Run a valuation
The Damodaran story-to-numbers framing, your thesis, and the framing questions.
When you start a new analysis, dbot works out the framing with you first. It captures your thesis and turns it into a few questions that the rest of the analysis is built to answer. This part matters more than anything else you do.
Your thesis is a short, testable claim about where the business is going and why. It's the story behind the numbers. A good one names a competitive advantage, says how that shows up in the financials, what it means for value, and the main risk that could break it.
You can write your own thesis or let dbot draft one for the ticker and edit it from there.
dbot won't just take your thesis at face value, and it won't always agree with it. Your thesis and the framing questions ground the research and point it at what actually matters for this company. dbot then tests them, and what guides the valuation in the end is the conclusions the research comes back with, not the thesis you started with. If the evidence cuts against your thesis, the numbers follow the evidence.
From your thesis, dbot writes a handful of questions, the ones that actually decide the answer. They line up with Damodaran's value drivers:
Every research step answers these with cited evidence, and the quality check holds the final valuation up against them. This is what keeps a report focused instead of just describing the company.
By default you don't pick a valuation method. After the research, a plan step looks at what the company actually is and decides how to value it: a single discounted cash flow for most companies, a segment-driven DCF for businesses with genuinely different divisions, a sum of the parts for conglomerates, or a dividend discount model for banks and insurers, where free cash flow is hard to pin down. Every shape shares the same research and finishing steps; what changes is how the company is modeled in the middle. If you do have a strong view, the review step lets you set a method preference that dbot treats as a strong default.
dbot estimates intrinsic value and shows its work: the story, the numbers, the counter-case, and the assumptions the market is implying. It won't issue buy or sell calls, and it ends every report with a few specific things to watch.