Journey to $5,000 a month trading prop firms with a bot

Today I have decided to start a journey with my bot (that I built) trading prop firms to see if it can pass their evaluations and eventually get payouts.

The reason I am doing this is so that I can be more disciplined with my trading. I often look at the charts myself and then take a trade on the same account and end up blowing the account. So I think if I can create a journal it will stop me from taking stupid trades and just let the bot take the trade.

I only have one account at the moment and I lost $1,000 on it today manual trading because i thought I knew better. This is supposed to be a purely passive income system so hopefully this journey will make me more disciplined to leave it alone.

Over the past few months I have trained the bot (which we will call Dave from now on) to trade 50K accounts which have an end of day drawdown of $2,000. This means if Dave loses $2000 the account is blown.

The aim is to trade every day if opportunities arise (which they normally do) and to trade 24 hours a day.

Dave has been paper trading since the end of August and his results look okay. He normally takes 1-2 trades a day.

THE SET UP

Dave is a python script create entirely by Claude AI (opus) with the help of my input. I can't give you any information about Dave as it has taken me quite a while to make him.
TradingView which is a chart trading software that you can download on to your PC is used for the Data. There is a free tier which Dave can use.
PickMyTrade is a copy trading platform which dave uses to route my trades to the broker. In future if this works out I can have multiple 50K accounts and PickMyTrade will be able to trade all the accounts at the same time. This costs $40 a month.
The 50K accounts are from Apex trading and they normally have deals on. Dave has 30 days to make his target of $3,000 to pass his evaluation (as I said I lost $1,000 today manual trading the account so he has to now make $4000 in a month to pass his eval. As you can imagine he is not happy with me. )

That is basically my software set up.

THE TRADING SET UP

Dave has two set ups that he can trade. The first set up is more frequent than the second.

SET UP 1

• Timeframes: 30-minute and 60-minute (15-minute excluded)
• Target: 2:1. Stop stays put — no break-even move
• Risk: $240 a trade, sized assuming a 3-point adverse fill, maximum 20 contracts, and no trade if the stop is under 4 points
• Rails: at most 4 trades a day, stop at −$750 on the day, flat by 16:58 New York, and nothing held longer than 8 hours

SET UP 2

• Stop: 25 points. Risk $200 = 3 micros
• Exit: that bar’s close, 30 minutes later. No target — it’s a one-bar trade
• A signal more than 2 minutes old is dropped rather than chased
  
  
• One trade at a time — a signal arriving while anything is open is skipped
• New entries stop once the day is −$500

That is basically it. Dave will start his journey at 23:00 GMT. He trades every weekday from 23:00 until 21:58 GMT.

Follow along if you are interested.

good luck

i am doing something simillar

Interesting journey, best of luck

Good luck bud, waiting for the results.

Following, I’ve always been interested in day-trading automations. Good luck with your journey, looking forward to the next update.

i am also doing trading now a days good luck to your journey

Good luck, waiting for results.

How long did it take you to build Dave? And did you come up with trading strategies yourself or base them on something existing?

DjonZ said:

How long did it take you to build Dave? And did you come up with trading strategies yourself or base them on something existing?

Click to expand...

A lot of the strategies came from watching you tube videos and then seeing if I could turn those strategies into a mechanical system. I created a lot of strategies and just let Claude AI capture the data. Once I saw the two strategies for Dave I started to concentrate on them and fine tune them to see what worked. So in all it has probably taken 3 months.

I have also created a new strategy today based on the 15 second chart. It had good results this afternoon but then had lots of losing trades this evening. So I have now put a cut of time of 18:00 GMT for this one to see how it performs. We are just going into the Asian session and will see how it has performed in the morning.

As for Dave we had no trades since 23:00 last night and then suddenly he decided to open a position at at 21:11 GMT tonight. As we are trading prop firms I have to get him the flatten the position before 22:00 as prop firm do not let you carry your trade over to the next day.

I am pleased to say he did make $165.00



I will try and update as much as possible for both my strategies.

Woke up this morning to find my trading view had crashed. The 15 second bot made +250 points last night which I lost out on. I am working off a laptop and will need to move this over to a VPS when I get some payouts, if I get some payouts

Good luck, man. That’s a hell of a market you've chosen to test a bot on. E-mini's, right?

What do you mean by "TradingView crashed"? The desktop app?

Somers said:

Good luck, man. That’s a hell of a market you've chosen to test a bot on. E-mini's, right?

What do you mean by "TradingView crashed"? The desktop app?

Click to expand...

Yes the desktop app which the bot reads from to place orders crashed. So no orders were sent. It is because of the memory resources on the laptop. I ahve removed some indicators from the charts which Claude AI advised to hopefully that might sort it out. But I really need to find a cheap VPS or buy a mini computer. Yes I am trading E-minis.

The London session was not good for the 15 second bot today. I will report back at the end of the day

Interesting. I didn't know you needed a TW desktop app to use their data. Their app is just a "webview wrap" kind of app.

I think Claude can manage to make requests to their servers without the app.

Well, if you are able to talk it into that.

Somers said:

Interesting. I didn't know you needed a TW desktop app to use their data. Their app is just a "webview wrap" kind of app.

I think Claude can manage to make requests to their servers without the app.

Well, if you are able to talk it into that.

Click to expand...

Tradovate provide an API but not for prop firm trading. The app which is a webview wrap kept logging me out after a period of inactivity. I have now resolved that by making phantom orders that would never get filled and then cancel them. Tradovate then thinks that I am constantly using the app and doesn't log me out.

So yesterday was a bit if a s**t show. Dave traded once and lost $157.50

and the new 15 second bot did not trade Asia as trading view crashed which lost me 175 points ($350), London was really poor and in the New York I took the laptop in the car and ran out of battery.

The 15 second bot during London and New York made 36 trades and made $672 but also lost $862. So overall lost $189.50. If it had traded Asia It would be $160.5 in profit.

I also spent the whole day tinkering with the 15s system and making manual trades which also cost me money. The whole reason I started this journey was to be more disciplined and then I go ahead and think I am better than the system.

So starting with the Asia on Sunday night and I am literally going to walk away and not even think about it to see what my profit of loss is by 21:00 GMT on Monday night.

With the 15 second system it definitely needed some tweaks and i learned to put some safeguards in place just in case tradovate or trading view crash. I now have a python script that sends a message to telegram if they go down.

I am also looking at Windows VPS to migrate to this week.

Will report back on Monday night

You should back test every strategy you find using years of data. This will help you find the best setups, win rate, etc. this is very important in trading

I must ask. You did test it on historical data first, right? Or on a paper account, at least?
It just looks like you are testing it right now with real money.

Somers said:

I must ask. You did test it on historical data first, right? Or on a paper account, at least?
It just looks like you are testing it right now with real money.

Click to expand...

I paper traded various strategies for a couple of months and Dave which combines 2 strategies was positive.

The 15 second strategy is forward testing. You need to get a premium plan from trading view to get 15 second tick data. So I am happy to lose $45 on a 50k apex plan to test it. I should know by the end of the week if it is a runner. The thing with back testing is you can't add your own money management. You can add your risk and reward but that's about it. My aim would be to see if the 15 second could make $1000 a day and then stop trading for the day. Asia looks better for those system at the moment so I might trade 2 contracts for that session and then drop down to 1 for London and New York. You can not add these filters into your back test.

I'm doing something similar But for the fun of it ... for right now, my bot is in testing and learning mode ... but is keep improve the win rates each day ... but I need more data so to can have a strong, smart bot can take 1/2 months of data Trades taken by the bot in the past 7 days: the winning rates. Yesterday: 5 trades, 4 wins and 1 loss

zocx said:

I'm doing something similar But for the fun of it ... for right now, my bot is in testing and learning mode ... but is keep improve the win rates each day ... but I need more data so to can have a strong, smart bot can take 1/2 months of data Trades taken by the bot in the past 7 days: the winning rates. Yesterday: 5 trades, 4 wins and 1 loss

View attachment 552939

Click to expand...

Keep going. You are trading crypto?

So a quick update.

I tried to switch to Tradovate to put all orders through the desktop app. But again it crashed in the early hours of Monday morning. So the Asian session was ruined again for the 15 second bot (which I am now going to call Bob from now on). So I have switched back to Pick My Trade to route the orders. The latency can be anything up to 30 secs which is not ideal but at least it is reliable.

So last night Bob did manage to trade almost all the Asian session up until 06:30 this morning when there was a crash.

The good news is that it made $402.96 trading 2 contracts. It has given some back in the London session that started about an hour ago but it is encouraging even if it is just 1 full night of trading. I am going to monitor the Asian session all week and if its win rate is good over this week. I might just switch to trading the Asian session.

Dave has also done well over the past 2 days he traded 5 times with 4 wins with a total profit of $914.00. (screenshot of Dave).

Still early days