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_idIntegerUnique loan identifier
principalFloatLoan principal amount ($)
ltvFloatLoan-to-value ratio (0.6-0.9)
tenor_yearsIntegerLoan maturity (3-10 years)
sectorStringIndustry sector (10 categories)
seniorityStringDebt seniority level
initial_ratingStringInitial credit rating (AAA-D)
initial_spreadFloatInitial credit spread (bps)

loans_spreads_timeseries.csv

Field Type Description
loan_idIntegerLoan identifier
monthIntegerMonth number (1-180)
spreadFloatCredit spread at month t
ratingStringCredit rating at month t
default_intensityFloatInstantaneous default probability

loans_credit_events.csv

Field Type Description
loan_idIntegerLoan identifier
event_typeStringEvent type (default, prepay, maturity)
event_monthIntegerMonth when event occurred
recovery_rateFloatRecovery rate if default (Beta dist)
loss_amountFloatLoss 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_idIntegerUnique fund identifier
fund_nameStringFund name
vintage_yearIntegerFund vintage (2010-2019)
commitmentFloatTotal commitment ($)
strategyStringInvestment strategy (5 types)
geographyStringGeographic focus (4 regions)
management_feeFloatAnnual management fee (%)
carried_interestFloatCarry rate (typically 20%)
final_tvpiFloatFinal Total Value to Paid-In
final_dpiFloatFinal Distributions to Paid-In
irrFloatInternal Rate of Return (%)

funds_cashflows_timeseries.csv

Field Type Description
fund_idIntegerFund identifier
quarterIntegerQuarter number (1-40)
contributionsFloatCapital calls ($)
distributionsFloatDistributions to LPs ($)
navFloatNet Asset Value ($)
tvpiFloatTVPI at quarter end
dpiFloatDPI at quarter end
rvpiFloatRVPI 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