DCF Valuation Model Builder
Build auditable DCF valuation models with WACC, scenarios, and sensitivity analysis.
The skill requires staged user confirmation, source comments for hardcoded values, and formula-driven models, while the repository states that outputs require professional review. However, it does not clearly control data flows for MCP, web, or local Excel data, sensitive financial-data handling, least privilege, rollback, or dependency security, so points are deducted.
The documentation covers the DCF workflow, formula rules, checks, and troubleshooting. However, the referenced recalc.py is absent from the supplied files, no key-path test suite is provided, and the Office JS guidance, IF-based scenario logic, and later INDEX-consolidation guidance are inconsistent. Edge-case handling and failure feedback remain incomplete. Static calibration caps this score at 10.
The intended users, DCF valuation scenario, input sources, and principal outputs are reasonably clear, and Office JS versus standalone xlsx environments are distinguished. Non-fit boundaries, exclusion triggers, Chinese-language support, and mainland-China reachability are not addressed, while core data retrieval may depend on external MCP and web services, so points are deducted.
The material has a useful structure with overview, constraints, workflow, correct and incorrect patterns, troubleshooting, and a validation script. The repository supplies an Apache-2.0 license and contribution guidance. However, the skill lacks its own version, changelog, named maintenance owner, dependency installation notes, and the referenced script is missing from the supplied evidence.
The skill clearly targets directly usable Excel DCF models, sensitivity analysis, and valuation summaries, with concrete formula, scenario, and output instructions. No generated model or statically verifiable representative result is supplied, implementation remains dependent on user execution, and documentation inconsistencies require review. Static calibration caps this score at 7.
Formula templates, a validation script, and repository-level plugin-validation workflows provide some auditable evidence. There are no committed key-path tests, real execution records, or independent third-party corroboration for the skill, and the validator is not shown to establish core modeling correctness. Static calibration caps this score at 5.
- Before delivery, confirm that recalc.py and all dependencies actually exist, and test formulas, scenario switching, sensitivity tables, and terminal-value constraints.
- Do not treat MCP or web-sourced data as verified facts; independently check sources, dates, units, diluted shares, net debt, and terminal assumptions.
- Add controls for sensitive financial data, permission scope, recovery after failure, and fallback procedures when services are unreachable from mainland China.
What does this skill do, and when should you use it?
This skill builds equity valuation models using discounted cash flow methodology and institutional-style Excel conventions. It can source data from SEC filings, analyst reports, user-provided inputs, web sources, and available MCP servers, then construct historical analysis, operating projections, free cash flow, WACC, terminal value, and the enterprise-to-equity bridge. The model includes Bear, Base, and Bull cases plus three sensitivity-analysis grids on the DCF sheet. The workflow requires user confirmation after data retrieval, revenue projections, free cash flow, WACC, terminal value, and the equity bridge rather than assuming an end-to-end build without checkpoints.
Reads user inputs, MCP-sourced data, SEC filings, analyst reports, and necessary web market data; validates net debt or net cash, diluted shares, historical margins, growth rates, and tax rates; analyzes three to five years of historical revenue, margins, capital intensity, working capital, and return metrics; builds Bear, Base, and Bull revenue, operating-expense, EBIT, and free-cash-flow projections; calculates WACC using CAPM, cost of debt, and capital-structure weights; discounts forecast cash flows and terminal value using a mid-year convention; creates the enterprise-value-to-equity-value and implied-share-price bridge; and produces an Excel model with DCF and WACC sheets, with sensitivity grids at the bottom of the DCF sheet for WACC versus terminal growth, revenue growth versus EBIT margin, and beta versus the risk-free rate.
- An investment-banking or equity-research professional needs an auditable five-year DCF model based on company financials and market data.
- An investor wants to compare intrinsic value and implied return under Bear, Base, and Bull growth and margin assumptions.
- A financial analyst needs to combine SEC filings, analyst estimates, WACC inputs, and terminal-value assumptions in one Excel valuation file.
- A model reviewer needs to check net debt, diluted shares, capital structure, terminal-value concentration, and sensitivity outputs.
- A user needs a flexible deliverable in which changing assumptions drives Excel recalculation rather than receiving static values precomputed in Python.
What are this skill's strengths and limitations?
- Covers the full DCF workflow from data validation through implied per-share value.
- Requires live Excel formulas for projections, margins, discounting, valuation, and sensitivity cells instead of static calculated hardcodes.
- Includes Bear, Base, and Bull scenarios, WACC calculation, mid-year discounting, and three sensitivity-analysis tables.
- Specifies input comments, formula-error checks, colors, borders, number formats, and sheet architecture for reviewability and auditability.
- Supports both Excel Office JS execution and standalone XLSX generation.
- The quality and availability of inputs depend on access to MCP providers, web sources, SEC filings, or analyst reports in the execution environment.
- The source provides no standalone automated test suite or verified sample output.
- The required stage-by-stage confirmations make it less suitable for users who expect an unattended end-to-end build.
- The standalone XLSX path requires Python, openpyxl, filesystem access, and the prescribed recalculation script; the Office JS path requires a live Excel environment.
- The supplied skill file is explicitly truncated, so unseen sections remain unknown.
How do you install this skill?
The README documents two collection-level installation paths. In Cowork, go to Settings → Plugins → Add plugin and paste https://github.com/anthropics/financial-services. In Claude Code, run claude plugin marketplace add anthropics/financial-services, then claude plugin install financial-analysis@claude-for-financial-services. The README also identifies Model Builder as a named agent containing DCF, LBO, three-statement, and comps modeling. The source does not provide a standalone installation command for dcf-model; the skill is located at plugins/agent-plugins/model-builder/skills/dcf-model/.
How do you use this skill?
After installing the relevant plugin, use a request such as: “Build a five-year DCF model for [company name or ticker], using available SEC filings and market data, and show Bear, Base, and Bull cases.” Provide a company name or ticker and revenue-growth assumptions, or explicitly allow consensus estimates; projection period, terminal growth, and specific WACC inputs are optional. Confirm each stage of the workflow: retrieved inputs, revenue projections, free-cash-flow schedule, WACC, terminal value and equity bridge, and sensitivity tables. For a standalone XLSX file, the skill requires Python/openpyxl and the command python recalc.py [path_to_excel_file] [timeout_seconds]; all formula errors must be fixed before delivery.
How does this skill compare with similar options?
The source explicitly distinguishes two implementation paths: use Office JS inside an Excel Office Add-in or Office JS environment, and use Python/openpyxl for a standalone XLSX file followed by the recalculation script. Both paths are expected to follow the same formula, commenting, layout, and sensitivity-analysis rules.