DPI (distributions to paid-in)

The formula

DPI = cumulative distributions ÷ paid-in capital

It counts only cash that has left the fund and reached the investor. No estimate, no mark, no judgement call — which is exactly why it is the hardest headline number to argue with.

A worked example

The same LP has paid in €3,000,000 and received €1,800,000 in distributions. DPI is 1,800,000 ÷ 3,000,000 = 0.6x: for every euro called, sixty cents have come back. Until DPI passes 1.0x the position has not yet returned the capital it consumed, whatever its TVPI says.

The comparisons people search for

DPI vs TVPI

TVPI counts distributions plus what the manager says the rest is worth. DPI counts only the distributions. TVPI minus DPI is RVPI — the unrealised remainder. A fund with a strong TVPI and a DPI near zero has produced marks, not proceeds.

DPI vs MOIC

MOIC is a total multiple including unrealised value, so it is closer to TVPI than to DPI. A realised MOIC, quoted on exited deals only, is the closer analogue — but it is measured against the capital in those deals, not against everything the LP paid in.

The recallable distribution trap

Some distributions can be called again by the manager. Whether those are netted out of DPI or left in it varies, and the choice can move the number materially in a fund's early years. If two sources disagree on DPI, this is the first place to look.

When does DPI matter most?

Late in a fund's life, and whenever TVPI looks strong. Early on DPI is near zero by construction and says little. As a fund matures it becomes the number that separates realised performance from a valuation opinion.

Worth knowing

• Also called the realisation multiple or cash-on-cash multiple.
• DPI of 1.0x means the LP has been made whole in nominal terms — before any time value.
• Unlike TVPI and IRR it contains no estimate, so it does not move when a mark changes.
• DPI + RVPI = TVPI, on the same fee basis.