
๐จ๐พโ๐ป Exploring AI, investing in stocks and Bitcoin $BTC ๐ญ Posting my thoughts. Not Financial Advice. DYOR.
United Kingdom
Joined March 2010
- Tweets2.9K
- Following1.7K
- Followers2.5K
- Likes18K
Every list of 'stocks to avoid' has one name on it that's being punished for the wrong reason.
Give me the ticker and the one metric everyone's ignoring
77
Rate hike odds for Wednesday's Fed meeting jumped from 70% to 90% in three days on one hot CPI print, and $VOO still gained about 1.1% Friday anyway.
A probability swing that fast is noise until it changes what I'd do Monday. This one didn't
80
Claude, GPT, and Grok all handle the easy prompt fine. The real gap shows up around tool call 40, when nobody's watching and one bad call can cascade.
Which one do you actually trust to run unsupervised?
84
I don't track every headline that moves a stock 2%. I track the ones that would've changed my position size if I'd known them a day earlier.
Tell me the last headline that actually would've changed what you did, not just how you felt
96
Realty Income ($O) raised its monthly dividend to $0.2715 a share, the 136th increase since its 1994 NYSE listing. That's 675 consecutive monthly payouts now.
1
103
Most agent failures aren't reasoning failures. The model reasons fine. They're context failures: the right instructions paired with wrong or missing information. Everyone blames the model. It's usually the pipeline.
72
Which sector is quietly the biggest chunk of your portfolio right now, on purpose or by accident?
77
Oracle ($ORCL) posted a $664 billion backlog Friday, and management says roughly half converts to revenue within their own stated window.
Backlog size is the headline. Conversion timeline is the number I actually filter for
115
Everyone says they'd trust Claude, GPT, or Grok with real work. Nobody agrees on what real work means, and that's the gap that matters
89
The point of a checklist isn't catching mistakes. It's removing the moment where you decide whether to bother running it
83
A plan written after you're already in the trade isn't a plan, it's a rationalization. Write it before you enter, then ask Claude to poke holes in it.
What's the last plan you wrote after the fact and called it strategy?
80
GM crew. Most people want a better strategy. What they need is the patience to run the boring one longer
82
The riskiest action in a Claude agent workflow isn't the wrong one. It's the one that can't be undone. Gate irreversible steps, a send, a delete, a trade, behind a manual check and let everything reversible run on its own
84
Most systems fail from too many steps, not too few. If a process needs a slide deck to explain, nobody's running it under pressure
84