YT YouTube2Text

YouTube2Text → Transcripts → I Backtested 1,500 Versions of the TTrades Fractal Model Across 16 Years (Full Results)

I Backtested 1,500 Versions of the TTrades Fractal Model Across 16 Years (Full Results) — Transcript

by Strange ICT · 2,716 words · 416 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. In the last video, I took the
0:03candle range theory, one of the most
0:05hyped strategies on the internet, and I
0:07ran 4200 variation of it through my
0:11backtesting engine. And the verdict, no
0:14edge, none. It doesn't work. 90% of the
0:18variations failed and came back
0:20unprofitable, and the other 10% that did
0:23work all died the moment they hit data
0:26that they have never seen. And of
0:28course, I got hate for it, most of it
0:30from traders who aren't profitable in
0:32the first place. So, let me say this
0:34once clearly. Whether you believe my
0:37backtest or not is entirely your choice.
0:40I'm not telling you to stop trading it.
0:42I'm not telling you to follow me. I'm
0:44not telling you what to do. I'm telling
0:46you what the data said. Do whatever you
0:48want with that, all right? Just remember
0:51it the day you hang all your hopes and
0:53dreams on that particular strategy.
0:56Now, the moment that video went up, my
0:58comments and my DMs filled with one
1:00request, do the T-trades fractal model
1:04next. So, here we are.
1:06Now, if you have spent any time around
1:08ICT mentors and educators, you already
1:11know the game. You take the setup
1:13exactly how they teach it, the sweep,
1:15the entry, the stop, the targets, all of
1:17that. And once it loses, they tell you
1:20that you didn't read the right context,
1:22your narrative was wrong, it was the
1:24wrong time of day.
1:26You just don't get it. So, it's never
1:27the model, it's always you.
1:30>> As if you're right on your bias, that
1:32means you got to know the direction.
1:33Otherwise, these things are going to
1:35fail in your hands. In the beginning,
1:36when you don't know what the bias is
1:37consistently, it's going to fail in your
1:38hands. It does not mean my concepts
1:40fail. It doesn't mean that the logic is,
1:42you know, inferior. It doesn't mean that
1:43they change the algorithm. It just means
1:45that you yourself, as the operator,
1:47don't know what you're doing yet.
1:48>> But here's what makes this model and
1:50this video different, and I love the
1:53author of this strategy because of it.
1:55These are his own rules, not mine. The
1:58author says right there in his videos
2:01that his rules are all defined.
2:04He titles his own lessons a mechanical
2:06framework, and he sells an indicator $30
2:09a month that draws the exact setup on
2:12your chart for you.
2:13So, think about it. If a computer can
2:15draw it in the form of an indicator,
2:18then the computer can back test it.
2:20There is no context or narrative left to
2:22hide behind.
2:24And every claim in this investigation is
2:26cited in his own words, linked to his
2:29own videos, and his website, so you can
2:31check by yourself all of it, backed by
2:34his own proof. So, in this video, I
2:36formalized every rule he teaches, and I
2:38ran over 1,500 variation of it on each
2:41five markets, the gold, the Dow Jones,
2:43the Nasdaq, EUR/USD, and USD/JPY across
2:4816 years of data. So, today we're going
2:50to find out if the TTrades Fractal Model
2:53actually work. So, stay till the end.
2:56Before we get into the video, make sure
2:57to check the current discount Bullox is
2:59running. One of the best and most
3:01reliable futures pro firms out there.
3:03They have three different account types.
3:05You can go to their website and read
3:07everything that you need. They have the
3:09lowest payout rejection rate in the
3:11whole industry. So, you can use the
3:13coupon code stranger to get a really
3:15nice discount on whatever account that
3:17you choose. And now, let's continue with
3:20the video. Now, everything I'm about to
3:22show you, I put it into one a big
3:24report, a PDF with every rule, every
3:27chart, every number. If you want to read
3:29it in details, or you want to try your
3:31own variation maybe, just leave a
3:33comment down below, and I'll send it to
3:35you. And same deal as always, if there
3:37is another strategy that you want me to
3:39put under the scope and run through the
3:41exact engine, drop it in the comments,
3:44or DM in my Twitter account, and I'll
3:46run the full thing, and show you the
3:47results, including the exact variation
3:50that made money, if there is a one. All
3:53right, so here's how we tested it. The
3:55data goes back all the way to 2010 up to
3:572026
3:59and we split it three ways. So, the
4:00first part is the in sample. This is the
4:03playground where we explore and build
4:06variations. The second part is the
4:08validation. All right, so the survivors
4:10have to prove themselves here. And third
4:13part is the out of sample sealed until
4:15the end. So, the survivors get only one
4:19shot at it.
4:20And the rules as always every trade risk
4:22the same unit one R. So, everything is
4:25weighted on a level field and it's going
4:28to be simplistic to read the data and
4:31every trade includes a spread and the
4:33real cost you'll actually face in the
4:36market.
4:37Because a lot of strategies looks
4:38profitable without costs and the moment
4:41they had real spreads, they all
4:43collapse. All right?
4:44So, what is the T Trades Fractal Model?
4:48Now, at its core, it's one simple story.
4:52The same one a lot of these ICT models
4:54tell. The market is full of stop loss
4:56orders sitting just above the obvious
4:59high and below the obvious low. Big
5:01players push price past a high to grab
5:03those stops and fuel their own
5:05positions.
5:06And once they're filled, price snaps
5:09back and runs the other way.
5:11So, it's basically a model for catching
5:13those grab and reverses moves.
5:16So, T Trades breaks every move into four
5:19candles. He calls them C1, C2, C3, and
5:22C4. The four stages of one single move.
5:26C1 is a setup candle. I'm going to go
5:28through the model how it works briefly
5:30and then I'm going to go through the
5:32results of the back test. Just so you
5:34have the right [snorts] context.
5:36So, C1 is a setup candle, the one before
5:39that prints the one before that prints
5:42the high or the low that that's about to
5:44get swept. Right? C2 is a star of the
5:47whole model. C2 is the candle that
5:49sweeps C1's high and closes back inside
5:52C1's range. He calls that candle two
5:55closure, a reversal closure, all right?
5:58So, that close back inside is his signal
6:01that the reversal is on and he calls it
6:04the swing point. C3 is the continuation,
6:06the next candle where the trade actually
6:08run and takes profit. C4 is the
6:11expansion and
6:13the move stretching further out, all
6:15right?
6:17And this works on two time frames at
6:18once. So, the higher time frame sets a
6:20trap, the lower time frames times the
6:21entry using the change in the state of
6:23delivery. So, he spun the pattern on the
6:26higher 4-hour and you take the entry
6:28down on the 15-minute with that model.
6:31And of course, there is daily bias,
6:32external and internal liquidity,
6:34discount and premium, point of interest,
6:36all of that. Every piece has been
6:38formalized from his own videos. So,
6:40don't worry about it.
6:43Now, this section is for the state of
6:45delivery. I made a section for it in
6:47details. If you want to go through it,
6:49be my guest. But everything is simple
6:52and
6:53well-defined. So, I give him props for
6:55that. There is no mystery here that you
6:58need to buy a course to find out. Now,
7:00the variations that we are on. First of
7:02all, before variating anything, we are
7:04on the baseline. The baseline is his
7:07exact settings straight from his
7:09playbook.
7:10So, the daily closure bias from this
7:12video, the hourly reversal at a fair
7:14value gap or a swept level, the correct
7:17half of the day's range, premium and
7:19discount, confirmed by a full change in
7:21the state of delivery, internal the
7:23close, stop at the swept high, a fixed
7:27to our target. So, that's the baseline.
7:29That's his exact recommended setup, all
7:32right? And then we are on the full grid.
7:34So, every reasonable version of every
7:37rule and every combination of them.
7:39So, the target, the stop, all of it.
7:41That give us around 1,500 versions per
7:44market. And And now to the results.
7:47Honestly, guys, I was generally excited
7:50about them. As you can see on this
7:52diagram right here, everything leaned to
7:54the profitable side.
7:56So, this is the profit factor on the
7:58x-axis, all right?
8:00And remember, profit factor is just for
8:03every dollar that you lose, how much do
8:05you make? So, above one, you're
8:07profitable. Below one, you are losing.
8:09And as you can see, more than 50% of the
8:12variations landed on the profitable
8:15side.
8:16That is the signature of a real edge.
8:18Unlike the CRT, the candle range daily
8:21we tested last time, go back to the
8:22previous video
8:24and compare the two diagrams side by
8:26side.
8:27Let me just pull this out so you can
8:29see.
8:30So, there you go.
8:31Big difference between the two.
8:33Really, really promising results. And
8:36then everything collapsed. So, here's
8:39what I did. I took the champion of these
8:41versions,
8:42the best one for each market. And here's
8:44the thing, it was actually the same
8:46variation that scored the best metrics
8:48on gold
8:49and the same variation that was a
8:51champion on the other assets, too.
8:54That's an even bigger green flag. So,
8:57it's not like
8:59there was a variation that
9:00had big metrics on gold and there was
9:03another different variations that
9:05stood out on the Dow Jones and another
9:07variation on USD. No, the same variation
9:11scored the best metrics on all assets.
9:14All right?
9:15So, I turned it into an MT5 expert
9:18advisor, a real robot. It was a simple
9:21task because my engine runs on Python,
9:23so we just translated to MQL5 and then
9:26everything collapsed on MT5 strategy
9:30tester. When I ran it on gold, I was
9:32expecting a a profit factor around 1.5.
9:36But when the results came back, it was
9:381.05.
9:40Barely above break even. Basically break
9:43even.
9:44So, I started investigating. Maybe the
9:46MT5 code wasn't the same as the Python.
9:49All right? Maybe I hadn't translated the
9:51strategy correctly.
9:53So, I ran every test and finally I found
9:55a problem and it was really stupid
9:57problem. I should have
9:59stopped that from the start.
10:01It was the look-ahead bias.
10:04So, the look-ahead bias means that the
10:06back test used information that in real
10:09life would not exist yet at the moment
10:11it made the decision.
10:13Basically, the strategy in my
10:15backtesting engine was quietly reading
10:18the future to act in the present. And
10:21this is one of the most common traps in
10:23all of algo trading.
10:26Where your strategy sneaks a look at the
10:28future and then makes its decision.
10:31But, when you run it on a platform, a
10:33real one like NinjaTrader, MT5, or even
10:36TradingView, so that look-ahead bias
10:39gets exposed because their engine can
10:41only move forward bar by bar.
10:46And this is exactly what happened. The
10:48strategy got exposed. So, after I fixed
10:50that, every single variation collapsed.
10:52Every metric collapsed as you can see on
10:54these charts.
10:56The profit factor on gold came out to
10:581.05, exactly as the MT5.
11:02So, now
11:03our Python engine and the MT5 uh
11:05strategy tester is the same. We have the
11:07same metrics with a 27%
11:10win rate.
11:11And all of the other currencies
11:12collapsed, too.
11:13And honestly, guys, I was really, really
11:15disappointed because I had real hope for
11:17this one, especially with all the
11:19teaching behind it.
11:21The well-defined conditions and rules
11:23and the logic underneath it.
11:25But,
11:26the data doesn't lie. Simple as that.
11:29It's right in front of you. If you want
11:30to read the whole report, be my guest.
11:33Leave a comment and I'll send it to you.
11:35I'm not going to waste any more time on
11:36it and I can't give it another chance
11:38because I already spent a huge amount of
11:41time formalizing every single thing from
11:44all of his videos and
11:47double-checking everything,
11:48triple-checking everything. There is
11:50nothing left to save
11:52the strategy.
11:54One final thing I want to show you
11:55because I get DMs and I see posts all
11:58over Twitter about this exact thing. So,
12:01imagine this scenario.
12:03Someone tests a strategy from a guru,
12:05maybe he learned on YouTube or bought a
12:07course. So, the first month, maybe the
12:09first two, it's not working for them
12:12and they assume it's their own fault.
12:14Maybe they are not executing the
12:15strategy the right way.
12:17Maybe they're not being patient enough.
12:19They're not reading [snorts] the right
12:20contacts. So, they still believe the
12:22strategy works, but they just need more
12:25work on themselves. And then after a
12:27month or two, the strategy starts making
12:29money, everything is good. Maybe you get
12:31funded, maybe you get a payout. You have
12:33finally cracked it.
12:35Your fear, your discipline, your
12:36psychology. You tell yourself that you
12:38have sorted it all out. You finally know
12:41how to run the strategy. And then a few
12:44weeks later, a month or two later, the
12:45strategy starts losing again. It was
12:48fine before, so what exactly happened?
12:51This is a huge problem and a common one.
12:54I have seen it hundreds of times in my
12:56journey and let me show you something,
12:58all right?
12:59This is a chart of a strategy that
13:01doesn't work. As you can see, this
13:03strategy is a total failure. It loses
13:05money from start to finish across six
13:07years of data.
13:09But let's zoom in here for a second, all
13:11right? Look at the equity curve right
13:13here from this point.
13:16Around Feb or March
13:18all the way to the end of 2023, the
13:20strategy was actually profitable. So, if
13:23you would have started learning uh this
13:25strategy and applying and trading with
13:27it in Mars
13:29maybe from someone on YouTube maybe from
13:32I don't know course you bought you would
13:33be living your best life right here. The
13:35strategy is working.
13:37You are making money. You got your
13:39payouts but all you really had was a
13:41regime that happened to favor the
13:43strategy and when the regime shifted you
13:46can see it the strategy collapse and
13:48just kept bleeding to the downside. A
13:50consistent profitable strategy guys
13:52survives across years of data. Six or
13:55eight years minimum green every single
13:58year. All right, that's what consistency
14:01looks like. Getting one payout or two
14:03payouts or even three payouts out of a
14:06strategy is not consistency. That's not
14:08what makes a strategy work. What makes a
14:10strategy work is consistency over years.
14:13So keep that in mind whenever you test
14:15something on your live account.
14:17Because you might be trading something
14:20that is a timed bomb. It's going to stop
14:22in the future and you're going to blow
14:24out all your accounts.
14:26So the verdict does the T trades fractal
14:28model have a real mechanical edge?
14:31One that you can actually trade based on
14:34every back test we're on the answer is
14:36simply no. It doesn't work. And I want
14:39to be fair
14:41with the author of the strategy
14:44this is important. The model doesn't
14:46fail because the logic is stupid. All
14:48right, no. The logic is excellent but in
14:51trading everyone can build his own
14:54logic. The Elliott Wave believes waves
14:57are how the market moves. That's his
14:59logic.
15:00And he believes it to the bone. Another
15:03trader trades trend lines and he
15:05believes trend lines are the sole truth
15:08of the market. Another uses ICT concept.
15:10Another uses supply and demand. So
15:13everyone here has his own version of
15:16logic and just because it sounds logical
15:19and your brain can digest it it doesn't
15:21mean that it works. All right?
15:25So, that's the three trades fractal
15:27model. If you got any other strategy
15:29that you want me to test, drop it in the
15:30comments and we will let the data
15:33decide. Same as always, happy trading
15:36and good luck.

About this transcript

This page contains the full transcript of I Backtested 1,500 Versions of the TTrades Fractal Model Across 16 Years (Full Results) by Strange ICT, generated from the public captions YouTube serves with the video. The transcript has 2,716 words across 416 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.