All Insights

Performance & Reporting

Benchmarking Private Equity Performance: IRR, TWR, PME, and What LPs Should Actually Compare

A practical LP guide to private market performance measurement: where IRR misleads, when TWR helps, how Public Market Equivalents work, and how to build a benchmarking stack that survives an investment committee.

Matthew S. · Founder, Allocator Desk · June 2026 · 11 min read

Most allocators inherit a benchmarking habit before they ever question it. A pension reports against a vintage-year quartile. An endowment shows IRR next to a public index. A family office prints MOIC and calls it a day. None of these are wrong, but none of them on their own tell you whether a manager actually earned their fee.

This guide walks through the four measures LPs use most, the traps in each, and a sequence for putting them together so the committee conversation moves past "what is our IRR" into "did we get paid for the risk and illiquidity we took."

Why private markets need their own toolkit

Public equity performance is easy to benchmark because cash is fungible and prices are continuous. A manager either beats the index or does not, and the dollar amount you put to work is exactly the dollar amount being measured.

Private markets break both assumptions. Capital is called over years, distributions come back unevenly, and NAV is an appraisal until the asset sells. Any single number you quote is a compression of a messy cash flow stream. The job of an LP is to use multiple lenses so the compression does not hide the story. For background on why the early years look worse than the end, see our LP guide to the J-curve.

IRR: useful, abused, and frequently misunderstood

Internal Rate of Return is the discount rate that sets the net present value of a fund's cash flows to zero. It answers a specific question: what annualized return did the actual dollars in the ground earn, given when they went in and came out.

Where IRR helps:

• Comparing funds against their own pacing assumptions.
• Measuring the impact of an early distribution or a subscription line.
• Flagging managers who sit on uncalled capital for years.

Where IRR misleads:

• A small early distribution can inflate IRR for the life of the fund.
• A subscription line that delays the first call boosts IRR without changing the underlying deal economics. We unpack the mechanics in our [capital calls guide](/insights/capital-calls-lp-guide-forecast-fund-track).
• Late-stage NAV markups inflate IRR even if no liquidity ever materializes.
• IRR cannot be aggregated cleanly across funds. Pooled IRR and average IRR can disagree by hundreds of basis points.

A useful discipline: never quote IRR without DPI next to it. If DPI is below 1.0 and IRR is above 20 percent, the return is still mostly paper. Treat that combination as a hypothesis, not a result.

TWR: borrowed from public markets, occasionally helpful

Time-Weighted Return strips out the effect of cash flow timing. It is the standard in public markets because the manager does not control when money arrives.

In private markets, TWR is less natural. The GP very much controls when capital is called and distributed, so removing that timing actually hides skill. Still, TWR has a role:

• For total portfolio reporting, where you want one number that combines public and private sleeves without distortion from pacing.
• For board-level reporting where the audience expects an annualized return that behaves like a public benchmark.
• For internal coverage teams comparing portfolio segments over the same window.

Do not use TWR to compare two private funds against each other. It will systematically reward managers who hold longer and penalize managers who return capital quickly, which is the opposite of what you want to reward.

Multiples: MOIC, TVPI, DPI, RVPI

Multiples are the simplest and often the most honest measures because they ignore time. They tell you how many dollars you got back, not how fast.

• MOIC (Multiple on Invested Capital): total value divided by capital invested, usually at the deal level.
• TVPI (Total Value to Paid-In): residual NAV plus distributions, divided by paid-in capital, at the fund level.
• DPI (Distributions to Paid-In): cash actually returned, divided by paid-in. The cleanest measure of realized success.
• RVPI (Residual Value to Paid-In): unrealized NAV divided by paid-in. The part still subject to mark adjustments.

The useful framing for an IC: TVPI is the promise, DPI is the proof. As a fund ages, DPI should be climbing toward TVPI. A fund that is eight years in with DPI well below 1.0 and a TVPI of 1.8x is telling you that almost the entire return is still unrealized paper.

PME: the benchmark that asks the only honest question

Public Market Equivalent analysis asks the question that matters: would the LP have been better off putting the same dollars into a public index, on the same dates, with the same withdrawal pattern?

There are several flavors:

• Long-Nickels PME: invests called capital into the index, withdraws distributions, and computes an IRR on the synthetic public position. Useful but breaks down when NAV exceeds the synthetic index value.
• Kaplan-Schoar PME: a ratio. The present value of distributions discounted by the index, divided by the present value of contributions discounted by the index. Above 1.0 means the private fund beat the index on the same dollars. Below 1.0 means it did not.
• Direct Alpha: solves for the excess IRR the private fund earned over the index. Translates the comparison back into a number the committee already understands.
• PME+: a Long-Nickels variant that scales distributions so the synthetic NAV matches the fund NAV at the end. Cleaner for late-stage funds.

For most LP committees, Kaplan-Schoar PME plus Direct Alpha is the right pairing. KS gives a single readable ratio, Direct Alpha translates it into basis points of outperformance. Pair them with DPI and you have most of what you need to defend a re-up. Our re-up decision framework walks through how to weight these signals.

A few practical notes on PME:

• The index you pick matters more than the formula. A large buyout fund benchmarked to the S&P 500 will look different than the same fund against the Russell 2000. Pick an index that reflects the strategy, not the one that flatters the manager.
• Currency matters. A euro-denominated fund benchmarked against a USD index without an FX adjustment is comparing two different things.
• For sector funds, use a sector index. Healthcare buyout benchmarked against broad equities is not useful.

Vintage-year quartiles: the industry default, and its limits

Most LPs also report against vintage-year quartile data from Preqin, Burgiss, Cambridge, or Pitchbook. This is fine as a sanity check but has three issues worth flagging.

1. Survivorship. The funds that report into commercial databases are skewed toward larger, more institutional managers. 2. Self-selection. GPs choose which funds to report. Underperforming vehicles get quietly omitted. 3. Definition drift. Vintage year, fee treatment, and currency conventions vary across databases. Two providers can disagree about the same fund's quartile.

Use vintage quartiles to spot outliers, not to make decisions. A second-quartile fund with a strong PME is more interesting than a first-quartile fund that lost to the index.

How to build a benchmarking stack that holds up

The LPs we see doing this well do not pick one measure. They run a small, consistent stack at every reporting cycle and at every re-up decision.

A workable default:

1. DPI and TVPI for the realization status. 2. Net IRR for the time-adjusted return, always shown next to DPI. 3. Kaplan-Schoar PME against a strategy-appropriate index for the public market comparison. 4. Direct Alpha to translate PME into basis points. 5. Vintage quartile as a tertiary check, not a primary signal.

At the portfolio level, layer TWR on top for board reporting and total-portfolio attribution.

Do this consistently for every fund, every quarter, and the conversation changes. Instead of arguing about whether a fund is "good," the committee can ask the better questions: is the realization pace tracking the pacing model, is the public market premium holding up at this point in the cycle, and is the manager earning the illiquidity premium we underwrote.

Common reporting traps

A short list of things we see go wrong in LP benchmarking, in rough order of frequency:

• Showing gross IRR without flagging it. Always net of fees and carry when comparing to public benchmarks.
• Mixing since-inception IRR with rolling TWR in the same table without labels.
• Letting a subscription line distort IRR without disclosing the levered-to-unlevered gap.
• Benchmarking a 2021 vintage today and concluding anything about manager skill. The J-curve is still doing most of the work.
• Using a public index that does not reflect the actual investable opportunity set.
• Quoting peer quartiles from one database without acknowledging the survivorship issue.

Most of these are cured by a footnote and a second number. None of them are cured by a louder presentation.

Where this fits into the broader LP workflow

Benchmarking is one piece of a larger feedback loop. The pacing model sets the expectation for when capital should be called and returned. The performance measurement tells you whether the actual results matched the underwrite. The re-up decision closes the loop.

For more on the connected pieces, see our notes on building a commitment pacing model, evaluating secondaries, and running a disciplined co-investment program.

The short version

IRR tells you what you earned on the dollars in the ground. TWR removes the cash flow timing for portfolio-level reporting. Multiples tell you whether the money came back. PME tells you whether the public index would have done it for you. Use them together and label each one clearly. The committee that argues about which number is the right number is asking the wrong question. The right question is what the four numbers together imply about whether to commit again.

Topics

• Benchmarking
• IRR
• TWR
• PME
• LP Operations
• Performance Measurement
• Private Equity