financial-analyst-mcp

Institutional-grade financial modeling for AI agents — deterministic engines reconciled to Excel.

От сообщества: Добавлен пользователем или импортирован; проверьте владельца перед подключениемРаботаетНужен API-ключГлобальныйБесплатноМожет изменять данные

Что умеет

  • Lbo.Model: Model a leveraged buyout (LBO) on a full 3-statement basis, from sources & uses through exit. Reconciled to Excel — see the /lbo/reconciliation attestation. Senior debt is sized by the mini
  • Waterfall.Distribute: LP/GP waterfall distribution across multiple periods — penny-accurate. Distributes cash in strict order: 1. Return of capital (pro-rata LP/GP by contribution) 2. Preferred return
  • Re.Multifamily.Underwrite: Multifamily acquisition underwriting proforma. Builds a year-by-year model: GPR → vacancy → EGI → OpEx → NOI → debt service → cash flow → exit. Tracks DSCR, CoC, LTV, cap ra

Какие данные видит

Нужен ли аккаунт

Нужен API-ключ из настроек сервиса

Institutional-grade financial modeling for AI agents — deterministic engines reconciled to Excel.

Corporate & PE: LBO modeling (sources & uses through exit, IRR/MOIC, sensitivity tables), DCF valuation, LP/GP waterfall distributions (multi-tier promote, IRR/MOIC hurdles, selectable day-count), XIRR on irregular cash flows, debt sizing (min of LTV/DSCR/LLCR), amortization schedules, Monte Carlo simulation, and FX-adjusted P&L.

Real estate underwriting: multifamily acquisition and ground-up development, hotel acquisition and ground-up development, industrial acquisition and ground-up development, SFR, short-term/vacation rental (STR), and fix-and-flip — monthly proformas to unlevered and levered project returns.

Each engine is checked against a SHA-pinned Excel workbook through a served parity process, with the results queryable at GET /{engine}/reconciliation. Deterministic and formula-traceable — no LLM in the math.

$0.25–$5.00 per call. Pay per call with USDC via x402 (no signup), or get an API key at financial-analyst.ai/keys/create.

Список инструментов сервера (17)

Технические названия из tools/list. Нужны только разработчикам.

lbo.modelModel a leveraged buyout (LBO) on a full 3-statement basis, from sources & uses through exit. Reconciled to Excel — see the /lbo/reconciliation attestation. Senior debt is sized by the minimum of the leverage, interest-coverage, and fixed-charge-coverage (DSCR) tests; mezzanine plugs senior up to the total leverage multiple. The annual model runs four facilities (senior, mezz, capex, revolver) with scheduled amortisation, cash sweep, origination fees and interest; taxes are levered (on EBIT - interest, with NOL carryforward); un-swept cash accumulates and the exit uses NET debt. Returns are computed for each exit year, and the equity gain is decomposed into value-creation drivers. All monetary inputs are in consistent currency units (e.g. GBP or USD millions). WHEN TO USE: - PE deal screening and IC-grade returns analysis - Independent sponsor assessment; DCM leverage/debt-capacity sizing - Comparing entry/exit multiples, leverage, and hold-period assumptions - Value-creation attribution for an IC memo (growth vs multiple vs deleveraging) OUTPUTS: - Sources & uses (senior sized by the binding test, mezz plug, equity check) - Returns for each exit year: IRR, MOIC, equity value - Value-creation attribution: EBITDA growth, multiple expansion, debt paydown, cash generation, transaction costs, management incentive - Full annual model: revenue, EBITDA, EBIT, interest, taxes, debt balances, cash COST: $5.00 per call (5 API key credits).
waterfall.distributeLP/GP waterfall distribution across multiple periods — penny-accurate. Distributes cash in strict order: 1. Return of capital (pro-rata LP/GP by contribution) 2. Preferred return (cumulative, non-compounding, ACT/365 on unreturned LP capital) 3. GP catch-up (100% to GP until GP = catchup_pct × LP preferred paid) 4. Promote tiers (1–5 tiers with IRR or MOIC hurdles) Uses Python Decimal throughout. ACT/365 day count matches Excel. TIERS STRUCTURE: hurdle_type applies to ALL tiers. Set the last tier's hurdle to 99.0 to capture all remaining distributions above the prior tier. Example: [{"hurdle": 0.15, "lp_pct": 0.80, "gp_pct": 0.20}, {"hurdle": 0.18, "lp_pct": 0.75, "gp_pct": 0.25}, {"hurdle": 99.0, "lp_pct": 0.70, "gp_pct": 0.30}] WHEN TO USE: - PE fund waterfall modeling and LP reporting - Real estate promote / carried interest calculations - Fund admin tools and carry verification - Checking GP catch-up mechanics against fund documents
re.multifamily.underwriteMultifamily acquisition underwriting proforma. Builds a year-by-year model: GPR → vacancy → EGI → OpEx → NOI → debt service → cash flow → exit. Tracks DSCR, CoC, LTV, cap rate annually. NOTE: expense_ratio is applied to GPR (not EGI), consistent with how fixed costs like taxes, insurance, and reserves behave in practice. WHEN TO USE: - Underwriting a multifamily acquisition for a lender or equity sponsor - Screening deals against CoC/DSCR/IRR thresholds - Sizing debt on a multifamily asset - Comparing entry vs exit cap rate assumptions
re.sfr.underwriteSingle-family rental (SFR/DSCR) investment underwriting proforma. Monthly cash flow model — debt service, rent, and expenses all computed monthly then aggregated annually for exact IRR reconciliation. Tracks CoC, DSCR, LTV, and cap rate annually as debt amortizes. Supports DSCR loans and bridge-to-perm via interest-only periods. IMPORTANT: monthly_expenses is a MONTHLY figure (taxes + insurance + maintenance + property management combined per month), not annual. WHEN TO USE: - Underwriting an SFR buy-and-hold investment - DSCR loan qualification analysis (tracks DSCR year-by-year) - Comparing SFR to multifamily or STR economics - Screening single-family rental portfolios
re.str.underwriteShort-term rental (STR/Airbnb) investment underwriting with GO/NO-GO verdict. Revenue is modeled from quarterly ADR and occupancy assumptions per quarter. Each quarter is a structured object with adr, occupancy, and days. Operating expenses are itemized — mgmt fee, platform fee, cleaning, supplies, insurance, property tax, HOA, utilities, repairs, and other. VERDICT LOGIC: - GO if Year 1 CoC >= coc_threshold (primary signal) - GO if IRR >= irr_threshold AND MOIC >= moic_threshold (override) - NO-GO otherwise WHEN TO USE: - Evaluating a vacation rental or Airbnb investment - Screening STR opportunities against CoC/IRR thresholds - Comparing STR vs long-term rental (SFR) economics - Underwriting a short-term rental portfolio acquisition
re.fixflip.underwriteFix & flip underwriting — solves backwards from desired profit to max purchase price. ARV = equal-weighted average of 2–5 comparable sales. Loan = min(arv_cap_pct × ARV, ltc_cap_pct × (purchase + repairs)). Output identifies which constraint is binding. WHEN TO USE: - Evaluating a flip before making an offer - Hard money lender underwriting a rehab loan - Solving for max purchase price given a profit target - Comparing repair cost scenarios
xirr.computeCompute annualized IRR (XIRR) over irregular cash flow periods. Handles any mix of daily, monthly, quarterly, or annual intervals. No assumption of regular periods — uses actual dates. WHEN TO USE: - PE fund IRR on actual distribution dates - Real estate returns with irregular cash flows - Project finance draw-down and distribution schedules - Venture investment IRR from funding rounds to exit - Any return calculation where Excel XIRR would apply OUTPUTS: Annualized IRR, MOIC, net profit, hold period, cash flow summary. COST: $0.25 per call (1 API key credit).
amortization.scheduleFull amortization schedule with summary statistics and milestones. Matches Excel PMT exactly: monthly rate = annual rate / 12, payment rounded to nearest cent. Interest at full precision, balance rounded to cents each period. WHEN TO USE: - Debt service modeling on any fixed-rate loan - Bridge or construction loan with IO period before amortization - Comparing loan structures (rate, term, prepayment scenarios) - Calculating total interest cost and payoff dates
montecarlo.simulateMonte Carlo simulation with correlated variables. Up to 100,000 trials. Define uncertain inputs with probability distributions, write a formula, receive a full statistical distribution of outcomes. Supports correlated normal/lognormal variables via Cholesky decomposition — captures real-world relationships like cost/revenue correlation or demand/price elasticity.
fx.pnlFX-adjusted P&L — decomposes total return into asset vs currency components. Answers: how much of my return came from the asset itself vs. exchange rate movement? All FX rates are caller-supplied — fully deterministic arithmetic, fully auditable. No market data dependency. WHEN TO USE: - International real estate (buy GBP, carry in GBP and USD, sell GBP) - Cross-border PE investments (invest EUR, exit USD) - Any investment where currency movement is material to total return - Attributing performance between asset manager (local) and FX (currency)
dcf.valueDCF (Discounted Cash Flow) valuation with Exit Multiple or Gordon Growth terminal value. End-of-year cash flow convention. Two terminal value methods: - "exit_multiple" (default): terminal EBITDA × EV/EBITDA multiple. Standard in PE and M&A. Requires terminal_ebitda and exit_multiple. - "gordon_growth": FCF_n × (1+g) / (WACC - g). Standard in equity research. Requires terminal_growth_rate; WACC must exceed terminal_growth_rate. WHEN TO USE: - Equity research intrinsic value - M&A target valuation - PE portfolio company valuation - Capital budgeting / project NPV OUTPUTS: Per-year FCF, discount factor, and PV; terminal value (undiscounted and PV) and its % of enterprise value (sanity check — typically 60-80%); enterprise value, equity value, implied share price; 9×9 sensitivity matrix (EV across WACC ± 200bps vs exit multiple ± 2x or terminal growth ± 100bps). COST: $1.00 per call (1 API key credit).
debtsizing.sizeDebt sizing across three asset-class methodologies. Sizes to the binding (minimum) of each class's applicable constraints — runs every leverage/ coverage/collateral test and lends to the tightest. Set asset_class to "cre", "private_equity", or "project_finance" and supply that class's required fields (others can be omitted). CRE — requires: noi, interest_rate, amortization_months, dscr, ltv, cap_rate, debt_yield - DSCR Constraint — annuity-based sizing off target DSCR (Excel PV() convention) - LTV Constraint — (NOI / Cap Rate) x LTV - Debt Yield Constraint — NOI / Debt Yield Private Equity — requires: ebitda, cfads, interest_rate, amortization_months, leverage_multiple, min_interest_coverage, dscr - Leverage Constraint — EBITDA x Leverage Multiple - Interest Coverage Constraint — (EBITDA / Min Interest Coverage) / Rate - DSCR Constraint — CFADS-based annuity sizing off minimum DSCR Project Finance — requires: construction_cost, cfads, interest_rate, debt_term_years, min_llcr, dscr, ltv - Loan-to-Cost Constraint — Construction Cost x Max LTV - DSCR Constraint — CFADS-based annuity sizing off minimum DSCR - LLCR Constraint — PV of CFADS stream off minimum LLCR OUTPUTS: binding constraint and loan amount, all three constraint values, and implied metrics at the sized loan (e.g. implied DSCR/LTV/leverage/LLCR actually achieved at that loan amount). COST: $0.25 per call (1 API key credit).
re.hotel.underwriteHotel acquisition underwriting — stabilized or light value-add / PIP repositioning. Bridge-to-perm structure: acquisition + optional PIP funded by a construction/bridge loan, refinanced into a permanent loan sized on stabilized NOI at month cl_term_months. RevPAR built from quarterly ADR × occupancy, plus other operated revenue and USALI OpEx. WHEN TO USE: - Underwriting a hotel acquisition (full- or select-service) - Modeling a PIP / brand-conversion reposition with a bridge loan - Sizing perm debt off a stabilized cap rate with a DSCR floor - Screening on DSCR / cap rate / IRR / MOIC
re.hotel.developHotel ground-up development underwriting. Full construction budget (land, hard + pre-opening, contingencies, capitalized interest, developer fee) funded by a construction loan, refinanced into a permanent loan at month cl_term_months. Quarterly ADR × occupancy ramp with a stabilized exit. WHEN TO USE: - Underwriting ground-up hotel development - Sizing a construction loan on blended LTC and capitalized interest - Testing yield-on-cost vs exit cap rate spread - Screening development returns (IRR / MOIC) with a stabilization ramp
re.industrial.underwriteIndustrial / warehouse acquisition underwriting with a per-tenant rent roll. Bridge-to-perm structure: acquisition + optional capex funded by a bridge/construction loan, refinanced into a permanent loan sized on stabilized NOI at month cl_term_months. Tracks lease-level rent, escalators, TI/LC rollover reserves and NNN OpEx. WHEN TO USE: - Underwriting a stabilized or light value-add industrial acquisition - Modeling staggered lease commencements / rollover - Sizing perm debt off a stabilized cap rate with a DSCR floor - Screening on DSCR / cap rate / IRR / MOIC and $/SF metrics
re.industrial.developIndustrial / warehouse ground-up development underwriting with a per-tenant rent roll. Construction loan on blended LTC (land + hard + soft + dev fee + capitalized interest), refinanced into a permanent loan at month cl_term_months. Lease-up from first occupancy with TI/LC reserves, NNN OpEx, and a stabilized exit. WHEN TO USE: - Underwriting ground-up industrial/logistics development - Sizing a construction loan on blended LTC and an s-curve draw - Modeling pre-leasing with staggered commencements - Testing yield-on-cost vs exit cap spread; screening IRR / MOIC
re.multifamily.developMultifamily ground-up development underwriting. Construction loan on blended LTC (land + hard + soft + capitalized interest), refinanced into a permanent loan sized as min(LTV, DSCR) on stabilized NOI / stabilized_cap_rate. Lease-up from C/O with a stabilized exit. WHEN TO USE: - Underwriting ground-up multifamily development - Sizing a construction loan on blended LTC and an s-curve draw - Testing yield-on-cost vs exit cap spread - Screening development IRR / MOIC with a lease-up ramp
financial-analyst-mcp: подключить к Claude, ChatGPT, Cursor · Connectors.fun