我終於找到了這個帖 ---- 這是當時在大千討論公司FA基本麵時發的一個X的post

我覺得這個X的post非常接近我認同的投資理念。雖然我們很難做到他說的10點，但培養這樣的approach對投資會有極大的幫助。

他用了NVDA做例子，來解釋如何build FA assumption，又如何去驗證。

This is an example of what a valid thesis looks like:

• Blackwell is the best product launch and will generate $200BN+ in revenues in '25 (v.s Street at $170BN)
• Scaling laws will hold post '25, driving performance by x percentage on a 1-year cadence.

Key is that the thesis is specific enough to track if it's playing out.

This is an example of what is NOT a valid thesis for NVDA

• Nvidia is a great company
• AI is the future and Nvidia is a winner
• Jensen is awesome

原文在下。

https://x.com/IvanaSpear/status/1858910469219905969