1,000
Loans
Individual credit facilities
50
Funds
Across 10 vintage years
180
Months
15 years of time series
40
Quarters
Per fund NAV history
Loan-Level Data
Simulated using Ornstein-Uhlenbeck for spreads, Cox-Ingersoll-Ross for default intensity, and Beta distribution for recovery rates.
loans_characteristics.csv
| Field | Type | Description |
|---|---|---|
| loan_id | Integer | Unique loan identifier |
| principal | Float | Loan principal amount ($) |
| ltv | Float | Loan-to-value ratio (0.6-0.9) |
| tenor_years | Integer | Loan maturity (3-10 years) |
| sector | String | Industry sector (10 categories) |
| seniority | String | Debt seniority level |
| initial_rating | String | Initial credit rating (AAA-D) |
| initial_spread | Float | Initial credit spread (bps) |
loans_spreads_timeseries.csv
| Field | Type | Description |
|---|---|---|
| loan_id | Integer | Loan identifier |
| month | Integer | Month number (1-180) |
| spread | Float | Credit spread at month t |
| rating | String | Credit rating at month t |
| default_intensity | Float | Instantaneous default probability |
loans_credit_events.csv
| Field | Type | Description |
|---|---|---|
| loan_id | Integer | Loan identifier |
| event_type | String | Event type (default, prepay, maturity) |
| event_month | Integer | Month when event occurred |
| recovery_rate | Float | Recovery rate if default (Beta dist) |
| loss_amount | Float | Loss given default ($) |
Fund-Level Data
Simulated with J-curve pattern for NAV evolution, vintage effects, and realistic cash flow dynamics.
funds_characteristics.csv
| Field | Type | Description |
|---|---|---|
| fund_id | Integer | Unique fund identifier |
| fund_name | String | Fund name |
| vintage_year | Integer | Fund vintage (2010-2019) |
| commitment | Float | Total commitment ($) |
| strategy | String | Investment strategy (5 types) |
| geography | String | Geographic focus (4 regions) |
| management_fee | Float | Annual management fee (%) |
| carried_interest | Float | Carry rate (typically 20%) |
| final_tvpi | Float | Final Total Value to Paid-In |
| final_dpi | Float | Final Distributions to Paid-In |
| irr | Float | Internal Rate of Return (%) |
funds_cashflows_timeseries.csv
| Field | Type | Description |
|---|---|---|
| fund_id | Integer | Fund identifier |
| quarter | Integer | Quarter number (1-40) |
| contributions | Float | Capital calls ($) |
| distributions | Float | Distributions to LPs ($) |
| nav | Float | Net Asset Value ($) |
| tvpi | Float | TVPI at quarter end |
| dpi | Float | DPI at quarter end |
| rvpi | Float | RVPI at quarter end |
Mathematical Models
Ornstein-Uhlenbeck (Credit Spreads)
drt = theta(mu - rt)dt + sigma dWt
theta=0.5, mu=3%, sigma=1%
Cox-Ingersoll-Ross (Default Intensity)
d lambdat = kappa(theta - lambdat)dt + sigma sqrt(lambdat) dWt
kappa=0.3, theta=2%, sigma=5%
Beta Distribution (Recovery Rates)
R ~ Beta(alpha, beta)
alpha=2.0, beta=3.0 (mean ~40%)
J-Curve (Fund NAV)
TVPI(t) = 1 - depth * (t/trough) for t <= trough
Trough at Q8, final TVPI: 1.3-2.0x
Download Data
Generation Statistics
Loan Statistics
- 1,000 loans generated
- 180 months (15 years)
- 288 defaults (28.8% cumulative)
- Average recovery: 40.25%
- 10 industry sectors
- 4 seniority levels
Fund Statistics
- 50 funds across 10 vintages
- 40 quarters per fund (10 years)
- Average TVPI: 2.72x
- Median TVPI: 2.61x
- 5 investment strategies
- 4 geographic regions