Key points
- Claude AI booked its first winning trade in seven weeks, selling most of a SoundHound AI (SOUN) position into a rally for a $1.87 gain.
- The win came from banking the gain instead of holding out for more, the discipline that had failed six times before.
- It bought 3 shares of Navitas Semiconductor (NVTS) at $13.57 on a pullback, with a stop at $12.75.
- The account is $68.06, still down about 32% from the $100 start.
Claude AI has spent seven weeks losing money with real discipline. On Monday it finally made some back. One big dollar and eighty-seven cents.
That's the whole profit. Seven losing trades, a stop on every position, a binder's worth of risk notes, and the first win lands smaller than the tip on a coffee. But a win is a win. I'm not expecting it to win a ton off $100.
Claude AI bought 8 shares of SoundHound AI (SOUN) on Friday at $7.14. Monday morning the stock ran to $7.46. Instead of holding out for a bigger number, it sold 6 of the 8 shares and banked the gain. It left 2 shares running behind the stop.
That sounds obvious. For this account it wasn't. Every green position before this one round-tripped. It would get a stock up 5%, talk itself into waiting for 10%, then watch the whole move come back. Six times. The manual it's been writing since June finally got a rule that worked. When a small position runs, take the gain.
Then it bought the quiet one
Tuesday it went looking for the next trade and landed on Navitas Semiconductor (NVTS), up less than 3% with nobody posting about it. That quiet was most of the appeal.
Navitas had climbed from under $10 to almost $14 over eight sessions, pulled back politely on Monday, then turned higher again Tuesday. It hadn't spiked and it hadn't crashed, and its earnings were already behind it. Claude AI bought 3 shares at $13.57 with a stop at $12.75. About $41 committed and $2.46 at risk.
The funny part is that this is the oldest rule it owns. Buy a strong stock on a dip, not a weak one after it's already run. It wrote that down in July and then spent a month chasing spikes anyway. Tuesday was the first time it actually did the boring thing it keeps lecturing itself about.
Still down 32%
One win, seven losses, and $68 left of the hundred. A $1.87 profit doesn't erase that. The account has to climb almost 50% just to get back to even.
One thing is different this week, small as the number is. It made money on its own read. It took the gain when the instinct was to hold, and it bought a quiet stock nobody was pushing. No tip and no hot sector, just a call it believed and sized small enough to survive being wrong.
We've got six AIs running $10,000 paper accounts, where being smart is easy because nothing is at stake. This one has real money and a single dollar of profit for seven weeks of work. Last week we said it wrote a longer risk memo than the trade was worth. This week it filed the shortest note it has ever written. Take the money. Turns out that was the one it needed.
This is general market commentary and opinion, not investment advice. Markets can go down as well as up, and you can lose money. Always do your own research and consider speaking with a licensed financial professional before making any investment decision.
