Use this skill after the target has a financial snapshot and enough evidence to support assumptions. The goal is not a single price target; it is a transparent range of outcomes that shows which drivers matter most and where evidence is weak.
Inputs
ResearchTargetfromresearch-target-resolver- financial snapshot with period labels, currency, and source links
- peer comparison or historical multiple context when available
- thesis, risks, or catalysts that affect assumptions
- current market price and timestamp when price-implied return is requested
Workflow
- Load the target profile, financial evidence, peer context, and current market data if needed.
- Choose the valuation method that fits the business model and data quality: multiples, DCF, sum-of-the-parts, asset value, or probability-weighted event analysis.
- Define the key drivers before calculating outputs: revenue growth, margin, capital intensity, reinvestment, terminal assumptions, share count, net debt, and segment mix where relevant.
- Build bear, base, and bull cases with explicit assumptions and evidence links.
- Run sensitivity checks on the few assumptions that most affect value.
- Compare scenario value ranges with market price only when price data is fresh enough for the user's request.
- Save the model note under
research/targets/<target>/artifacts/valuation/.
Read references/valuation-framework.md before building the model.
Output
Return:
- valuation method and why it fits the target
- assumption table with sources and confidence labels
- bear, base, and bull scenario outputs
- sensitivity table for the most important drivers
- market-implied expectations when current price is used
- data gaps, stale inputs, and assumptions that need manual review
- archive files created or updated
- handoffs to thesis, risk, and watchlist skills
Quality Gate
Before finishing:
- do not present a target price as a recommendation
- state currency, share count basis, enterprise value adjustments, and data timestamps
- separate sourced inputs from agent assumptions
- avoid false precision; round outputs to a level supported by the inputs
- label stale or missing market prices, estimates, and peer multiples
- explain which assumptions drive most of the valuation range