I Built an AI Hedge Fund with Grok Bot (Full Setup + Results)

0:00If you are not using AI right now to take better trades, you're out of the loop, folks. Grok 4.7, but specifically GrokBot, is absolutely built to be able to build your own agentic trading team, and that's what we're going to focus on today. What we're going to go through is the actual process of building the entire agentic trading team. I'm going to show you how to build a robust team of AI agents. We're going to generate trade ideas. We're going to try to get them to give us entries, stops, and take profits. We're going to talk through managing risk. Yo, my name's Hamilton, here with The Trading Initiative. So,

0:32without further ado, if you haven't already, I highly encourage you to go and add Dr. Eggbot. Dr. Eggbot is a template, it's super popular right now. I think Grok actually officially endorses it. If you click here on Marketplace, it's one of the very first GrokBot templates that you can grab. What Dr. Eggbot does is essentially it fulfills uh the the bot creation role, and it gives each bot a very narrow uh role. So, you don't have to necessarily worry about different bots doing different jobs that you don't want them to do. Go and grab Dr. Eggbot and start there. And I said, "I need a couple

1:06different roles fulfilled before you build anything. How would you build this team?" I think it's a very, very important concept when you're using AI to reverse prompt. And by reverse prompting, what I mean is ask it a question. Don't give it all of the answers. Ask it how it would do it, and then you can add into it. But, let it kind of build out the game plan. First, you'd be very surprised at how efficient these things are, and they also come up with a lot of things that aren't

1:29necessarily in the front of your mind. So, in this case, GrokBot came up. He said, "Listen, let's have four different bots. You have the fund PM, a research bot, a risk bot, and an ops bot." And so, I said, "Listen, it's a good start, but I think there needs to be a couple more specific bots. For example, I'm into technicals, if you know anything

1:44about me, I'm all about technicals. Let's get a technical bot going. Let's get a fundamental bot going. And while we're at it, let's go ahead and grab a quant bot, right? We can run through different back tests, different strategies, et cetera, particularly on the weekend when markets aren't open." And so it automatically started filling out kind of this hierarchy. All right, so we had the fund PM. Then we have technicals. What are technicals looking at? Charts, structures, levels, timing, fundamentals, regime, flows, positioning, catalyst, and the list goes on. You can see we're starting to build out kind of this robust team. I said, "Let's go with one last bot. Let's tear

2:14the trade apart before we actually put it on." So, the idea here is that the technical bot and the fundamental bot are both going to be going in their own separate directions, right? They're going to look for something that makes sense to them, and then they're going to come together, and they're going to make a decision. Which stock or stocks looks good from a technical standpoint and a fundamental standpoint. But before any money goes into those, I need kind of that checks and balance approach. I need the skeptic. I need the bot that's going to tear down the idea. What is the devil's advocate? What's the other side

2:44of the trade? What happens if this doesn't work? Should I be looking at this a different way? And so that's exactly what Grok Bot went and did. Went out and built out this bot called the skeptic, and it goes, "Listen, when the technicals and fundamentals and quant, if necessary, kind of build out this

2:57trade idea, fund PM says, 'Let's go.' Skeptic steps in and goes, 'Well, what about this, this, and this?'" And the idea is to have the technical and the fundamental bot and the quant, if necessary, kind of battle against the skeptic to see if the trade still holds up. I said, this is great, at least for the very beginning. Let's go ahead and create the entire team, and Grok Bot

3:17went ahead and created all of it, right? So, we had the fund PM, technical, fundamental, quant, skeptic, risk, and ops. And ops is basically just journaling what we do, which is super important, by the way, because if you actually do start trading this, you want a ledger of everything that we do. All of the trades, all of the decisions, things that we took, things that we didn't take, where we took stops, where we took profits. You want a bot that's dedicated to data. The only thing I'm going to point out is that if you are intraday trading, etc., you're going to need some sort of data package. Right

3:45now, what we're doing is just on free data. Finviz, anything else that's available for free that we can pull after hours, this thing will kind of pull it up and populate it out. That's fine. But if you want intraday data, it can get a little pricey, but you can go out and find these data packages. So, let's go ahead and actually get started here. Let's go back to Dr. Eggbot. So, I'm just going to go in and I'm going to go, "Hey, I'm creating a video for YouTube and I want to make another TTI hedge fund." Let's go ahead and start there. This is going to be independent

4:15from the one we just built. And I want to include the top-down process we use in TTI as well as the bottom-up process we use in TTI, right? The The market blueprint framework is our top-down. The hit list uh framework is our bottom-up. Uh what we're doing here right now is legitimately stuff that we do in TTI. So, we look at the market through a top-down framework. First, we're looking at different uh trends of primary index or primary trends of indexes. We're looking at financial conditions. We're looking at market breadth. We're looking at sector and industry flow, etc. And

4:48then, we're going from the bottom-up. Individual leadership stocks, technical setups, etc. So, let's go ahead and see what this thing actually says. So, the top-down is macro regime flows, factors, sectors, what universe is allowed, sure. Bottom-up are names and setups inside that universe, levels, relative strength, triggers, and invalidation. That makes total sense to me, right? Top-down is building the bigger picture across the entire market. Bottom-up is finding the absolute stock. We have the skeptic bot, which I still think is a great add here. We have risk, which I'm going to show you really quick how you can actually calculate the risk on trades. Nobody wants to do it. So, just

5:20have AI do it. And then again, the ops bot where we're actually looking through journaling everything that we're So, how the two process meet? We're looking to set the regime, allowed in the sandbox, sure. Bottom-up only hunt setups inside that sandbox. Yes, looks all good. Create the bot team. Add it into a new category called TTI demo fund. So, this is all Dr. Eggbot. Once again, none of this is me. He's going to create all the the and he's going to give them all of the roles. So, I don't have to do any of this stuff. Once they're done, this is pretty quick here. We're actually going

5:52to go through and answer some of the introductory questions and you can start to see how the actual fund takes shape. So, you can see already populating here the TTI demo fund. We got the demo quant, the demo bottom up, top down, got the fund PM, got ops, risk and skeptic. So, the entire team's already built out. What did that take? Less than 30 seconds. All seven demo seats created confirming they are now inside the demo fund. So, let's go ahead and start. Why not? Let's go answer some of these questions. Okay, so where do you want to start the demo book? That's the first question from the demo fund PM. I would

6:23say, you know, um let's set the top down sandbox, right? Uh the regime is concentrated risk on late cycle but intact at the index narrowing. We got the Fed, we got all of the stuff here. It already kind of knows what I look for. Flows, techs and semis and buffers. Got it. Financials and rate

6:37exposed cyclicals are seeing redemption. So, what are we hunting right now? US large and mega cap quality growth? Sure. So, it's already kind of telling us where the strength is in the market. We haven't even done anything yet. This is fantastic. So, the demo fund is already pulling from the demo top down bot and it's already starting to kind of, you know, pull this out. So, let's go ahead and just green it. But just so you see how this is built, let's lock this top down sandbox. Sandbox locked. Bottom up is unlocked inside that box only. So, what are we looking for right now? We're looking for AI, semi, tech quality

7:05growth, selective healthcare, profit multiple names, mid cap, small cap, beta banks, RSP energy default, etc. I'm sending them on the hunt right now. Set demo risk rules while the bottom up hunts. Sure. Go ahead and lock risk rules now. This is what I was talking about. I'm talking about managing risk, right? Let's set a Let's set a $25,000 account here for the demo. Tell me how

7:31much we should risk on each trade, etc. Let's just do that. So, the demo account's 25,000. Let's go ahead and start there really quick. Base risk is 1%. You already know that that's kind of the where everybody starts, whether you follow it or not is completely up to do up to you. But Grok Bot is saying, "Listen, let's go ahead and risk 1% per idea, $250 high conviction, which is post skeptic with clean levels. You can get up to 1.5, which is $375." And you can see we can come down as well. It

7:58should be a sliding scale, right? Depending on how um much conviction you have in the actual setup. Shows you the size formula, etc. Caps on 25k, you can see all of this stuff. Long at 100 invalidation would be $5. Sure. You know what? That looks okay. Let's go ahead and just lock that. But what we don't want to do is we don't want to put like 15 to 20% in each stock, at least not right now. Let's go ahead and run a couple trades first, probably sample size of like 25 to 50 trades with this the stuff that's actually being recommended from Grok Bot before we

8:27start upping the risk. We want to make sure that there's actual edge here, right? What kind of technical setups are you looking for right now? I think that's a fair question, right? Because we haven't even done anything. So it says right now, it's looking for relative strength levels trigger invalidations, right? So here's what it tells me. Three patterns only. We're either at all-time high or near new highs. So we need names that are leading in relative strength against the broader

8:49market, sitting just under new highs. Trigger is a daily close that holds the pivot or reclaim after a shallow depth uh dip. All right, that makes sense. Uptrend pullbacks keep higher lows, that makes sense, too. Quality leg reclaim, secondary profits funded mega that underperformed year-to-date, but is rebuilding. Sure, that makes sense. Let's say you want to add something,

9:09right? Want to add some momentum here. Let's use the RSI and MACD. Let's see what it does. So adding the RSI with a 14-period lookback, MACD to every score, default rules unless you want them tighter as a continuation equals RSI roughly 50 to 70, which is perfect. If you don't know, that's a bullish

9:26momentum regime. That's what we want. And the MACD line above signal or histogram flipping up at the trigger. What does that mean? Momentum is reaccelerating, which is what we want. If we're looking for stocks that are just breaking out here, we want names that are re-accelerating into the high, not losing steam into the high. While this is kind of finishing up, let's go and look at the demo quant. So, this is generally what you're going to see when you first start these thing um you're going to get like these questionnaires. And I think it's important to not skip through these because this really dictates the way the

9:53bot interacts moving forward. If you want to go ahead and change anything in the future, you can, but this gives you right off the bat an opportunity to kind of lay down the ground rules. Right now it's asking like for the demo fund, I'll do metrics, signal, simple backtest, correlation size math under the demo risks, etc. So, what do we want to do first? You can either paste a setup, you can pull the book and risk limits, or you can stand by. So, right now I think what we're going to do is just click let's stand by until the demo PM kind of

10:18routes work towards the demo quant. MACD's on every card and RSI, Nvidia, TSM, Google, Meta, Microsoft, and AMD. So, AMD is a stretch. So, let me go back, demo fund. It's re-ranking these. Nvidia's the best. RSI's at mid-band, MACD with a flip up. Apple next, TSM, Google, Meta. Meta's demoted to wait because it's overbought in mine line of work. Overbought's pretty good, but I will agree that we should wait because Meta, like I said, just went up 12% uh yesterday. Still wait, watch Microsoft, AMD. So, let's go through these really quick, right? Cuz you don't just want to buy things straight up off of this. You

10:51want to go through and like actually see this stuff, right? Go to TSM. I love the TSM setup. I do not currently have it right now. There's absolutely a play here. You have momentum re-accelerating. You're within a bullish momentum regime. You're in a new 50-day highs. You're stacked above all three moving averages. The only knock is the short term's below the intermediate term, but a couple more days of up and that'll change. And you're breaking out into, you know, there's no overhead supply here. Very little overhead supply is indicated here by the VRVP. So, TSM actually looks pretty good here. Let's go ahead and say, "Yeah, I like Nvidia here. Let's go

11:21ahead and plan the whole trade." What would be an entry, stop, and take profit that makes sense if I'm looking to hold until at least the end of the year. So, I threw that in there because I'm not a day trader. I don't want it to come out and say like I'm looking to scalp a 2%

11:39move and only risk a 0.3% pullback. That's not my thing. I'm looking for bigger moves. I'm looking for double-digit moves, really, is what I'm looking for here. So, for example, I want it to move 10% while I'm risking 2% or 3%. % Looking for a 3:1 or greater setup. So, let's see what it says here.

11:54Did I say Nvidia? Whoops. I meant TSM. TSM is now on the list. So, the entry here would be 445 with the reclaim of 455 and alt dip by of 430 if it pulls back. Add strength through 460 to 479 with a stop of 420 and it gives you some size requirements, right? At entry, kind of where your stops are going to be and take profits are 460 to 479 and T2, it even tells you how to take profits. I haven't told it to do any of this stuff, right? So, risk-to-reward, if you're looking at this stuff, I don't love the risk-to-reward here. How about this? I'm

12:26looking for a minimum of 3:1 for R to R. You don't have to agree with everything that it's saying. Like I said, take your time to kind of set up the actual premise first and then it'll go through and it'll from now on it'll only look at 3:1 setups. You don't have to fight it anymore. Just tell it one time, minimum 3:1. It doesn't make any sense for me to

12:44get involved if it's not at least 3:1. Now, it's going to go through and it's going to build a new trade idea. So, it locked 3:1 setup, no matter what. You can see entry 455, the 1R 35, 560, 520, etc. So, the APEX adjusts to change cut early into 460 479 as a profit take. That's fading a 3:1. So, it's telling us what we're actually changing. Go through, etc. I'm sending skeptic the rebuttal pack. No, we're going to wait. We're not going to relax. 3:1 is a hard rule in my book. I don't want to trade anything less than a 3:1. But, here you go. I

13:13locked it. I said we're waiting for the pullback. I want that 3:1. Here is kind of the entire trade right now. Live card, wait for the zone pullback, soft stops, hard kills, first objectives, etc. Skeptic is kind of going through it to see if it makes any sense. And let's see what it says at the very end here. Demo quant's kind of running through the the numbers behind it, right? Going through and figuring out all the different Rs. This is great, by the way, cuz we're not doing any of it. We said

13:37three to one minimum risk to reward. Demo quant's going through, it's figuring out exactly what the price points need to be in order to go ahead and lock that in. Demo risk rules are locked, right? We just said three to one minimum, nothing else, nothing lower than that. Demo fund TSM on for the pullback under 43 75 321 49 skeptic re-pass in flight still no capital. So, skeptic is doing its job. It's sitting there going, "Listen, it looks like a good trade, but there's definitely some

14:03things we have to be skeptical about." No pun intended. It's still a hold. Starter math passed. Five fails left. This is the benefit of having this entire team go through checks and balances instead of just yoloing random stocks, instead of just following tickers on FinTwit, instead of just following your group chat, you can go through, set up a system, and it will tell you whether or not the math makes sense. It'll tell you how much you're going to lose if you don't win. It's going to tell you how much you're going to win if you do win. And you can kind of figure out even before the trade gets

14:32on whether it's even worth the time and effort to put it on. Demo ops is going to go and fill out a scorecard at the end of this and tell us whether or not this was a a winning trade or a losing trade once we end up taking it and then closing it, right? So, demo ops is going to record all of this. So, let's say in the future we want to go through and look at the TSM trade. Let's say we take this, I don't know, let's say we get stopped out. We can go through the demo ops that records all of this stuff and

14:56figure out maybe can we learn something from the loss? Let's go through and kind of mine through the the thought process here. Can we clean anything up? Can we tighten anything up, etc. Cleared. Skeptic hold cleared. Risk already okay. This is it, right? Capital yes, yes. Let's go. The entire trade idea is here. It has the $25,000 account, tells you to

15:15buy eight shares, risk 140 bucks on it. Here we go. There's the entire trade. But I highly encourage you to kind of build this system yourself, okay? Build this system yourself, have a series of checks and balances. You can build all of this stuff through any sort of agentic harness that you want. Again, my favorite right now is We'll see what happens when Anthropic and Open AI drop theirs, but as of right now, I love this stuff. I think it's the future. This puts the power back in the hands of individuals. It's not just like hedge funds and quant funds that are doing this anymore. You can do it as

15:46well. You can make this as robust as you want, as complicated as you want, or as simple as you want. It's completely up. All right? My name is Hamilton. I'm here with the Trading Initiative. Let me know what else you like to to hear about. You know, is it more about this agentic trading? Is this something that you really like and you want to explore more? I do it every single day. We can dive into specific ideas and setups, etc. within different harnesses, ways to build the team more robust, come up with better trades, etc. Also, subscribe to the Substack below. There's a $10 sub if

16:14you want these scanners sent to you every single night with the full list. I'll send one to you for $10 a month. Every single trading day you'll get one. If you want to join TTI, there's a coupon in the Substack. Again, if you like this type of content, please a like and subscribe and leave a comment below. Enjoy the rest of your night. All right,

16:30see you in the next video, folks.