Every private fund pitch shows performance in a handful of acronyms: IRR, MOIC, DPI and TVPI. Each measures something real, and each can be made to flatter. Reading them together, and knowing what any one of them conceals, is one of the most useful skills an allocator can build.
The trap is to fixate on a single number, usually the IRR, because it is the largest and most quotable. The discipline is to read all four as a set, because they only tell the truth in combination.
What each one measures
MOIC, the multiple on invested capital, is the simplest: total value divided by capital in. A 2.0x means the money doubled, ignoring time. TVPI, total value to paid-in, is the fund-level version of the same idea, counting both distributed and still-held value against what LPs paid in. DPI, distributions to paid-in, counts only cash actually returned: real money in your pocket. IRR, the internal rate of return, is the annualised, time-weighted return, which rewards getting money back sooner.
The relationship between DPI and TVPI matters most. TVPI includes unrealised value the GP has marked itself. DPI is only what has been paid out. A fund with a high TVPI but low DPI is showing you paper gains it has not yet turned into cash, and the gap between the two is a measure of how much you are trusting the GP's own valuations.
How IRR can mislead
Because IRR rewards speed, it can be inflated by timing tricks. A quick early win, or the use of a subscription credit line that delays calling LP capital, can lift the reported IRR without improving how much money LPs ultimately make. A high IRR paired with a modest MOIC often means the gains were fast but small.
The honest read is to pair IRR with MOIC and DPI. A strong IRR that has actually returned cash, high DPI, is real. A strong IRR sitting on unrealised marks and financial engineering is a claim, not a result. Ask always: how much of this has actually come back.
The quartile game and its limits
Funds are ranked against peers of the same type and vintage year into quartiles, and top-quartile status is the industry's badge of quality. It is useful, because manager dispersion in private markets is huge, but it is also gameable: benchmarks differ, vintages are defined loosely, and a fund can be top-quartile on one metric and mediocre on another.
Use quartiles as a screen, not a verdict. The right comparison is against the correct vintage and strategy, on the metric that matters for your goal, cash returned for income, total multiple for growth, and always with an eye on how much of the record is realised versus marked.