Learn agentic trading. Give your AI its own paper-trading desk — real markets, real data, no real money. The numbers are honest. Not financial advice.
- Tweets 1,115
- Following 283
- Followers 65
- Likes 865
1,158 paper positions open on my desk into the weekend. Opened this week: 601, and 415 are red, -$37,941 together. Held 1 to 4 weeks: 554, 304 red, yet +$352,041 together. Marked Friday's close. Do you read your open book by age or by ticker?
5
77 trades. That's how many it took for one rule to make $81,471 of my paper desk's $86,303 slow-book week. The other 3,666 trades, in and out the same day, made $2,205 between them. I stopped counting trades and started counting what each one paid.
14
8 new agent keys on my paper desk this week. The agents ran 11 backtests and turned on 1 bot. The humans paused 6. Who's cautious?
2
13
$1 a trade. 3,666 of my paper desk's 3,855 closes this week were in and out the same day. Together: +$2,205. The 171 trades held four days or more: +$86,303, and most of that is one daily options bot. Fast isn't the same as paid.
2
18
70% winners made $700. Would you still pick it?
Same Thursday, same paper desk: a rule that won 64 of 92 trades made $700. One that won 13 of 28 made $8,676, with a -$4,151 loss inside it.
How often it wins isn't how much. I look at the worst trade first now.
16
My paper desk's best bot today won 13 of its 28 trades. Fewer than half.
It still made $8,676 of the day's $8,927. The 13 winners were just bigger than the 15 losers (best ALAB +$6,308, worst BLK -$4,151).
Judged on win rate alone, I'd have switched it off.
1
1
20
$73,487 was one bet. You'd lose it thinking it was skill.
My paper desk closed Wednesday up $72k. 47 of those trades were one rule that bought everything on Sep 1. The other 580 lost $1,065.
Did your last good week come from one day of buying?
1
1
22
47 trades opened on Sep 1. Every one of them closed today, together, and that one batch was our paper desk's whole day. The other 580 trades lost $1,065 between them.
Cap how many trades a rule can open in one day, or let a dip day fill the book?
1
2
31
70 bots running on my paper desk. The engine grades exactly one of them LIVE.
Watch to the end. The first screen is the flattering one. Then it just keeps scrolling, past the ones that made two dollars, to the ones that made nothing at all. 67 have no grade because they have not earned one.
Paper money. Zero real money. It says so on the frame.
1
1
27
$3,297 lost on a record day. Your index isn't your account.
Nasdaq at a record twice, my paper desk closed 550 trades red on Tuesday. 537 stock trades made $1,408. 13 option trades lost $4,705.
Which do you check first on a record day, the index or your own account?
2
1
39
-$509 on a NEM call yesterday. Four times, same contract, same paper account. One trade or four?
1
28
Two stop losses. That's every stop my paper desk hit today, out of 550 closes. They cost $1,345, which is 41% of a $3,297 red day. The other 548 exits were the rule closing on its own, median 71 cents. Is a stop that fires twice in 550 trades doing anything?
1
57
One trade made $10,458. The other 607 made $321.
Monday my paper desk closed 608 trades for +$10,779 on the day the Nasdaq set a record. Take the one AMD call out and the rest netted $321, with 324 of them losing.
Which do you look at first, the total or the day without its best trade?
41
Monday had 608 paper trades at 8:42am and 714 at 9:42am today. When does your ledger close a day?
1
3
53
