At any point before full realization, the fund reports a Net Asset Value (NAV), the current fair value of everything it still holds, plus any uncalled/undistributed cash. Under ASC 820 (US GAAP fair value measurement) and the IPEV (International Private Equity and Venture Capital Valuation) Guidelines, an unrealized portfolio company is marked using the same tools you already own from Phase 3, a DCF, a comps multiple, a recent-transaction reference, with an added illiquidity lens (M3.10). This is a real, disciplined valuation exercise, audited annually, and it is still an estimate made by the party whose own fee-and-carry economics benefit from a higher mark. That conflict of interest is structural, not a moral failing of any specific GP, and every serious LP diligence process interrogates the marking policy (does the GP mark down promptly on bad news, or only at exit?) for exactly this reason. Hold the distinction sharply: distributions are cash the LP has already banked; NAV is a mark the LP is trusting. Every multiple this module defines next inherits that split.
Advertisement
The multiples: PIC, DPI, RVPI, TVPI, MOIC: and Gross vs Net
Four ratios, all denominated against Paid-In Capital (PIC), the cumulative capital actually called from LPs to date (not the full commitment, which may be only partly drawn):
- DPI (Distributions to Paid-In) = cumulative distributions ÷ PIC. The realized, cash-on-cash multiple, the only one of the four that is money already in the LP's account.
- RVPI (Residual Value to Paid-In) = current NAV ÷ PIC. The unrealized multiple, entirely a fair-value mark, as the NAV discussion explains.
- TVPI (Total Value to Paid-In) = DPI + RVPI = (cumulative distributions + NAV) ÷ PIC. The total value multiple, realized plus unrealized.
- MOIC (Multiple on Invested Capital): at the fund level this is arithmetically identical to TVPI (both data providers and LPAs use the terms near-interchangeably here); but MOIC is also used loosely at the individual-deal level (a single portfolio company's own proceeds ÷ cost), where "TVPI" has no meaning at all, because TVPI is specifically a fund/vehicle-level aggregate term. Confusing the two levels is a genuine, common error: flagged again in Common Mistakes.
What is the difference between gross and net multiples?
Gross vs Net modifies every one of these. Gross figures are computed on cash flows before fees and carry, typically the capital actually invested in portfolio companies and the gross proceeds at exit. Net figures are computed on what the LP actually experienced. LP's own paid-in capital (which includes the fee-funding portion of every call) against what the LP actually received after both management fees and carried interest have been deducted. Net is the only number that tells an LP what they actually earned; Gross tells you how the underlying portfolio performed, isolated from the vehicle's economics. Watch for a further wrinkle when comparing across data providers: "Gross" is not perfectly standardized, some providers net out fees but not carry, others net out neither, so a reported "Gross IRR" from one database is not automatically comparable to another's without checking the definition (Common Mistakes, item 6). Worked Example 3 computes all four multiples, Gross and Net, for both running funds, Python-verified.