it's safe to
15:46run and so on and so on. So before doing
15:49anything, I actually really recommended
15:51that you read through uh you know, the
15:53disclaimers and the important parts of
15:56this page so you really know what to do
15:58before you do it. But as you can also
16:00see here, this page has, you know,
16:034.3,000
16:05stars. So this is a connection that is
16:08already used by, you know, thousands and
16:10thousands of traders. But yeah, a good
16:13habit to develop, you know, as soon as
16:15possible in the AI age is to always, you
16:18know, read through everything and really
16:20understand, you know, what you are
16:22installing before you do it. Okay. Okay,
16:25so what you now need to do is that you
16:26need to download the Claude app and you
16:29do that on claude.com/d
16:32download and you also need to download
16:34the tradingview desktop app and you do
16:37that on tradingview.com/desktop.
16:41Then you want to open up a new chat
16:43inside claude and then you want to paste
16:46the prompt that I showed you guys
16:47earlier that is in the article. This is
16:50the prompt that will guide you through
16:51the installation process step by step.
16:55So you paste this prompt right here and
16:58here depending on your claude plan. As
17:01you can see right here, I'm currently
17:02using the best model which is Fable 5.
17:05But depending on when you watch this
17:08video, Fable 5 might not be available.
17:11Uh but what I always recommend doing
17:13when you use AI uh is if possible you
17:16want to always use the best model
17:18because the difference between you know
17:20a model that is at the very top and a
17:22model that is just a little bit below
17:24there can actually be a very big
17:25difference in efficiency. So if you can
17:28um uh I recommend to actually use the
17:30highest model uh especially when you do
17:32things like installing things and you
17:34know uh using it for trading I highly
17:36recommend the the highest model. So I'm
17:38going to use Fable 5 here. And as you
17:41can see uh in this installation process
17:43I will not show the whole process but it
17:45should be able to guide you through
17:47everything you need. So for example now
17:49it asks ask you you know are you on a
17:51Windows or Mac? So I'm on Windows so I
17:54press Windows. Uh it asks you you know
17:57do you have the trading view desktop app
17:59installed installed? Yes it's installed.
18:02Yes, cloud is installed and so on and so
18:04on. And it will start to work through
18:07work you through this whole process. And
18:09as you can see here, step one, you want
18:10to check if node.js is installed. It
18:13shows show you everything, you know,
18:15everything really step by step. But all
18:18right, so I really hope the installation
18:20process went well. If you have any
18:23questions regarding the installation
18:25process, I actually highly recommend to,
18:28you know, ask Claude about it. it's, you
18:30know, much more intelligent than me in
18:33many cases. Uh, but of course, if you,
18:36you know, stumble upon any problems
18:38also, you can leave a comment down below
18:40and I will try my my really best to
18:42help. But now we have open up Trading
18:44View. Here we are, as you can see, we
18:47are on Bitcoin on a daily time frame and
18:52Claude is connected to Trading View. And
18:56here, as you can see, I usually like to
18:58have uh this on my second screen. So, I
19:00usually have Claude on one screen and
19:02Trading View on another. But so you guys
19:04can see everything at once, I will keep
19:06them both at the same screen. Uh so here
19:09I'm just doing a simple test. I'm asking
19:11is Claude uh connected to Trading View.
19:14Uh sometimes you need to allow for
19:16things to happen. And as you can see
19:18right here, yes, Trading View is
19:20connected and healthy. I can see your
19:22open chart, Bitstamp, Bitcoin USD. So it
19:25actually sees it. Uh it sees that we are
19:28on the daily time frame. Uh it says I
19:31can control your charts, your
19:32indicators, p uh pine script alerts,
19:35everything. So it already starts to feel
19:38a little bit like magic. But we are just
19:41getting started. But right so now then
19:44now when you have training view
19:46connected to Claude, let's start using
19:49them together. And let's begin here by
19:51taking a look at how to prompt when
19:54Claude sees your chart. So these are
19:57basically some tips on what to think
19:59about, you know, when you prompt Claude.
20:01It can be very overwhelming in the
20:03beginning because you can pretty much
20:04ask it to do to do everything. But it's
20:07important before you prompt to
20:08understand what Claude can uh what
20:11Claude sees and knows and what it
20:13doesn't. So here what Claude already
20:17sees is basically the assets. So for
20:19example, are we looking at Bitcoin or
20:21are we looking at Tesla or Forex or
20:23whatever? It can also see, you know, the
20:25time frame. It can read every candle and
20:28so on and so on. So the way to prompt it
20:30here is to go straight to the question.
20:35And what I like to do here is that I
20:37like to ask, you know, one specific
20:39request at a time. So ask for one task
20:42at a time. And what you can also do is
20:44that you can share your view and let
20:47Claude challenge it. So you can for
20:50example, you know, say that if you think
20:52the chart is in an uptrend, you you can
20:54say, you know, you can ask, for example,
20:56Claude, are we in an uptrend right now?
20:58And basically see if it agrees with your
21:01analysis. And what I also like to do is
21:03to name the concept clearly. So for
21:06example, when we prompt Claude, are we
21:08talking about market structure? Are we
21:10talking about supply and demand? Are we
21:11talking about liquidity? Be, you know,
21:14concise, but be be very precise. So,
21:17these are basically my beginner tips in
21:19how to prompt Claude. So, let's try a
21:22super simple prompt just to get started
21:25right here. Let's just write, you know,
21:27what is Bitcoin trading at right now.
21:31And once again, while Claude loads a
21:33little bit, I do once again want to
21:35emphasize that I personally highly
21:37recommend to use the highest model when
21:40it comes to using Trading View because I
21:42I have been trying to use a little bit
21:44of lower models. I tried Opus 4.8 and
21:48set 5. And even though this is just one
21:51step down in the quality of the model, I
21:54can actually already notice a pretty big
21:57difference in uh you know the accuracy
22:00of of the outputs. But here you can see
22:03that uh it has loaded uh completely and
22:06it says Bitcoin is trading at 63,187
22:10right now on the bit Bitstamp spot price
22:14and as you can see that is actually
22:15exactly the price Bitcoin is trading at
22:17right now. So it's super super cool. It
22:21really does have the uh the live
22:24data. But right. So now then now let's
22:28take a look at the first workflow or in
22:31other words the first way in which we
22:34can use claude inside trading view. And
22:37my first recommended way to use it is to
22:39use it for market structure analysis.
22:43And if you didn't know, market structure
22:45is basically when we look at the chart
22:47as a whole. So this is when we are
22:49answering questions like, you know, are
22:51we in an uptrend, are we in a downtrend
22:53or maybe in a sideways market. We look
22:56at things like trends like break of
22:58trends. You might have heard of, you
23:00know, break of structure or for short
23:02BOS and change of character or for short
23:06CH, OC. So these are all things we can
23:10you know uh ask Claude to analyze. But
23:13here, if you are completely unfamiliar
23:15with market structure, in order to get
23:18the most out of your AI, I really think
23:20it's important to have a good
23:22understanding of market structure
23:24yourself. And luckily, I do have a full
23:27course on market structure. So, I will
23:29make sure to link that course both up in
23:31the video card as well as the
23:32description below. Uh, so I highly
23:34recommend to check that course out if
23:37you don't know market structure. If you
23:39have a good understanding of market
23:40structure, the AI will be much more
23:43useful. So, we are still in an era where
23:46the best traders are basically traders
23:49that have good knowledge on their own
23:51and they use the AI. If you don't have
23:54knowledge on your own and use the AI,
23:56you can still make devastating mistakes.
23:59But here, just as a few examples, you
24:01can, for example, verify, you know,
24:02swing points. Uh, but your structure
24:05knowledge is the quality check. So your
24:08own knowledge is still you still need
24:10your own knowledge as I said and if you
24:12need more market structure knowledge I
24:14highly recommend as I said to check out
24:16my market structure course but all right
24:19so now let's try a market structure
24:22prompt right here so I'm going to write
24:25find the most important swing points on
24:30this chart use the visible candle
24:36candles and By visible candles, I say
24:38that Claude should look at the the chart
24:41that I can see right here and analyze
24:44just this chart. Uh mark the major swing
24:48highs and swing lows and tell if we are
24:54in an uptrend or downtrend and then
24:58press enter. And now we can just sit
25:01back and relax and let Claude do its
25:04thing. uh currently sometimes you will
25:07need to you know uh click allow uh for
25:11some actions for cloud to do some act
25:13some actions and because we are using
25:15the best model here it will take you
25:18know a lot of time to think but I really
25:21really think it's actually worth it you
25:23know it's worth to wait you know extra
25:25time to get more intelligence and better
25:28analysis and now you can actually see
25:30that without me doing anything uh the
25:34chart is uh claude is actually starting
25:36to mark up you know major swing highs
25:39and even the levels and major swing lows
25:41and even the levels. So you can see here
25:43it has marked you know multiple highs
25:45and lows here. So here you can see all
25:47in the chat you can see all the major
25:49swing points and you can see here that
25:51the verdict verdict is that we are in a
25:55downtrend right here and here to be
25:57honest for a first try I actually think
25:59that this is very good. I think it found
26:02this major swing high super clearly. You
26:05know the swing high right here super
26:07clearly. This is how I would have drawn
26:08it as well. It found the lower high
26:11right here super clearly and then yet
26:13another lower high. So very clean uh on
26:15the highs here on the lows.
26:19I wonder a little bit about this one.
26:21I'm not sure why it drawn uh draw this
26:24higher low right here. But it clearly
26:26found the swing low right here. It found
26:28a lower low and then a lower low. But
26:30here, remember that my prompt was pretty
26:32vague. I just said, you know, find the
26:35most important swing swing points. And
26:37that that can be hard for the model to
26:39understand. So if you want more specific
26:42rules, you can be more specific with
26:44your questions. So for example, what do
26:47you actually mean by a swing low or
26:49swing high? And you can also be more
26:51specific about the range. So maybe start
26:54finding swing swing highs and swing lows
26:56from all the way back in, you know,
26:58October. And just to show you guys a
27:00little tip if you are lazy is that you
27:03don't need to, let's say you want to
27:05remove the drawings. You don't need to
27:06do this on your own. You can just say
27:08remove all the drawings you just did.
27:13And then just press enter and Claude
27:16will automatically remove everything.
27:19And now, as you can see, all the nine
27:21drawings, the seven lab labels and the
27:23dashed lines are removed and the chart
27:27is clean. Thank you for that. All right,
27:31so now let's take a look at our second
27:34workflow, our second way to use AI
27:38inside Trading View, and that is to draw
27:41support and resistance, or in other
27:44words, key levels. And what I like to do
27:47here is that I like to use claude to
27:50find strong levels or the most
27:53significant levels. These are levels
27:55that, you know, many traders are paying
27:58attention to. Um, and the reason I like
28:00these levels is that these are the
28:02moments where we can really start to see
28:05some predictable behaviors. So, I really
28:07like to draw out these strong levels.
28:11So, you can use it to, for example,
28:12find, you know, tested swing points. You
28:16can fi find areas with maybe high
28:19volume. You can find for example round
28:22numbers or psychological levels
28:25and in simple terms claude will find
28:28them. They can explain why they are
28:30important or you know why they are not
28:32important and of course it will draw
28:36them automatically on your chart. But
28:40here it's always a good exercise to, you
28:43know, not trust the levels blindly.
28:45After Claude has drawn the levels, you
28:46want to pause and check the levels
28:48yourself if they sort of make sense.
28:52Okay, so now let's jump back in Trading
28:54View. Let's actually draw out the chart
28:57a little bit more so we can get the more
28:59significant levels. Let's open up Claude
29:02right here. And one thing I really think
29:04is useful in Claude is that you actually
29:06don't have to write your prompts. If you
29:09want to, you can just speak instead. So
29:12now I'm going to use the recording. So
29:15uh wait, I'm going to uh remove it. So
29:18I'm just going to uh get it clean. Okay.
29:20So now I'm going to give my prompt.
29:23Okay, Claude. So I want you to find the
29:26most significant levels on this chart. I
29:30want you to start from the beginning of
29:332024.
29:35And what I want you to do, I want you to
29:37draw out the levels that are really
29:40important. Think about the levels that
29:42many traders are paying attention to and
29:44levels that are likely to get, you know,
29:48uh, very strong reactions in the future.
29:50So, think about things like all-time
29:52high levels or levels that have been
29:54clearly tested multiple times. uh also
29:57think about uh you can also consider
29:59volume. So levels where we haven't seen
30:01lots of volume are more significant you
30:04know take your time and make sure that
30:06each level make sure that you have good
30:09reasons for drawing each level
30:12and then press done. So as you can see
30:16now when you speak you can get much more
30:18sort of precise with your prompts. So
30:21let's hope that Claude does this well.
30:23Let's press send message right here. And
30:26don't worry if it works for a while. You
30:28know, when you give this kind of
30:30prompts, I actually like to see it
30:31working for a while because that means
30:33that it's using, you know, a lot of its
30:36intelligence, uses a lot of its test
30:38time compute, which is a fancy way for
30:42basically saying that the AI is
30:44thinking. All right. So after a lot of
30:47thinking uh the model is now done and we
30:51have multiple multiple levels here on
30:52the chart. We have a total of 12
30:55drawings and here uh uh Claude draw the
31:00resistance lines with red and the
31:03support lines with green. So the levels
31:06to the upside with red and the downside
31:09with green here. And what is very very
31:11nice about this is that you will get the
31:13explanation for each level. So for
31:16example, the first resistance is pretty
31:19clear. It's the level at around, you
31:20know, 126,000. Uh this is the all-time
31:24high level. And as you can see, it even
31:26says right here it says here you have
31:29the all-time high. It's triple tested
31:31here in between
31:3324,000 to 126,000. And as you can see,
31:38it was tested one time, two time, three
31:40time. Maybe if I did this myself, I
31:42would preferably have drawn it so that
31:45it touches, you know, maybe draw it out
31:47as a zone like this would make so make
31:50sense in my opinion. But very nice that
31:52Claude found it. Another very important
31:54area is the one here in between around
31:57112,000 to 108,000. It's these two red
32:01lines right here. And once again, I
32:03would have probably drawn this out as a
32:06area right here. But I think that
32:07overall, this is actually super super
32:10impressive because as you can see, this
32:12was indeed a super important area. You
32:14can see we had one touch, two touch,
32:15pretty much uh three touch, four touch.
32:17We broke this one and then we sort of
32:19flipped this previous resistance right
32:22to become support for a while. It also
32:25found this resistance right here, this
32:27clean touch right here that previously
32:29acted as support. Uh, and here it
32:32actually drew out a zone.
32:35Here you can see it actually drew out
32:37drew out a major support zone where we
32:40have multiple touches here recently. So,
32:43to be honest, guys, for being, you know,
32:45a one-time prompt. This right here is,
32:49you know, in my opinion, it's so good.
32:52You know, it's not perfect. It's not,
32:53you know, exactly how I would have drawn
32:56the levels, but it has found the most
32:58significant levels. Maybe it overdraw,
33:01maybe it drew a little bit of too much
33:03lines. Uh, in my opinion, but it found
33:05the most important levels. Um, and here
33:09you can always improve. So, for example,
33:10next time I prompt it, I might instead
33:13of ask asking it uh for drawing lines, I
33:16might ask ask it to draw zones. So, I
33:18can write something like, you know,
33:20focus on support zones like you drew the
33:22one at, you know, uh, 57,000 to 60K. And
33:27this is one of the magical things about
33:29AI is that you the more you prompt the
33:31AI, the more you will get used to it.
33:34And uh the better the AI will become and
33:37it can make you better. So it can
33:39basically be a positive cycle where you
33:41make the AI better and the AI make you
33:44better and so on and so on. But overall,
33:46I think that Claude gets a big thumbs up
33:50for the support and resistance workflow.
33:53And now, as always, if you're too lazy
33:55to delete all the drawings yourself, you
33:57can just write remove the drawings now,
34:01please. And as you can see, this was
34:03super quickly. Just after like 5 to 10
34:05seconds, all the drawings looks removed.
34:08It says done. The chart is now clean and
34:11the drawings are removed. Okay, so now
34:14it's time to take a look at our third
34:17workflow. And this right here is using
34:20Trading View indicators and indicator
34:24analysis. So here it starts to get
34:26really really interesting because you
34:28can use Claude to for example
34:31automatically add indicators but you can
34:33also automatically generate for example
34:36buy and sell signals. You can
34:38automatically change the settings as
34:41well. Claude has the power to pretty
34:44much do anything a human can do. So, as
34:48I said here, Claude can, for example,
34:49add indicators by just typing or even
34:53using voice like I really do. You don't
34:55need to worry about any menus and stuff
34:57like that. And what Claude also can do
35:00is that it can read the live values and
35:04also compare them to price action. So,
35:07it can not only read the indicators, but
35:09it can also tell you what the indicators
35:12actually tell you about the price. Yet
35:15another thing it can do is that it can
35:17check confluence. And in simple terms,
35:20confluence in trading is basically when
35:22you have multiple signals pointing in
35:25the same direction. So for example, if
35:27you have one, let's say you have a
35:29bullish price action signal like a
35:31candlestick pattern, but at the same
35:33time you might have let's say you have a
35:36bullish signal on the MACD as well and a
35:39bullish signal on the Fibonacci. Now we
35:42have three bullish signals happening at
35:44the same time and that is what is known
35:46as confluence. The more signals
35:48happening at the same time the more
35:50confluence we have. So you know one type
35:54of confluence is if the indicator for
35:56example agree with the overall trend and
35:59so on and so on. And here the claude
36:02tool works with you know pretty much any
36:05indicator. So you can for example use
36:07you know the RSI, the MACD, moving
36:09averages and so on and so on. Uh but
36:12when it comes to using just as with many
36:15other workflows when it comes to using
36:16indicators I highly recommend to learn
36:20about the indicators before you, you
36:22know, just blindly use AI. But luckily I
36:25actually have full trading courses on
36:27all of these indicators on the channel.
36:30Uh so I will make sure to pop up a
36:32playlist right here. In this playlist,
36:34you can find all my most important uh
36:37trading courses right here on on
36:39YouTube. And all of these trading
36:40courses are completely for free. Uh so I
36:43highly recommend to, you know, no matter
36:45what indicator you use, you know, dive
36:47in to the course so you really fully
36:49understand the indicator.
36:51Okay. So now we're back here in Trading
36:53View. We are on Bitcoin and we have our
36:57Claude assistant ready to work here. And
37:00let's just start with something simple.
37:03Let's add an indicator to our chart. And
37:05to make it really simple, let's just
37:07add, you know, let's add a simple moving
37:09average. And as you guys might know by
37:12now, I prefer to actually prompt using
37:14my voice because you can get uh more uh
37:17details and it's also faster. So now,
37:20let me prompt our assistant here real
37:22quick. So, we press record. You can also
37:24click Ctrl D.
37:26Uh sorry about that start. Ignore what I
37:29said. Ignore Ctrl D. Uh what I now want
37:32you to do is that I want you to open up
37:34a simple moving average indicator. And I
37:38want you to do this using the indicators
37:40tab. So you open up the real uh simple
37:43moving average, not using the API.
37:47And what I also want you to do at the
37:49same time is that I want you to change
37:52the color of the simple moving average
37:54to be purple. Um and I want it to be the
37:57thickest line possible. And remember it
38:00should be based on 50 days and make sure
38:04that it works. Uh and yeah that is your
38:07task. And then we just send this message
38:10right here. And hopefully cloud will be
38:11able to to understand this prompt. All
38:15right. All right. All right. It says
38:18done. The SMA is now live on your
38:20Bitcoin USD chart. Uh so it looks pretty
38:25good here. It looks like it added the
38:27simple moving average. Uh to double
38:29check, we can go up here to the settings
38:32tab just to check that it is indeed, as
38:34you can see, it's 50 days. Uh we are on
38:37the daily time frame, so that is
38:38correct. And if you look here at the
38:40style tab, you know, fable, I asked for
38:42the thickest line and I asked for
38:44purple. Uh so it basically solved the
38:48whole prompt in one shot. That was very
38:52uh very very impressive.
38:54Uh but now okay let's sort of try to
38:58test this AI assistant to the limit. So
39:02what I now want to try to do is that I
39:04want to try to make it add a MACD
39:08indicator and without me telling it uh
39:12you know the settings I want it to
39:14research and find the right settings and
39:17change the settings for me. So let's try
39:20this prompt right here. Hello Claude.
39:24very good job with the uh adding the SMA
39:27to the chart. That was really, really
39:29good. Now, I have a harder task, but I
39:32think you can do it. What I want you to
39:33do is that I want you to add a MACD
39:36indicator using the same method. Use the
39:38indicators tab and add a normal MACD,
39:41but then I want you to change the
39:43settings to Linda Rashki's settings. Uh,
39:48and to find out the settings, you know,
39:51you you will probably have to look it
39:52up. I'm not going to tell tell you the
39:54settings. Uh you need to find them out
39:56yourself, but I'm sure you can do it. Uh
39:59so add the MD, change the settings to
40:02Linda Rashki settings. And yeah, that is
40:05your task. And then we hit enter and let
40:09Claude do its thing. Uh its thing. This
40:12is a pretty complex complex task, so I
40:15assume it will have to think for quite a
40:18while. Uh but hopefully it can uh it can
40:23do that. Uh so it's already looking
40:26here. It added a few tasks. You can see
40:28it's searching the web. Uh it's looking
40:31for the right settings here.
40:34So I think it it will be able to do it.
40:37Uh but I will be back when Claude has
40:39done its job. All right guys, so Claude
40:42is already done here with the task. This
40:44was actually pretty fast. Uh, I wonder
40:47if you can see the thinking time. Uh, it
40:50was pretty damn fast to be honest. Done.
40:52The MACD is on your chart with Lind
40:54Dashki's settings. So, here it looks
40:59like it actually researched that the
41:00correct settings is the famous 310
41:03oscillator. That is correct. It has a
41:05fast length of three, a slow length of
41:0610, and a single signal length of 16.
41:10So, let's double check that the work
41:11here is correct. We go to the settings
41:13tab. We go to to inputs and just like
41:16that beautifully we have the 31016 MACD
41:20which is uh the Linda Rashki uh MACD
41:23settings. So Fable was able to do both
41:26of these things. And guys, this is just
41:29scratching the surface. I've showed you
41:31how to add indicators and how to
41:33automatically sh check settings, but you
41:35can of course also ask uh Claude any
41:38question about the indicators. Uh so
41:41here you have I I highly recommend to
41:44you know experiment as much as possible.
41:46The the opportunities are basically
41:48endless.
41:50But right so now the next workflow I
41:53want to take a look at is actually
41:56multi-timeframe
41:57analysis. So we can actually use an AI
42:01to analyze a chart not only on one time
42:03frame but we can use you know two time
42:05frames, three time frames or even more.
42:08This can uh if we if done correctly this
42:11can actually improve our odds of success
42:15and as a result we can make more money.
42:18Um but here just some basics here with
42:21time frames. We usually use a higher
42:23time frame for bias and a lower time
42:26frame for entries. And what I mean by
42:29this is that we can for example let's
42:31say we identify a you know uptrend on
42:34the daily time frame then we can
42:37actually zoom in to maybe you know the
42:394hour time frame or maybe the 1 hour
42:42time frame to get precise entries to get
42:45more specific entries. Of course, it's
42:48very possible to enter at the same time
42:50frame where you find the trend, but in
42:52many case cases, you can actually find
42:54better opportunities if you combine
42:56multiple time frames at the same time.
42:58And here the magical thing with claude
43:00is that Claude can switch time frames
43:02and uh compare the structures
43:05automatically. So for example, you can
43:08let uh you know Claude analyze maybe uh
43:11four chart windows at the same time and
43:13it can sort of uh put all of the
43:16different time frames together and give
43:19you the result of the analysis.
43:22Okay, so now we're back here once again
43:24on Bitcoin. I will try to just keep on
43:28this chart the whole course to keep it
43:29consistent. We are still on Bitcoin. We
43:31are still here on the daily time frame.
43:36And now let's actually try something.
43:39Now I want to give it some kind of
43:41prompt that it can use to analyze
43:44Bitcoin on multiple time frames.
43:47So let's say something like this. Uh
43:52great job with the MACD. I have now
43:54deleted it to make the chart a bit more
43:56clear. But what I now want to do is that
43:59I actually want you to analyze the chart
44:01you're looking at right now on three
44:04different time frames. I want you to
44:06analyze it on the daily time frame. I
44:08want you to analyze it on the 1 hour
44:10time frame and the fiveinut time frame.
44:13And if you are unclear about you know
44:16what period uh I'm talking about or you
44:20know anything else, you can ask me
44:23questions before you start this task.
44:27So here I used a technique that can be
44:29very useful and that can make your AI
44:31assistant understand you better and that
44:33is to simply ask Claude to ask me
44:36questions so it's more clear about what
44:38I'm looking at uh or what I want because
44:42when I just say when when I simply say
44:44you know analyze the charts on multiple
44:46time frames it might be a bit unclear
44:48even even if you know I were to ask you
44:51to do that you would be like uh but yeah
44:53what what do you mean by analyze on
44:55different time frames? Should I focus
44:57on, you know, uh, what I'm currently
44:59looking at on the screen or what is the
45:01time frame? Maybe you you would have
45:03some questions. So, let's put this to to
45:05to Claude and let's see what we work
45:08with here. Okay. So, here it's asking
45:11what kind of analysis do you want on
45:13each time frame? Uh, let's say, you
45:16know, full technical read, but keep it
45:21very short, simple, and concise. Uh, how
45:25should I deliver the analysis? Uh, let's
45:28use a chat summary. It can actually also
45:30have, you know, a big report file or
45:32chat plus screenshots, but I'm going to
45:34keep it simple.
45:36Uh, should I mark up the chart? Uh,
45:39let's just keep the chart clean and get
45:41the full analysis in the chat. So, here
45:44it actually asked me three questions. I
45:46answer the questions. So, now we just
45:48can sit back and relax and let Claude do
45:51its job. All right. So Claude is done
45:55with the analysis. And while it was
45:57working, uh, yeah, it you could actually
46:00see it flipping to the different time
46:01frames and analyzing. Uh, but here we
46:05get a short summary. I'm not going to
46:06read the full one. Uh, but here you can
46:08see that on the daily time frame, it's
46:11bearish, but trying to find a bottom
46:13bottom. And you know, just by looking uh
46:16briefly, it looks like we might have
46:17some sort of bottom here. Just from
46:19experience, I might uh it might be some
46:21divergence here. If we uh have a
46:24momentum indicator on the chart, it
46:26looks like we might have some
46:26divergence. Uh so that looks, you know,
46:29correct just by looking. And here on the
46:321 hour time frame, it says that we are
46:34neutral to a bullish range.
46:38And if you look here at the 1 hour time
46:39frame, you can indeed see that we are in
46:42actually this this was good. You know,
46:44we are in a pretty clear trading range.
46:47We can see that by marking our uh
46:49resistance and support and we can also
46:51see that of course by looking at the
46:54flat moving average. A flat moving
46:56average indicates that we are in a
46:58range. Uh so very good job so far in my
47:02opinion. Uh last but not least it's uh
47:06on the five minute time frame it says
47:07that we are looking bullish uh intraday.
47:10So let's see how the five minute chart
47:12looking. And as we can see here, it's a
47:16pretty solid analysis because we have
47:17recently broken out of this sort of five
47:20minute range right here. Got a strong
47:22breakout. So we are indeed, you know,
47:25looking pretty bullish here on the 5
47:27minute. So just from a quick look, I
47:29haven't looked at this uh in depth, but
47:31from a quick look, it looks like it has
47:33a pretty solid understanding even when
47:36looking at multiple time frames at the
47:38same time.
47:40But right, so now it's finally time here
47:43for workflow number five with AI and
47:48that is how to build a rulesbased
47:52trading strategy. And of course when it
47:54comes to building a trading strategy,
47:56this is a long process and I do actually
48:00have a full course where I go uh you
48:02know step by step I give you guys a
48:05process you can use to build your own
48:06trading strategy. But even when it comes
48:08to building your own trading strategy
48:10and getting the rules, Claude can be
48:13super useful. Uh but first of all, let's
48:16just talk a little bit about the basics
48:17about the trading strategy. So first of
48:19all, a real trading strategy needs to
48:21have exact rules. So we need to know
48:25when to enter, where to set our
48:27stop-loss and where to place uh our uh
48:30target. So we need you know some
48:32condition from for buying, selling and
48:35taking profit. And here this process we
48:39can actually start with just a simple
48:41sentence to Claude and then Claude can
48:44help us make the rules exact. And of
48:47course before we trade with real money
48:48and so on and so on we need to back test
48:50this the strategy and it's also a very
48:53good idea to first uh before you trade
48:56with real money you can trade with fake
48:58money that is called paper trading. And
49:00I do have a a video where you can learn
49:02how to get started with paper trading. I
49:04will make sure to link that video up in
49:05the video card. and the description. Uh
49:08so here a key part of building your
49:10trading strategy is that you want to
49:11keep on improving it in until every rule
49:15is objective uh and testable at least
49:18preferably and especially if you're
49:20going to use AI to automate the the
49:23strategy. Uh you want to make sure that
49:26the rules are actually you know uh
49:28correct and that the strategy of course
49:30works. uh and here claude can help you
49:34make this uh process that's usually weak
49:38maybe done in an afternoon that is the
49:40power of AI yes in the beginning it will
49:43take longer but when you really master
49:45AI it can you know speed up so many
49:49workflows so many processes so don't
49:52underestimate this
49:55right so now let's actually take up
49:56claude right here you can use claude not
49:59only in trading view. You can pop it up
50:02wherever you want on your computer. As I
50:04said, you can have it on your main
50:05screen. Uh I usually have it on my
50:07second screen, but to show you guys, I'm
50:09having it here on the main screen. Uh so
50:11now,
50:13let's actually just uh oh uh let's
50:16actually prompt Claude to try to build a
50:18very simple trading strategy uh using
50:20the simple moving average we already
50:22have. Uh so let's do it.
50:26Okay. Hello, Claude. Now, I want you to
50:29help me build a super simple trading
50:32strategy. And the trading strategy,
50:34let's make it as simple as possible.
50:36Let's use the simple moving average we
50:38have on the screen. And let's just say
50:42that every time the price goes above the
50:45moving average, we want to buy and every
50:48time the price goes below the moving
50:50average, we want to sell. Uh, and I
50:54don't want you to do anything with
50:55Trading View right now. I just want you
50:57to set up in a very concise way uh my
51:00trading strategy in text uh so I can
51:03review it before we do anything else
51:05with it. And then I press done like that
51:10and enter. And now we just let Claude
51:13cook for a while and let's see uh the
51:16output. Okay. So after just a few
51:19seconds uh Claude is already done here
51:22with trading strategy. So the strategy
51:24SMA 50 cross uh so here it basically
51:27lists up everything. So the chart is
51:29Bitcoin USD the indicator 50 period
51:32simple moving average of the close entry
51:35when the price crosses uh above the SMA
51:38open a long position and exit when the
51:40price crosses below the SMA close the
51:42position. Uh uh position handling uh
51:45always either long or flat. Okay so this
51:49was actually something I didn't specify.
51:51So maybe the model could have guessed
51:54that I meant uh go long every time uh
51:56it's above the moving average and go
51:59short every time it's below. But now
52:01it's that we buy every time it's above
52:04and we sell our position every time it's
52:06below. Uh so it's a little bit
52:08different. And here signal confirmation.
52:11This is very important and this is
52:12something I didn't specify to the model
52:15but here it has chosen to uh uh base its
52:19calculations on the candle closes. So
52:21the candle need to close above the
52:23simple moving average to confirm the
52:26signal and this is very important. We
52:28need specific rules and it was good to
52:30see that fable added that specific rule.
52:32And here it's very important that I'm
52:34not you know here I'm not saying that
52:37this is a good trading strategy at all.
52:39In fact, I do think especially if you
52:42count with the commissions uh for
52:45trading, I do expect that this one will
52:48at least, you know, I think it will it
52:51might be a little bit profitable, but I
52:52think it will at least underperform,
52:55you know, the buy and hold strategy. So,
52:57I think that is usually a good benchmark
52:59to compare your trading strategy with
53:01either uh an index like the S&P 500 or
53:04with just buying and holding. But for
53:06now, this is all we need. All right. So
53:09now I want to switch the attention and
53:10take a look at how you can back test
53:13your trading strategy with AI. And if
53:16you don't know back testing is basically
53:18taking a look at how your trading
53:21strategy would have performed in the
53:23past. So how would your rules uh have
53:27performed in the past? And in this case
53:29we can use you know an instant back
53:31test. So we just you know look back uh
53:33basically using for example your AI
53:35assistant claude which we use throughout
53:38this whole video. Uh but you can also
53:40use your own for example drawn trades.
53:43Maybe you have you know uh long and
53:45short position tools in trading view or
53:47you have your own data of past trades or
53:50anything like that. You can also use for
53:52example trading views uh uh bar replay.
53:56It now has I think it has actually a
53:58builtin uh back testing now. But here
54:00the magic is that claude can calculate
54:03the win rate, the riskto-reward and
54:05expectancy and you don't need to be able
54:08to code at all. So this is the magic of
54:12claw. But right so now let's jump back
54:14here into trading view because now I do
54:17actually want to use the trading
54:19strategy we built earlier in the course.
54:22Um so the trading strategy it's a very
54:25simple one. It's simply that we buy when
54:27the price closes above the simple moving
54:29average and we sell when the price goes
54:31below the simple moving average. Now I
54:34want Claude to automatically back test
54:36how this strategy would have performed
54:39in the past. Uh so let me prompt it
54:42using my voice. All right Claude. So now
54:44I want you to do a full back test of
54:47this trading strategy. So, I want you to
54:50use, you know, Trading View's builtin uh
54:53strategy tester and I want you to use
54:55the exact rules we recently uh put out
55:00and you can actually do two uh sort of
55:03back tests or you can have two
55:04variables. One can be used uh without
55:07any commissions. So, how would this one
55:09have performed without commissions? And
55:12the second test can be to to make it
55:14more realistic. So, have some
55:16commissions for every buy and sell. uh
55:18you can simulate it on let's say a
55:20$10,000 account
55:23and let's say that you based base it on
55:26a sample size of 100 so 100 trades uh so
55:31we have a good sample size and yeah that
55:33is your task make sure to do it properly
55:36and no shortcuts and uh yeah we'll see
55:39you soon but yeah now we need to let
55:42Claude the cook here for a while this is
55:43a pretty big task you know testing 100
55:45trades uh but you can See, it's already
55:48starting to produce some code right here
55:51for the back test. Uh, so super excited
55:54to see how this one turns out. And while
55:57you watch this, like you just mentioned
55:58that, think about all the sort of
56:01opportunities this opens up. You can
56:03just, you know, build your own trading
56:05strategies, back test them, and yeah,
56:08just having the power of Fable to, you
56:11know, normal retail traders like me and
56:13you. I think it's, you know, such a, you
56:15know, good thing. I'm super excited
56:17about the AI AI in general. And yeah,
56:20this is this is just the beginning. If
56:23you made it this far in the video, uh I
56:25always forget to say this, but you know,
56:26drop a like on the video. It really
56:28helps out more than you think. Uh if you
56:30have, you know, found any help in this
56:32course so far, uh you know, hit that
56:34like button if you want to. And I will
56:36see you guys when Claude is done with
56:39this. All right. All right. All right.
56:41So, it looks like the analysis here is
56:44complete. And as you can see on the
56:47numbers, they look super super
56:49impressive. Uh let's see here how it
56:53went. As you can see, the back test is
56:55complete. Uh both of the back test. And
56:58if we take a look here at the net profit
57:00for this trading strategy, it was
57:04without commission. So without fees, it
57:06was a net profit of $6 million and uh
57:10without fees, a net profit of $4.8
57:14million. So, it's insanely insanely
57:17profitable. Uh, but what we can notice
57:20here is that this trading strategy seems
57:22to go from all the way back here in May
57:262015. And remember that Bitcoin has gone
57:29up a lot since then. So, I do want to
57:32ask a question
57:34uh later on. You can also see here for
57:36example max draw down here at around 50%
57:40uh average% per trade. And here it also
57:44explains you know the strategy. It says
57:46this is a classic trend following
57:48profile. It actually loses 70% of the
57:51time uh but makes it all back and more
57:55on the few very profitable trades. So
57:57you can see this in the graph here that
57:59it goes pretty much sideways, you know,
58:01losing a bit, losing a bit and then it
58:03spikes up, losing a bit, losing a bit,
58:05spikes up, losing a bit, spikes up. So
58:07it's basically a few very very
58:10profitable trades that makes this
58:12trading strategy
58:14uh that make this trading strategy work.
58:16And here in trading view you can see
58:18that uh Claude Fable automatically built
58:21this whole uh back test. And here you
58:24can get some more information. You can
58:25see total profit, total loss, the
58:28commission here, uh the total P&L and
58:31you get all this information right here.
58:33But if you are unsure about uh anything
58:36you can always ask cla and one thing
58:39that is very important that I want to
58:40ask right now is this. Okay. So this
58:45looks very very impressive on paper you
58:48know uh it sounds amazing getting that
58:51uh type of uh annual return but please
58:55compare it with just a buy and hold
58:58trading strategy. How much would I have
59:00profited if I simply just held Bitcoin
59:03during this whole time?
59:06Because this is super super relevant. So
59:08maybe if the strategy of just holding
59:10Bitcoin was better than this trading
59:12strategy, then you know it wouldn't
59:14actually make a lot of sense. If you can
59:16can just buy and hold the asset, then
59:18all of this sort of fancy back testing
59:21and buying and selling would maybe just
59:24have made it worse. So let's see here
59:27if Claude can compare uh this trading
59:30strategy to just a buy and hold trading
59:32strategy. Okay. So here straight from
59:34the strategy testing's own benchmark you
59:37can see that a buy and hold strategy
59:40would have also made you uh you know
59:42multiple millions. You can see it would
59:43have made you uh 26,000%. But you can
59:46actually see that in this case uh the
59:48simple SMA 50 cross uh trading strategy
59:51even with fees uh highly outperformed
59:55the buy and hold and that was actually a
59:57little bit surprising for me. Uh so here
1:00:00simply buying for 42 bitcoin at $236
1:00:03in May 2015 and never touching it would
1:00:06have made you a multi-millionaire too.
1:00:07The strategy roughly 1.8xed
1:00:11buy and hold. So the SMA strategy was
1:00:14actually almost twice as profitable. But
1:00:17here it also says the strategy's edge
1:00:18here isn't really buying. It's the not
1:00:21holding through bare markets. And this
1:00:24actually gives us a little bit of
1:00:26insights into the usefulness of
1:00:29indicators like the SMA. So it says here
1:00:32that the reason the SMA strategy was
1:00:34more profitable was actually that it
1:00:37made us avoid the very very bad
1:00:40drawdowns. So for example the worst draw
1:00:43down during the buy and hold was a draw
1:00:45down of 83%.
1:00:48So as you know when it when it comes to
1:00:50risk management if you fall down you
1:00:52know 83% you almost need to 10x your
1:00:55account to come back and what the SMA50
1:00:58strategy uh basically did was that it
1:01:01allowed us to avoid the worst draw
1:01:04downs. So here we're actually getting
1:01:06some deep insights into the usefulness
1:01:09of uh indicators like moving averages.
1:01:13And now I just wrote, you know, great,
1:01:15thank you. Now remove the strategy
1:01:17because I'm lazy and I just I'm just
1:01:19going to let Claude H clean up the chart
1:01:21here. Right. So now let's take a look at
1:01:23one of the best use cases for AI in
1:01:26trading and that is to use it for risk
1:01:30management. When it comes to trading,
1:01:34one of the I would say the most
1:01:36important skill you can learn is risk
1:01:38management. Because even if you have a
1:01:40profitable trading strategy, you might
1:01:42have a trading strategy that actually
1:01:44would make you, you know, very rich over
1:01:47time. If your risk management is not in
1:01:49shake, you might never reach that point
1:01:52because you might, you know, uh,
1:01:53completely wreck your account. Risk
1:01:56management is something you can't do
1:01:58without as a trader. So being able to
1:02:01enhance your your uh risk management
1:02:03with the AI is a blessing. But what can
1:02:07an AI do? Well, it can help you with
1:02:09things like for example position sizing
1:02:12math. So it can instantly calculate, you
1:02:14know, how much you should buy, uh you
1:02:17know, how much you should sell and so on
1:02:19and so on. Um it can also help you, for
1:02:22example, build a complete uh trading
1:02:24plan, you know, where to set your stop
1:02:26loss and everything like that. We talked
1:02:28a little bit about it earlier in the
1:02:30course.
1:02:31It can help you come come up with
1:02:33checklists. This is actually a very
1:02:36powerful. It might sound a little bit
1:02:37old school but having checklist is an
1:02:41underrated thing not only in trading but
1:02:43you know just in general in my opinion.
1:02:44So it can help you create these uh
1:02:46pre-trading checklists as you can see
1:02:49right here. Things that has to be true
1:02:51before you enter the trade. And here, as
1:02:53I just mentioned, risk management keep
1:02:55you in the game long enough so your edge
1:02:59can actually compound and you can make
1:03:03money.
1:03:04All right, so now we are once again back
1:03:07in trading view. And to demonstrate what
1:03:09an AI can do when it comes to risk
1:03:11management, let's just uh for example
1:03:13open up an imaginary
1:03:16uh trade here. So let's say that for
1:03:18some reason we identified a trade right
1:03:20here. Let's say we set our stop loss
1:03:22just below the low right here. And let's
1:03:25say we use a risk-to-reward ratio of
1:03:27three. So, I'm just uh putting up a
1:03:30random trade right here. And now I want
1:03:33to see if uh if uh Claude can actually
1:03:37help us with let's say position sizing.
1:03:39All right. So, I'm just about to enter
1:03:41the trade. You can see on the screen
1:03:44right here. Uh I hope you can see it. uh
1:03:47the entry level, the stop loss, and the
1:03:50target level. Uh I currently have an
1:03:52account of $10,000.
1:03:55Um and I'm uh I want to risk 2% per
1:03:59trade. Can you tell me exactly how much
1:04:02Bitcoin I should buy. As I said, I want
1:04:06to risk 2% per trade. And as you can see
1:04:09on the chart, I have my target and stop
1:04:11loss right here. All right. So, I think
1:04:13that should be good enough. Let's see if
1:04:16Claude can tell us exactly how much to
1:04:19buy. Okay. Okay. So, we are done here.
1:04:22Uh it says I can see your long position
1:04:24tool. Uh let me pull the exact level.
1:04:26So, let's double check here. Entry
1:04:2963,926.
1:04:32Uh and as you can see, it's correct. The
1:04:35target 71,752.
1:04:38That is correct. And the stop loss is
1:04:40correct. So here the position size for
1:04:432% risk on $10,000
1:04:46is you know we want to risk 200 and risk
1:04:48per uh bitcoin is 2,69.
1:04:54And here it calculated the size to be
1:04:570.0767 0767
1:05:00or in other words around 4,00 $4,900
1:05:04worth of Bitcoin which is currently
1:05:06around 49% of our $10,000 account. So we
1:05:11don't need to use any leverage. We can
1:05:13just use our own equity. And here it's
1:05:16actually smart enough to notice here
1:05:18that our position sizing tool is
1:05:20actually set to our 1,00 account. Uh so
1:05:23it actually asks the question you know
1:05:25do you want me to update update the tool
1:05:27settings automatically to change it to
1:05:29the to the correct size that is you know
1:05:31an amazing an amazing thing. Uh but for
1:05:34now let's not do it because let's
1:05:36continue here with the course
1:05:39but right so now let's switch attention
1:05:41to something very important and that is
1:05:44using AI safely. What are some best
1:05:48practices to to use AI in a safe manner?
1:05:52And the first very important thing is
1:05:55that we are at least in my opinion some
1:05:57will disagree but in my opinion we are
1:06:00still in a stage where you need to
1:06:03verify. So you can let Claude for
1:06:06example you know find support and
1:06:08resistance levels but I think it's still
1:06:10always important to verify that these
1:06:13levels are actually you know good. So if
1:06:16you watch this video maybe you know five
1:06:18years from now or 10 years from now
1:06:20maybe by then you know AI is so good
1:06:22that humans don't even need to to verify
1:06:24but we are definitely still in a stage
1:06:27where you know uh an AI is used as an
1:06:31assistant. You might have heard the
1:06:33phrase that you you you your job will
1:06:35not be replaced, you know, by an AI, but
1:06:38it will be replaced by an by a human
1:06:40that uses AI better than you. And
1:06:44trading is about to enter that stage as
1:06:46well. The best traders will be the
1:06:48traders that are not only good themsel,
1:06:51but they also have a good AI assistance.
1:06:54I think that is where trading is going.
1:06:57Um, but here Claude's live prices, but
1:07:02not really the news behind them. And I'm
1:07:05a bit hesitant if I should even write
1:07:07this because, you know, of course,
1:07:09Claude can nowadays also find the news.
1:07:13But if you only use the trading view uh
1:07:16and claude connection and don't tell
1:07:18Claude to you know uh, you know, go
1:07:21deeper and find out news and so on and
1:07:23so on, it will not do it automatically.
1:07:25So here once again this is kind of like
1:07:27a skill issue. If you are good at using
1:07:30AI you know Claude can see you know both
1:07:32live prices and news and everything like
1:07:34that but you need to still be cautious
1:07:37and it's up to you to you know prompt
1:07:40claude in good ways and use cloud in
1:07:42good ways and so on and so on. Also here
1:07:44keeping on the same theme AI is very
1:07:48good right now at you know preparing
1:07:51back testing trading strategies uh even
1:07:54doing analysis as as you have seen in
1:07:56this course it can do you know basic
1:07:58support and resistance it can it can do
1:07:59market structure but it's still very
1:08:01important you are responsible for
1:08:04everything you are the final sort of
1:08:07boss that decides yeah okay here is
1:08:10claude's analysis and claude's back test
1:08:12but you are the one deciding if you
1:08:14should click buy or if you should click
1:08:16sell, you are fully fully responsible.
1:08:21And by adopting these sort of safe
1:08:24practices, you can hopefully get the
1:08:26full upside of AI without, you know,
1:08:29these careless mistakes that can
1:08:31definitely happen if you are not
1:08:34careful.
1:08:36But all right guys, this video this
1:08:39course is getting pretty long. So I
1:08:41think it's time to actually wrap up. Uh
1:08:44first of all, I want to thank you all so
1:08:46much for watching. Uh I know this has
1:08:49been a little bit of a different course,
1:08:50but trading is moving more and more
1:08:53towards AI. So I thought it was
1:08:55definitely time to, you know, start
1:08:57incorporating AI more into the channel.
1:09:00And if you like this course and if you
1:09:02want to see more of it, please let me
1:09:03know that down in the comment section
1:09:05below. Also, another thing, if you made
1:09:08it all the way, if you watched the full
1:09:10course all the way to the end, make sure
1:09:12to drop down below in the comments,
1:09:14write full course. Uh that has been a
1:09:17thing I have been doing for a while and
1:09:18it's super, you know, fun to see so many
1:09:20people confused about it. People are
1:09:22thinking, you know, why is everyone
1:09:24writing a full course? Uh and it's also
1:09:26a way for me to, you know, see that you
1:09:28watch the full course. So, it's always
1:09:30nice to see these comments. So, make
1:09:32sure to comment that down below. And
1:09:34also, if you have any questions, of
1:09:36course, let me know. Another thing, this
1:09:38one is so underrated, but if you did
1:09:40enjoy this course, dropping a like on it
1:09:43helps more than you think. So, if you
1:09:45have a spare a spare like in your
1:09:47pocket, I think it's free. Uh, at least
1:09:50right now it's free. So, if you want to
1:09:52drop a like, do that. And I will see you
1:09:55in the next video. But, all right, guys.
1:09:58So being able to use AI is super
1:10:01important, but the other piece of the
1:10:03puzzle is of course the trader that is
1:10:06using the AI. So, it's very important to
1:10:10improve your core trading knowledge, and
1:10:12you can do that with this free playlist
1:10:15right

About this transcript

This page contains the full transcript of MASTER AI Trading in 70 Minutes (Full Claude + TradingView Course) by Mind Math Money, generated from the public captions YouTube serves with the video. The transcript has 12,171 words across 1,638 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.