Your AI Quant for Structured Finance
Graam analyzes deals, builds forecasts, and generates cashflows automatically. Give it a prospectus and loan tape—get complete analysis in minutes.
What Graam Can Do
Institutional-grade structured finance analysis through natural language
Collateral Analysis
Stratify loan tapes by FICO, LTV, vintage, DTI—instantly. Compare deals side-by-side and flag outliers or concentration risks.
“Compare the FICO distribution between @eart213 and @eart253”
Loss Forecasting
Build default and prepayment models from vintage curves or peer comps. Adjust assumptions and project lifetime losses.
“Estimate 48-month cumulative losses for @eart253 using 2021-2023 vintage data”
Model Validation
Back-test predictions against trustee remittance data. Generate actual vs. expected charts and error metrics.
“How does my default model perform against actual trustee data for @eart213?”
Waterfall & Cashflow
Parse prospectus terms to auto-build a waterfall. Run cashflow scenarios and compute tranche-level yields and coverage tests.
“Build the waterfall for @eart253 and project Class B IRR under 2x loss stress”
Interactive Analysis
Work in notebooks where every output is editable and re-runnable. Explore what-if scenarios, refine charts, and export reports—all in natural language.
“Now adjust the CDR assumption to 3% and show me how that changes the loss forecast”
Data Sources
Institutional-grade data from trusted sources
ECONOMIC DATA | ST. LOUIS FED
Federal Reserve Economic Data
U.S. SECURITIES AND
EXCHANGE COMMISSION
Official SEC filings (EDGAR)
BLS
Bureau of Labor Statistics
Market Data
Stocks, options, interest rates, earnings transcripts, 10-K filings
Your Data
Upload custom datasets, spreadsheets, and proprietary data sources