MCP Server¶
The Intangible Valuation MCP (Model Context Protocol) server exposes all valuation methods as tools for AI agents.
Overview¶
The MCP server provides AI agents with access to:
- Time value of money calculations
- Discount rate construction (build-up, CAPM, WACC)
- Relief from Royalty valuations
- Purchase Price Allocation
- Monte Carlo simulation
- Decision tree analysis
- Goodwill calculation
- Royalty benchmarking
Setup¶
Installation¶
Configuration¶
Add the MCP server to your AI agent configuration:
Running Standalone¶
Available Tools¶
Time Value of Money¶
| Tool | Description |
|---|---|
present_value |
Calculate PV of a single future cash flow |
future_value |
Calculate FV of a present amount |
annuity_pv |
Calculate PV of an ordinary annuity |
perpetuity_pv |
Calculate PV of a perpetuity |
growing_annuity_pv |
Calculate PV of a growing annuity |
terminal_value |
Calculate terminal value (Gordon Growth or Exit Multiple) |
Discount Rates¶
| Tool | Description |
|---|---|
build_up_discount_rate |
Build-up method with risk premiums |
capm_discount_rate |
CAPM cost of equity |
wacc |
Weighted Average Cost of Capital |
dlom_finnerty |
Finnerty model for lack of marketability |
control_premium |
Control premium calculation |
tax_amortization_benefit |
TAB present value |
Valuation Methods¶
| Tool | Description |
|---|---|
relief_from_royalty |
Relief from Royalty method with TAB |
reproduction_cost |
Depreciated reproduction cost |
replacement_cost |
Depreciated replacement cost |
purchase_price_allocation |
Full PPA waterfall |
goodwill |
Goodwill as residual |
monte_carlo_valuation |
Monte Carlo simulation |
decision_tree_valuation |
Decision tree analysis |
Example Usage¶
An AI agent can call the tools like this:
Tool: relief_from_royalty
Arguments:
revenue_projections: [1000000, 1100000, 1200000, 1300000, 1400000]
royalty_rate: 0.05
discount_rate: 0.12
tax_rate: 0.25
useful_life: 5
Result:
value: 387654.32
method: Relief from Royalty
tab_factor: 1.23
Architecture¶
The MCP server is built with FastMCP and exposes all valuation functions as tools with:
- Full parameter validation via Pydantic
- Rich descriptions for each tool
- Structured JSON responses with calculation steps
- Error handling with detailed messages