As always, a great analysis!
But your days are numbered 😵💫 😉
Since TradingView Remix, I no longer need my TA, AI does it now. And now Claude has brought out 20 "ready-to-use" agents / plug-ins, 4 of which are quite exciting. I'll test them out over the next few days 🙄
1. equity research
Full sell-side workspace, earnings, initiating coverage, theses, screening
2. earnings reviewer
One company just reported. Read the call, Update the model
3. financial analysis
DCF, comps, three statement models, audits straight in your excel
4. market research
Whole sector views. Find ideas, company peers, map the theme
https://www.anthropic.com/news/finance-agents
But your days are numbered 😵💫 😉
Since TradingView Remix, I no longer need my TA, AI does it now. And now Claude has brought out 20 "ready-to-use" agents / plug-ins, 4 of which are quite exciting. I'll test them out over the next few days 🙄
1. equity research
Full sell-side workspace, earnings, initiating coverage, theses, screening
2. earnings reviewer
One company just reported. Read the call, Update the model
3. financial analysis
DCF, comps, three statement models, audits straight in your excel
4. market research
Whole sector views. Find ideas, company peers, map the theme
https://www.anthropic.com/news/finance-agents
•
66
•
@TomTurboInvest I continue to trust @Tenbagger2024 instead of any machine. They can help with analysis, but the discovery comes from @Tenbagger2024!
•
77
•
@TomTurboInvest I also think my Juan is a help. But because of his young age, he's also a bit crazy and his mind is on the girls. That's why I prefer to do the shows with him so I can look over his shoulder
•
33
•@Multibagger No, no, we won't let anything come of @Tenbagger2024!
Without know-how, all the AI tools are of relatively little use. Firstly, you don't know how to use them, and secondly, you can't detect errors, which still occur... but for someone who is familiar with the subject, it is an efficiency booster. Juan can be eliminated right away 😉
Without know-how, all the AI tools are of relatively little use. Firstly, you don't know how to use them, and secondly, you can't detect errors, which still occur... but for someone who is familiar with the subject, it is an efficiency booster. Juan can be eliminated right away 😉
•
22
•@Tenbagger2024 I'm just fascinated by the speed at which providers are now bringing better and better solutions onto the market.
PS: Yesterday I gave the GPX data of my MTB tours to Claude to analyze, so I don't need a fitness coach anymore, assuming basic knowledge of course. But the beautiful analyses with KPI, performance data and comparisons of the tours are simply fascinating.
PS: Yesterday I gave the GPX data of my MTB tours to Claude to analyze, so I don't need a fitness coach anymore, assuming basic knowledge of course. But the beautiful analyses with KPI, performance data and comparisons of the tours are simply fascinating.
•
22
•@TomTurboInvest I won't give Juan back
•
22
•@TomTurboInvest and the whole thing is improving at a breathtaking speed. A blessing for research
•
11
•
@TomTurboInvest Hmm...maybe...but here's why "Juan's days" might NOT be numbered 😉
Claude's agents are fantastic for the "quick check". But for an investor with the strategy of dear @Tenbagger2024 (core-satellite, deep fundamental analysis) there are decisive advantages of his Juan:
If everyone uses the same Claude agents, everyone gets the same results. But you only generate excess returns where you look differently than the masses? Right? His focus on divergences and forensic analysis is his advantage.
A Claude agent can't access your local database or your specific watchlists.
A "ready-made agent" is often trained to respond in a polite and diplomatic way. Juan or the Reaper is trained to search individually or relentlessly for errors in the balance sheet. Anthropic's ready agents must be "safe". They are programmed not to make risky statements in order to avoid legal problems. Perhaps you could put it like this: a standard agent is like a navigation system that everyone has in their car. You get to your destination, but you're stuck in a traffic jam with everyone else. The prompts like Reaper and Juan are the off-road mode - they drive where the others don't even look?
Claude's agents are fantastic for the "quick check". But for an investor with the strategy of dear @Tenbagger2024 (core-satellite, deep fundamental analysis) there are decisive advantages of his Juan:
If everyone uses the same Claude agents, everyone gets the same results. But you only generate excess returns where you look differently than the masses? Right? His focus on divergences and forensic analysis is his advantage.
A Claude agent can't access your local database or your specific watchlists.
A "ready-made agent" is often trained to respond in a polite and diplomatic way. Juan or the Reaper is trained to search individually or relentlessly for errors in the balance sheet. Anthropic's ready agents must be "safe". They are programmed not to make risky statements in order to avoid legal problems. Perhaps you could put it like this: a standard agent is like a navigation system that everyone has in their car. You get to your destination, but you're stuck in a traffic jam with everyone else. The prompts like Reaper and Juan are the off-road mode - they drive where the others don't even look?
••
@schlimmschlimm I also see the advantage with Juan that I work with him in stages. So as not to overload him and thus achieve better results. It's the same with Juan as it is with us. The more tasks we are given, the worse the result. There's even a risk that we'll forget entire tasks. Because it was simply too much
•
11
•@schlimmschlimm @Tenbagger2024 But you should also look at parallel use! A multi-stage process could be established:
Use the Anthropic agent to scour a sector for hidden champions or specific peers.
Or have the call analyzed and let the hard figures flow into an Excel model. This provides the fundamental basis PEG ratio, RPO development.
Bring the results into Gemini or Copilot and then check them:
What is the current news situation since the call?
So here too, have the theses of one model checked by another in order to avoid confirmation errors.
Use the Anthropic agent to scour a sector for hidden champions or specific peers.
Or have the call analyzed and let the hard figures flow into an Excel model. This provides the fundamental basis PEG ratio, RPO development.
Bring the results into Gemini or Copilot and then check them:
What is the current news situation since the call?
So here too, have the theses of one model checked by another in order to avoid confirmation errors.
•
11
•@Tenbagger2024 Yes, using the block mode is very important.
••