Full transcript
0:00This is something I've never shown on
0:03the internet before. This is probably
0:06two years old. It is a manual back test
0:09I did on Google Sheets. And if you look
0:11at the bottom here, there are quite a
0:14few of these.
0:16Um, and I bring this up because what I'm
0:19about to show you is the back test I did
0:21with AI where we back tested
0:2510,000 plus strategies
0:28and 5 6 million trades. And that's only
0:32valuable if you have a point of
0:34reference. So this sheet in front of
0:36you, every day after school or every day
0:39after work, I would come home and I
0:41would open Trading View. I would do a
0:44split screen with ES and then Q. I would
0:47take my strategy. I would test it out.
0:49And each one of these [snorts] is
0:51actually annotated.
0:54So, all of these pictures are
0:56screenshots I took of Trading View
0:58because I annotated the annotated the
1:00charts. I took a screenshot. I threw it
1:02in. I took some notes on what I saw. I
1:07filled in my time of entry, my pair and
1:11the result of the trade. And I entered
1:14my little formulas so I could understand
1:16my gains and my win rates. And this is
1:19how I back tested. And this whole sheet
1:22probably took
1:2410 to 20 hours because there's a lot of
1:27charts here and every single chart is an
1:30annotated chart. I was obsessed with
1:34learning how to trade and learning what
1:37edges would work. And so I dedicated
1:40myself to the craft.
1:43And that same dedication
1:45carries into the opportunity we have in
1:48AI. And what I'm about to show you is a
1:535.6 million trade back test with 10,000
1:57strategies.
1:59And I'm going to show you the most
2:01profitable strategies
2:04of anything that's out there. The best
2:07strategies to get paid, the best to pass
2:09accounts, the best to get payouts, the
2:12highest winning, the most money-making,
2:14the highest win rates. There's
2:15strategies in here with 100% win rates.
2:18There's strategies in here with 17R
2:22as an average winning trade.
2:25And I'm going to dive deep into all of
2:27this and explain how we can actually
2:31apply this. And before I get too deep, I
2:36want to take a second
2:41and preface this video.
2:44If you don't know who I am, my name is
2:46Luke. I've been I'm known on the
2:49internet as Lil Fish. I've been day
2:51trading for over three years. I've been
2:52using AI for around four years. I've
2:55made over five figures in payouts. And
2:57this year, I started using AI to do AI
2:59assisted trading. And you'll see why
3:01that's important, how that plays with a
3:03lot of these strategies. So, that means
3:05for me, I wasn't watching the charts. I
3:07was getting notifications of valid trade
3:09setups, then using my human discretion
3:11to decide whether or not to place or
3:13skip the trade.
3:16That led me to my biggest payouts ever.
3:20And now I'm working towards fully
3:24automated trading. So Claude designs the
3:26strategy. Claude places the trades.
3:29Claude just this past week passed two
3:31funded accounts and placed a winning
3:34trade on a funded account for me. The
3:36objective here is to have AI do the
3:38entire thing and AI to be that magical
3:41money printer that we're all that we are
3:43all after. But before I kind of go into
3:47this, I really want to speak to you
3:48because I really want you to win. And if
3:51you're here to just be entertained,
3:52that's fine. You're going to see a lot
3:53of really cool stuff and a lot of
3:55interesting information that you can
3:56find useful. But information without
4:00implementation
4:01is just entertainment. And that's fine.
4:03If you're here for that, that's fine.
4:04But if you actually want to win, this is
4:07information that you should be taking
4:08notes on and applying to your edges. The
4:12only reason I am not is because I
4:14already have formulas in place that I
4:17believe are going to lead me to make
4:18more money.
4:20I'm testing these. They're in the
4:22process of working. They are leading me
4:23to pass accounts and make winning trades
4:25and getting me very close to payouts.
4:28But if you're nowhere around there, if
4:30you're totally a beginner, if you've
4:31never used AI to back test,
4:34take notes on these. These strategies
4:37are digital gold, if you will. This
4:39information, these edges of what AI was
4:42able to find in the weeds is really
4:44helpful when it comes to how you should
4:48approach algorithmic or even
4:50discretionary trading. What is the best
4:52approach out there?
4:56So, let's get into it because there's a
4:57lot of fascinating information here.
5:12So,
5:145.6
5:16million trades,
5:1810,000 strategies.
5:20This was done over seven years of data
5:23on CFD data. CFD because you can get CFD
5:27data for free. This includes futures,
5:30but just their CFD version.
5:34[clears throat] Let's get right into it.
5:36Let's let's not hesitate any longer.
5:40Here's all the information we're going
5:41to look through and then we're going to
5:43look into
5:45the best strategies out there, the
5:48highest winners, and the ones that you
5:51should use.
5:56To start off, here are the most
5:59fascinating things that I find that you
6:01can implement right now.
6:03Cost divided by stop. So, when you enter
6:07a trade with a high position and a small
6:09stop-loss, you are more likely to eat
6:12into your edge because the commissions
6:13are so high. And I get a lot of DMs of
6:15people saying, "Oh, like I have a
6:17profitable edge, but the commissions are
6:19eating away. your stop loss is too
6:21tight. And this edge, this graph here,
6:24represents that. The smaller your stop
6:26loss is, the more you're paying in
6:28commissions and the better your edge has
6:30to be in order to win. So, this is
6:33saying that a commission that is less
6:35than 3% of the stop cost is more likely
6:38to win than anything else.
6:41Which market?
6:43So, across all the assets tested, NASDAQ
6:47is a dramatic winner. And that makes
6:49sense because this is one of the most
6:51liquid if not probably I don't I haven't
6:53looked it up but this is probably the
6:55most liquid asset of any to trade.
6:59That means they're the highest volume
7:01and the most points to catch. So NASDAQ
7:04is going to be your winner. Which time
7:06frame?
7:07This is a bit of a killer because
7:10there's a lot of people who love the one
7:11minute or they love the 3 minute or they
7:13love the 302. When I was making most of
7:16my money, I never dropped below the 30
7:18minute. The 30 minute was my
7:21executionary time frame. And really, I
7:23favored the 60 minute. And this proves
7:25that system to be true.
7:28Depending on what time frame you're
7:30using to trade, the higher the time
7:32frame, the more winning strategies there
7:35are. And that coincides with lower
7:38frequency trading because everybody
7:40loves to I'm going to trade every single
7:43day. That is not the most profitable
7:46approach. It just isn't.
7:49You will make more money if you trade
7:51less on the higher time frame. That is
7:54something I have experienced. That's how
7:55I got my payouts. I was executing on the
7:5730-minut time frame and across 10,500
8:00strategies. This proves that to be true
8:02even more. So,
8:04which session?
8:07Across all the sessions, London has the
8:11worst performance, second to worst,
8:14second to Asia. And again, that makes
8:16sense going along with NASDAQ because
8:19these are high volume, high liquidity
8:22times of day. And the highest performing
8:25strategies, all were New York AM
8:28session. This is the market open. This
8:30is that liquidity sweep at 9:30 a.m.
8:32that we'd love to see. Hopefully, you're
8:34already seeing a bit of a pattern here,
8:35and that's going to play into some of
8:38our best strategies.
8:40The New York AM session is the best in
8:44terms of all of these strategies
8:45combined. And I'm going to get into all
8:47of these different categories because I
8:49did not go out and hand select 10,500
8:52different strategies.
8:54That would be ridiculous. How the stop
8:57loss is set. I found this one to be
9:00interesting because I personally for my
9:03biggest payouts was doing
9:05structure-based stop- losses. Wick low.
9:07That's where it always goes. And
9:11this suggests that to not be the best
9:15approach.
9:17Marginally,
9:19fixed stop losses. Fixed meaning no
9:22matter where I enter, I'm doing a 15
9:24point, 50 point, 100 point stop loss.
9:27nothing else.
9:30Nothing to do with the candles, nothing
9:32to do with the volume, nothing to do
9:34with the time. That is the highest
9:37performing strategy or approach. Second
9:41to this was a surprise to me. Time.
9:45No matter when you enter,
9:48you just let it ride.
9:52Very interesting. You let it ride for 80
9:55minutes straight or 80 candles straight
9:58and then you exit no matter what. I
10:01found that to be fascinating because
10:03that is something you never see because
10:06prop firms make it in a way where if you
10:08do that you are highly unlikely to be
10:11profitable because this strategy, this
10:14timebased exit is going to lead to
10:16dramatic winners and dramatic losers.
10:19And in the prop firm environment, you
10:21have consistency rules. You have daily
10:24loss limits that prevent those huge
10:26winners and huge losers from happening.
10:29And that appears to be one of the most
10:32profitable approaches.
10:35How you enter
10:38stop, confirm, market, limit, all are
10:41around the same with limit entries being
10:44the worst performing.
10:48These are the 50 different families of
10:50strategies. I asked for it to pick the
10:51top 50 retail strategies. And what
10:54should you immediately see from here?
10:56The groups mean cumulative R per
11:00variant. Every single one is negative.
11:04Every single one of these has a negative
11:06expectancy.
11:10This is the reason that when I made my
11:12money and I chose to build an AI
11:14assisted system, it was AI assisted and
11:17not fully automatic
11:20because I knew that from my experience,
11:23I have a discretionary edge that when I
11:26see a trade setup, I could say hm yes or
11:29hm no. And I might not always be able to
11:32put into words why I feel that way, but
11:34that decision that I make leads to the
11:37success I have with the model. Because
11:39the model I traded was, let me see if I
11:42can find it.
11:44Was this liquidity sweep and reclaim C2
11:51zero survivors
11:54with a median trade result of
11:56negative.175.
12:00This is not profitable. Like any of
12:02these, these are not profitable. And
12:04these are including the commissions,
12:06which is a crucial part of strategy. You
12:08can't disregard commissions. The people
12:10who disregard commissions and then go to
12:12the live markets are why my DMs are
12:14filled with my strategy is not
12:16profitable when I don't know why.
12:19But this is the reason that I built AI
12:21assisted. I get the notifications that I
12:23make the decision. I'm avoiding all
12:25emotional aspects. I'm shortcutting to
12:27the only part where I'm valuable, which
12:29is that decision.
12:31That being said, I am actively in the
12:33process of proving myself to be inferior
12:36to fully automatic systems built with
12:38Claude.
12:45This is what I find more interesting
12:49than anything on this dashboard, and it
12:51is how likely it is to pass a 50k eval.
12:55So, there's 10,000 strategies. 10. I
12:58like imagine that for a second. That
13:00screenshot that I showed, that Google
13:01sheet that I showed of all the different
13:03trades, that was one strategy.
13:06This did it 10,000 times. 10,000 times.
13:13And all of those strategies only.6%
13:2060 of 10,000 strategies
13:24have greater than a 50% chance of
13:28passing.
13:33You have a 99.4% 4% chance that if you
13:38pick a strategy that is entirely
13:39automated, it is less than 50% likely to
13:43pass an account.
13:46So, choose wisely and make a decision on
13:49whether or not you want to keep yourself
13:51in the loop or not. This is why I've
13:54built systems like discretionary
13:56trainers before because I want to get
13:59better at my ability to decipher what
14:02strategies are winners or what trades
14:04are winners and what trades are losers.
14:16Look at all of these negative
14:17strategies.
14:19This should alarm you if you trade
14:22strictly algorithmic and not in a
14:24business prop manner
14:27because you need to do your due
14:28diligence in using AI or testing other
14:31strategies
14:33or using your discretion which is what I
14:36did.
14:38Now for all the strategies
14:42we can see that a lot of these have a
14:44incredible amount of trades. 11,000
14:47trades for all of these,
14:50which is unreal. And I'm going to break
14:53down what's interesting to me here and
14:56how this plays into the proper game
14:58because some of these strategies may
15:00look freaking awesome. We're going to
15:02see that, especially when we look at
15:03average RR and some of these just
15:06freaking crush it. 31 RR as an average,
15:10that kills it. But it's a 7% win rate.
15:15You can't pass an eval with a 7% unless
15:18you do I don't know what's that's like
15:2213 tries,
15:2414 15 tries to pass one evaluation
15:29that has a single day pass and no
15:31consistency. So this strategy would
15:33actually suck
15:36and it only won one time. probably a
15:39huge win
15:43because the trade count matters.
15:48Net R per trade.
15:53Man, I don't even know where to start
15:54because there's just so much here that I
15:55want to I want to talk about and I want
15:56to look at because you can look at all
15:58of these at face value. And this is
16:00what's so important. I get DMs all the
16:02time of people saying like, "Hey man, I
16:04have a 60% win rate and I'm not
16:06profitable." 60% win rate alone is not
16:09anything to tell me. Like I cannot get
16:11anything from you telling me you have an
16:13epic win rate because look at all of
16:14these 100% win rates, 80% win rate
16:16strategies and none of them are
16:18profitable. You have an 84% win rate but
16:21you're losing 85 but you're losing all
16:26of these 80s that are red. 80%. That
16:29sounds great, but look at all these
16:30losers because average RR and net R per
16:34trade are crucially important. And let
16:36me really scare you here because look at
16:38how many of these are likely to pass an
16:40eval
16:440000.
16:46They're not even getting to 1%. Most of
16:48these 1.6% chance of passing an eval.
16:54Forget even getting a payout. You're not
16:55going to pass the account to begin with.
16:59Okay.
17:03Best trades,
17:06best strategies,
17:12total net R.
17:17These are what I think are some of the
17:18best strategies,
17:22but they're not going to work on prop
17:25firms. And you can probably already see
17:27that because properforms are stacked
17:29against you already, which trading is
17:31already stacked against you. This is
17:32even more so. So you might be able to
17:34make gains with these in the live
17:35market, but it's highly unlikely a
17:38proper market. And you can see that
17:39immediately from the win rates. Win
17:41rates under 50% are almost impossible to
17:44make money with in the proper
17:46environment because you have consistency
17:48rules.
17:50Consistency rules have only showed up in
17:52the past year or two. The consistency
17:53rule is, if you don't know what that is,
17:55it means if I make $1,000,
18:00I can't have, if I have a 50% consist
18:03consistency rule, I can't have made over
18:06500 in a single trade. I must have taken
18:09two trades to make up that total or
18:11three trades. And so when a prop firm
18:13has instant day funding but a 20%
18:16consistency rule, that means you must
18:18have had five $200
18:21winning days. And if you lose and you
18:24make more on one day, if you
18:26accidentally make 300 on one day, now
18:28you have to make 1,500 before you can
18:30take a payout. Consistency rules will
18:32eat you alive.
18:34And so win rates that are low and have
18:37big average wins like these, volume
18:40spike breakout is the most profitable of
18:43any strategy. 10,500 strategies. Volume
18:47spike breakout is going to give you the
18:49most money and you will not be able to
18:52pass an eval and you will not be able to
18:54get a payout. Look at these statistics.
18:560% 0%. Look at the P&L chart over time.
19:01Very profitable. It's clear you can't
19:04win with it in a prop firm because of
19:06this number being 14%. And this number
19:09being eight
19:11because you're making nine times the
19:13amount you risk 14% of the time. And
19:16that's awesome and that's profitable.
19:18But that is not going to comply with
19:21what's required of profs.
19:26So you can't use this
19:34Now, let's look at the ones most likely
19:36to pass an even.
19:40Notice how it's another volume spike
19:43breakout. What do we immediately see
19:45here? We see high win rate, lower
19:50risk-to-reward. Now, unfortunately, this
19:53is only 27 trades. So, this is pretty
19:56unreliable data. We need much higher
20:00trade quantities before we look to use
20:02that. But we can see here, I mean,
20:07you could set this up to give you a
20:09notification when this trade exists. And
20:11so the few times a year when this
20:13strategy exists, you should take this
20:16trade.
20:19You're going to get what 27 trades
20:21across
20:23seven years. It says 0.07 trades a week.
20:26So, what does that mean? Once every 20
20:29weeks, two to three times a year, you're
20:32going to get one of these trade setups,
20:34and you should take it then because it's
20:35probably going to win,
20:38but uh you might have to wait an entire
20:41year to get this trade. So, if you want
20:43this criteria, here it is. Volume,
20:45spike, breakout, the family settings,
20:47the directions. How here's your entry,
20:50your stop, and your target.
20:52Here's your session, your daily flat,
20:55your maximum amount of trades per day.
20:57So, I know there's only going to be two
20:59signals a year, but if for some reason
21:01you get two per day, you could take
21:03both, but not the third. And you skip
21:06Wednesday and Friday,
21:11but we want a higher quantity.
21:14121 New York window sweep model. Let's
21:18take a look at this one.
21:22So this this looks great on the outside,
21:25but it's been unprofitable for the past
21:27two years, and it still is a lower
21:29frequency. You're taking one trade every
21:313 weeks.
21:35Here's your criteria. It's in the New
21:37York open because that's when we're
21:38having most of our volume, most of our
21:39liquidity.
21:43But it's profitable if you use it in
21:442024.
21:522026. We're on a hot streak.
21:56Random control.
21:58I love that. I just clicked on this one.
22:03A random strategy with a 150 trades
22:06happens to beat 99% of strategies.
22:13What is this telling you? What
22:14information is this giving you about
22:16automated strategies?
22:19Because I'm using fully automated
22:21strategies right now to pass accounts
22:24and getting close to payouts.
22:29So, why am I getting close to payouts
22:33with complete automation when after I
22:36did a 5 million trade back test, which I
22:38didn't do before,
22:40and almost nothing is yielding positive
22:43results.
22:51I can think of two things that this is
22:54telling me. Number one, human discretion
22:58is an asset and you should use it to
23:01your advantage.
23:03And number two, I just shared a tool
23:05with my community, Ocean AI, this
23:07morning. And this tool is to calculate
23:12the probabilities of passing evals and
23:16getting payouts including the cost of
23:18evals and the size of payouts as quick
23:21as possible in a business fashion.
23:27Treating profits like a business. How do
23:29we pass and get paid without being
23:33profitable? We don't even need to have
23:36good edges in the market. We don't even
23:38need to have an edge in the market, but
23:41we can still win and get paid.
23:45The strategies I'm using to pass
23:46accounts and make money on funded
23:48accounts hardly have an edge at all. In
23:52fact, a lot of the strategies that we're
23:53looking at perform better, just less
23:55frequent.
24:05Unreal amounts of data. Unreal. Now, I
24:08show a lot of this
24:14to expand your understanding of what you
24:18can do with AI
24:20because a 5 million trade study
24:26took me
24:29one night
24:31to do.
24:35And I did smaller ones that are the
24:36reason that I'm passing accounts now.
24:42But you should look at this and you
24:43should say, "Wow,
24:46look what I can do if I take advantage
24:48of the opportunities at hand. Look what
24:51I'm able to build with these tools."
24:56because I already showed you what I had
24:58to do manually
24:59and how in an afternoon I can do 5,000x
25:04that.
25:10I hope this is just a testament to use
25:12AI to your advantage.
25:16I talked about it a lot yesterday
25:22for
25:23the video last week.
25:27AI is an enhancer.
25:33If you are lazy, you can still get these
25:35results. And they are more than the
25:38results that I who was working hard two
25:40years ago was getting. You can beat me
25:43from two years ago who is more driven in
25:46an afternoon.
25:49But I am still the same me and I am
25:53still the same driven person. And that
25:56is why I am succeeding with AI and
25:59making more than I ever have.
26:02This year I've made more money trading
26:03than I ever have. And I attribute 95% of
26:07that to my use of AI and my
26:10optimization.
26:11eliminating myself from the loop, giving
26:14more work to AI to do, and keeping the
26:18only part of me that is valuable, which
26:19is my taste.
26:23I would argue I'm working and investing
26:26myself more than ever.
26:30I was spending
26:32hours a day to get 10 trades on a manual
26:35back test. I'm still spending hours a
26:39day, but instead of 10 trades, I get 5
26:42million.
26:46And if you want to reap the rewards
26:50that I'm reaping,
26:52you need to be that highly highly
26:55agentic person,
26:58but with the tools at hand.
27:02Because I just showed you a whole bunch
27:04of information.
27:06NASDAQ is the most profitable.
27:09New York AM session is the most
27:11profitable. Those things are true
27:13because they have the highest of volume.
27:16What does that mean? News events are
27:18likely to play into that.
27:20Small stop- losses get eaten away by
27:22commissions. Every retrail strategy is
27:26going to yield, not every 99.4% of
27:30retail strategies are going to yield
27:32negative expecties.
27:34So, do you take those strategies and
27:36approach them in the market anyways and
27:39figure out the percentages and the RRS
27:42you need to still get paid or do you add
27:45yourself in the loop and build some sort
27:46of automation where you only interact
27:49with what's valuable
27:51where you only do what gets you paid?
27:53You make the final decision
27:56all off of your taste.
27:59That's how I made most of my money in
28:01trading, my taste.
28:04eliminating the emotion.
28:12I made an entire video breaking down
28:14this back test, not this back test
28:16specifically, but back testing with one
28:17of my students on how I do this.
28:21And in my community, we're doing things
28:23like this all the time, back testing
28:25crazy amounts of edges and talking
28:29directly with students on how do we use
28:31the information that we have to get paid
28:34because that's our goal. It's fun. It's
28:37cool to look at a dashboard with a lot
28:39of green and red scare squares, but how
28:40do we get paid? How do we take this
28:43information, apply it, live test it, not
28:45spend any money until we're ready to
28:47take that risk and then get paid?
28:54If you're interested in being a part of
28:55that community, there's a link down
28:57below to apply.
29:00I only take people who are serious about
29:02it. I only take people who really want
29:04to make something
29:06and build something. I really care about
29:10my students and I really want to see
29:11them win
29:15because I've won and I know what that
29:17feels like and I want to deliver that to
29:20you. That's why I'm making these videos
29:21as well because I want to show you I
29:24want to break the mindset because so
29:25many of you guys are just going to AI
29:27and saying like make me profitable bro.
29:31[laughter]
29:32You don't think everybody else has had
29:34the same exact idea?
29:37You need to go deeper than that. You
29:39need to apply this information and
29:40connect the dots. go through phases of
29:45a systemic approach to actually get
29:49paid.
29:53This was done with cloud code using
29:55fable 5.1 ultra code and it was done in
29:59a single night and it's more trades than
30:02I could ever hope to back test. I back
30:04tested a thousand manually in my
30:05lifetime and to look at the screen that
30:07says 5.6 six million trades is
30:10absolutely unreal. AI is your
30:12opportunity to beat everyone around you.
30:14It's your competitive edge and you
30:15should be taking the absolute most
30:17advantage of it that you can.