YT YouTube2Text

YouTube2Text → Transcripts → Building a Profitable Strategy w/ Claude - Day 90: My 7 Winning Strategies

Building a Profitable Strategy w/ Claude - Day 90: My 7 Winning Strategies — Transcript

by Strange ICT · 2,928 words · 405 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Hello guys and welcome back to the
0:01channel. If you are new here, here's
0:03where we are. In the last episode, I
0:05told you that I stopped chasing the one
0:07perfect strategy, the unicorn that works
0:09in every market and every regime because
0:12honestly, that thing is almost
0:13impossible to find. And instead, I
0:15started building a book, a team of good
0:18modest strategies that don't move
0:20together. When one of them struggles,
0:22another one carries it. And the idea is
0:25simple. a team of modest strategies can
0:28beat one great strategy. So for the last
0:31few weeks, that's exactly what I did.
0:33And today I'm revealing my winning
0:35strategies. As you can see, this is my
0:37research terminal. Everything in this
0:39video lives here. Every strategy, every
0:42result, and the official reports for
0:45each one, seven strategies, five
0:47different families, three markets, and
0:499,500
0:51trades. Now, before we start, I got a
0:53present for you. I'm giving you the
0:55exact historical data that I used to
0:57build these strategies. Almost 15 years
1:01of data for each pair, clean and ready
1:03to back test all in one folder. All you
1:06have to do is to subscribe, like the
1:08video, and leave a comment down below,
1:10and I'll send you the link. Now, before
1:12we start, make sure to check the current
1:14discount bonus is running. They offer
1:16three options right now. the classic
1:17evaluation, a onetime fee evaluation
1:20with no activation fee after passing and
1:22a straight to funded option. You can use
1:25the coupon code stranger for the best
1:27available discount. All right, so when I
1:30started this series, I was only trading
1:32futures, okay? I was trying to find
1:34strategies on NQ and only intraday.
1:38That's where we found the opening range
1:39breakout, our first successful strategy
1:42in this series, the AM momentum. Um, and
1:46I loved it. Futures are great, really
1:48tight spreads and clean executions.
1:50Great for intraday sculping and so on.
1:53But the moment I shifted my attention to
1:55the book, I hit a wall. A book needs
1:58variety. Different markets, different
2:01behaviors, strategies that don't move
2:03together. And with one market and the
2:06fact that I can't uh hold overnight,
2:08there is only so much you can find,
2:10especially when you need um seven or 10
2:15uh strategies with low correlation
2:16between them. So I opened the door to
2:19CFDs, gold, the Japanese yen, the
2:22NASDAQ, forex pairs. Um I could hold
2:26trades overnight and I could size every
2:28trade to the exact risk that I wanted uh
2:30using the right lot size. And that's
2:33where things started to click. Now, as
2:35you can see here, every strategy in the
2:37book belongs to a family. I have five of
2:39them. Front runners, breakouts, session
2:42breakouts, opening momentum, and mean
2:45reversion. Now, one thing to get out of
2:47the way, I will show you the logic of
2:49each strategy, how it enters, and how it
2:52exits, and the results, but I'm not
2:54sharing the exact settings, and it's not
2:56me being protective or greedy. But pro
2:59firms detect identical trades uh coming
3:02from different accounts. They have
3:04really sophisticated systems for that
3:06and they can ban everyone involved. Now
3:09the first family is the simplest idea in
3:11the book. All you have to do is to wait
3:13for price to break out and only then
3:16jump in. It's um the late entry, okay,
3:19the boring entry, the one that you can't
3:21really uh brag about on social media or
3:24whatever. and it has one of the most
3:26solid edges that I found as you can see
3:28on the metrics. And this right here is
3:30the gold breakout strategy. It measures
3:33how wide gold normally moves and it only
3:36buys when price escapes that band. It
3:39means that it enters when there is real
3:42volatility on the market calculated
3:45based on how price usually moves on
3:48gold, you know, and it runs on three
3:50time frames at the same time. So I
3:53tested on three time frames and when I
3:55run the correlation it turned out that
3:57those three time frames although the
3:59same strategy but it actually has a
4:01really low correlation between them. So
4:03I grouped them into one strategy that
4:05trades on different time frames. So same
4:07idea but three speeds and it only wins
4:1130% of its trades as you can see but
4:13when it wins it wins three times what it
4:16loses. So many small loes and a few huge
4:19winners. It made plus 220% over almost
4:22six years on 1% risk per trade. But I'm
4:26not going to hide this. This strategy is
4:28ugly. It has a 47% draw down. But
4:32remember that draw down because when we
4:34put it into the book, everything is
4:36going to change. Now the same engine
4:38that we use, we I tested it on the
4:40dollar against the Japanese yen and it
4:42worked there also. So this is a good
4:44confirmation that this edge is robust
4:47when you take it into a different pair
4:49and it works. So this one takes about 40
4:52trades a year and 4 and a half% draw
4:55down really good draw down and it has
4:58the cleanest curve of the whole family
5:00and that's when you know an idea is real
5:03again when you take it to another
5:04markets and it works. Now the next one
5:06is probably the oldest idea in forex.
5:09Asia tends to build a range and London
5:12breaks it. If you watch the previous
5:14video, you already know this idea. So
5:16the strategy basically records the range
5:18of the Asian session and when London
5:21opens, a buy order waits just above that
5:24range. If it breaks, we are in and we
5:28are out before the New York afternoon
5:29and it never holds a trade overnight. It
5:32has couple of tweaks to it, but I think
5:34this is enough for you to build a
5:36strategy based on this. it made uh plus
5:39108%
5:41uh over almost 6 years between in sample
5:44and out of sample. Now uh something you
5:47might have noticed uh some of these
5:49ideas didn't really come from me. The
5:52London breakout is a public idea that's
5:54been around forever and my best strategy
5:56the one that I'm saving for last started
5:58with uh something that I read on Twitter
6:02couple of months ago and this is the
6:04thing about AI. Claude is insanely
6:07powerful at back testing ideas, building
6:09them, breaking them, finding the
6:11problems and making them better. But if
6:14you ask cloth to come up with a strategy
6:16from its own understanding of the
6:17market, you will get genetic stuff. The
6:20best ideas come from real traders,
6:22someone who has been watching one market
6:25for 10 or 15 or 20 years. and sometimes
6:29maybe drops a small observation in a
6:32social media post, a reply or in a video
6:35or in a forum. You know, that's where
6:38you can come up with really really
6:39powerful ideas and build a system around
6:41it. So, I'm building a system that goes
6:44out and finds some it reads Twitter,
6:46Reddit, and decades old forum threads
6:49the way um an investigator would
6:52following the leads, the replies and
6:54even charts in the images. then it hands
6:58me the best pieces and they can become
7:00strategies that we can actually back
7:02test and deploy in live markets. It's
7:04still under construction and in next
7:06videos I'll show you a live example how
7:09it works. So make sure to subscribe. So
7:11back to my roots the New York opening
7:13momentum. Um this one is a variation of
7:17the opening range breakout. the first
7:19profitable strategy that we found in
7:21this journey and this only trades uh the
7:24first hour after New York opens and only
7:26when momentum is already extreme in one
7:30direction. Okay, so it is a momentum
7:32strategy. It follows the move and it
7:35doesn't fight it or try to fade it and
7:37it skips the moves that have already
7:39stretched too far and every position is
7:42flat by the close. Now if you look at
7:44the curve, it has a profit factor of
7:462.88.
7:48This is the highest in the whole book
7:50and a draw down of only three and a
7:52half%. About 37 trades a year. So the
7:55frequency is not um extremely high.
7:58Doesn't trade much but when it does it
8:01actually means it. And honestly this one
8:03is really powerful. Um I have given you
8:06the seed of the ideal extreme momentum
8:08at the open. You want to follow that. Um
8:11if you spend some time uh in the lab
8:13with it, you can definitely build a
8:15really nice strategy around it. Now, uh
8:18before I show you the last two
8:19strategies in the book, I need to show
8:21you something else, which is the ones
8:23that didn't make it. Uh the book was
8:25about 15 strategies. Those passed every
8:28single test in Python on seeing data,
8:30cost, stress, placebo test, and
8:32everything. And then I rebuilt each one
8:35as an MT5 expert advisor and run it
8:38through a real broker tester. And only
8:41seven survived. Some of them died from
8:44the swabs, the fee that you pay for
8:46holding a trade overnight. So, for
8:48example, the NASDAQ breakout made around
8:50$11,800
8:53when trading and paid $9,500 in
8:56overnight fees. That is insane. The edge
8:59was there, but the swap ate almost all
9:02of it. Um, this one, uh, the DAO
9:05breakout also looked great in Python. on
9:08the broker's tester a profit factor went
9:11all the way to 0.7. I didn't so I didn't
9:14model the swaps the right way in Python.
9:16That's something a lesson that I learned
9:18the hard way. Uh some of them died
9:20because of the price feed. So for
9:22example in forex every broker has its
9:24own price. So for example on this one
9:26two brokers agreed on the signals only
9:2861% of the time. So the edge was thin
9:32that it lived in one broker's prices and
9:35nowhere else. Also, some of them died
9:37from real ticks. So, there is a big
9:39difference between running a back test
9:41on one minute open, high, low, and close
9:44and real ticks. Some strategies really
9:46doesn't care, but some of them actually
9:48sensitive to this. Always volatility
9:50strategy on real ticks. All right, so
9:53two left in the book, and this one is
9:55the only mean reversion strategy in the
9:57book that I have, and it trades on the
10:00daily chart. So basically when the
10:02NASDAQ has a sharp sell off, you know,
10:05it buys the next open, it can hold for a
10:08few days and get out no matter what. Uh
10:12it's rare. It doesn't trade every single
10:14day, of course. About 20 trades a year
10:17and it's the smallest earner in the
10:19book. So why do I keep it? Well, because
10:22it behaves nothing like the others. when
10:24the breakouts are struggling. This one
10:27is usually doing its own thing, you
10:29know, and in a book that's exactly what
10:31you want. Strategies with different
10:33behaviors and that don't move together.
10:37And now to the last two strategy to the
10:40last family and it contains my two best
10:44strategies. So I want to show you the
10:45exact thought that led to this strategy.
10:48So let's say that we are in a downtrend
10:50and there's an obvious swing high on the
10:52chart from a previous pullback. There
10:55are basically four things that can
10:56happen there. One, price pulls back,
10:59takes that high, and shows no reversal
11:02sign and just keep going. I don't trade
11:04that. Two, price takes the high, shows a
11:08reversal sign, and I take the trade and
11:10it works. Three, price takes the high,
11:13shows a reversal sign, I take the trade,
11:15and the sign was fake. price keeps going
11:18and I lose. And for the one that changed
11:21a lot, which is most of the time that
11:24high actually gets taken. So I asked
11:27myself, why am I waiting for it to get
11:29taken and then fighting it? If it gets
11:32taken most of the time, why not write
11:35the move into it? So that's how the
11:38frontr runner uh family works. It was
11:40inspired by a Twitter thread that I read
11:43a couple of months ago um about where
11:47people tend to trade the levels that
11:50people are waiting for to actually get
11:52involved either in a a sweep and
11:55reversal or traders who want to trade
11:58the breakout. So I was thinking why
12:00don't just trade and take a buy or sell
12:02to that exact level. So basically it
12:05places its order just before the level
12:07not on the break. Then the breakout
12:10crowd shows up and they push price
12:12straight into our trade, you know, on
12:14gold. These are the metrics and this
12:16right here is official report straight
12:18from the MT5 uh strategy tester on IC
12:22market. Um, and this is the second
12:24strategy in the family. Um, basically
12:27the same thing but applied on the US
12:29dollar against Japanese yen. And these
12:31are its profits and the MT5 uh [snorts]
12:34report. Now, this is my favorite thing
12:38about uh laying everything out like this
12:40in this terminal. You get to see what's
12:43missing. So, for example, these are the
12:46families systematic trader can run.
12:48Trend following, carry, seasonality,
12:52pairs, news, and more. And I have five
12:55out of [snorts] possible 14 or even
12:57more. And I trade only three markets. US
13:01dollar against Japanese yen, gold, and
13:03US 100. And also only three of my seven
13:07strategies can actually sell and trade
13:09both directions. The rest are long only.
13:13So if everything falls at the same time,
13:16most of the book is on the wrong side. I
13:19do have one mean reversion strategy, but
13:22it barely trades and too much of the
13:24book is on the yen and gold. So
13:27obviously the book is not finished. It's
13:29a version one. Yes, it can be deployed
13:32live and I tend to do that this week or
13:35next week. And this right here is the
13:37risk ladder. A conservative, low,
13:39moderate, medium, high, and very high.
13:43The more risk you take, the more it
13:45makes. And the deeper the draw down is
13:47at medium, for example, 1% per trade. Um
13:50the $10,000 made about $60,000 in
13:54profit. That's plus 600%
13:57um with a 6% maximum draw down. But of
14:00course we did everything we could to
14:02make them honest and the results from
14:05here the data was split into three
14:08parts. One to build on, one to validate
14:10on unseen and one kept completely unseen
14:14until the very end. So every strategy
14:17went through plus ebo test look ahead
14:18checks um cost stressing um and more. I
14:23do expect this book to have a positive
14:25edge like I said, but it would be naive
14:27to pretend it will make exactly these
14:30numbers live. The real world will be
14:32rougher than any back test. And remember
14:35that gold breakout strategy that had 47%
14:38draw down inside the book when you risk
14:40it um the right way, you can barely see
14:43it. You know, that's what the team does.
14:46Something else that you should uh notice
14:49right here is everything here is fixed.
14:53Every trade is sized of the original
14:55$10,000. So 1% per trade is always going
14:59to be 1% per trade no matter what equity
15:01is and the profits are never reinvested.
15:05So that's how P firms account really
15:07works. um you get one or 2% or 3% and
15:11then you request a payout and then you
15:14will go back to the original balance and
15:16so on. So there is no compounding. The
15:19account stays the same size. But watch
15:21what happens when you let the profit
15:24works. That's compounding. On the
15:28conservative level, the fixed account
15:30ends at about $25,000.
15:32With compounding, it goes to 44,000.
15:36And on the low level 40,000 fixed
15:39becomes almost 200,000 with compounding.
15:43So as you can see this these are the
15:45same strategies, same trades but the
15:48only difference is that the account is
15:50allowed to grow and nobody takes
15:53anything out of it. This is the power of
15:55compounding guys. So if you have your
15:57own account on a real broker, even a
16:00smaller one like $1,000, this is
16:03something to really think about. Uh a
16:05prop firm gives you a bigger account.
16:07Yes. But a personal account gives you
16:10compounding and a lot more freedom you
16:14know. Uh both have their place. But just
16:17know what you are choosing.
16:20Um now the next step is taking this book
16:22live. How and where and when? I'll tell
16:26you when I get there. And the book will
16:28keep growing more families, more
16:30markets, and more sies that can that can
16:32actually sell, not just long. Um, and
16:35the system I told you about is going to
16:37help me find them, you know. And if you
16:39want the data folder that I mentioned in
16:42the start of the video, all you have to
16:43do is to subscribe, like the video, and
16:46leave a comment down below, and I'll
16:48send you the link. And if you are new
16:51here, this book didn't start today. The
16:53series playlist is right here. It shows
16:55how we found the very first strategy and
16:58how all of this started. And this
17:00journey is still going. So, make sure to
17:02subscribe so you don't miss the next
17:04one. Now, happy trading and see you in
17:07the next

About this transcript

This page contains the full transcript of Building a Profitable Strategy w/ Claude - Day 90: My 7 Winning Strategies by Strange ICT, generated from the public captions YouTube serves with the video. The transcript has 2,928 words across 405 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.