Data Composition and Sources
PitchBook gathers fund returns data primarily from individual LP reports, providing a detailed baseline for estimating fund activity. Return profiles can vary among LPs due to factors like fee discounts and timing of commitments. The granularity of LP-reported returns on the PitchBook Platform offers valuable insights, although discrepancies in calculating fund-level returns must be addressed. To be included in pooled calculations, a fund needs at least one LP report within two years of its vintage and reports in at least 45% of applicable periods. Benchmarks determine fund returns based on all available LP reports in a given period, using straight-line interpolation for missing data in approximately 10% of reporting periods.
Expansion of Dataset and Inclusion Criteria
Starting from PitchBook Benchmarks as of Q4 2019, the dataset has been expanded to include funds with reported IRRs even if they do not meet pooled calculation criteria. In the Q2 2021 report, additional improvements were made to the inclusion criteria for reported IRRs, leading to some changes in vintage year data counts compared to previous iterations. Reporting lag and fund liquidation necessitate the extension of cash multiples and IRR information in certain cases. Consistency is maintained across editions, but fund classifications may change as new data is gathered and new funds are added to the dataset.
Definitions and Calculation Methodologies
Various definitions and calculation methodologies are used in the PitchBook report methodologies. Fund counts indicate the number of funds in a sample, where some funds may have reported IRRs but lack sufficient cash flow information for pooled calculations. Median calculations show the middle data point in a sample group. Standard deviation measures the dispersion of reported IRRs for a peer group. Vintage year represents the year of a fund's first investment. IRR reflects the discount rate of historical cash flows to achieve a net present value of zero. Horizon IRR is a capital-weighted calculation showing IRR over a specific time range, and DPI and RVPI multiples measure distributions and unrealized returns, respectively.