I'm a retired tech executive who spent the last year building a personal research tool to help with my own due diligence. It combines two very different data streams: 5 years of historical financial data scored quantitatively, and daily financial press analysis processed through a systematic sentiment engine.
I've finally gotten the system operational with multi-brokerage sync, but I've run into a philosophical problem regarding the final rating it gives a stock.
Here is a real example from my current dashboard:
Fundamental Score: 65/100 (solid but slowing margins, decent debt-to-equity)
Sentiment Score: 85/100 (strongly positive outlook on a new product line and management restructuring)
My instinct was to combine these into a single weighted number — say 70% fundamentals / 30% sentiment. But a conversation in my intro thread made me question whether that's even the right goal. Maybe the output shouldn't be one rating but two, with a flag when they significantly diverge — forcing the investor to think rather than just follow a number.
So my question for the veterans here:
-
Should these ever be combined into a single score, or does collapsing them destroy the signal?
-
If you do combine them, how do you weight fundamentals vs. sentiment in a value framework?
-
Is a large divergence between the two more valuable as a warning flag than either score alone?
I'm not looking to automate the decision — I'm trying to build a tool that removes the heavy lifting while keeping the human judgment in the equation. Happy to share more detail on how either score is generated if useful.