How to Answer the DDQ Performance Section: IRR, MOIC, TVPI and the Follow-Ups LPs Send Next

The email from the investment consultant arrives late on a Thursday, a week after the DDQ went out, and it is polite in the particular way that makes an IR team nervous. Your track record response, it says, shows a net IRR of 13.8% for Fund II, while the deck their investment committee reviewed in the spring shows 12.4% for the same fund, so could you confirm which figure is correct, and whether either one includes the effect of the subscription line? Someone on your team opens both files side by side and realises, with a sinking feeling that has nothing to do with arithmetic, that both numbers are right: one is net IRR with the credit facility and the other is net IRR without it, and neither document says so. Nothing was miscalculated, yet the fund now owes an allocator a round of explanation, and that allocator has started, quietly, to wonder what else might not line up.

That scene, along with every figure in it, is illustrative, but it captures the real difficulty of the performance section. Finance and the fund administrator produce the figures every quarter; the hard part is presenting them so that every number arrives with its definition, its basis and its as-of date attached, and stays identical across every DDQ, RFP and pitch deck that leaves the firm.

What the performance section is really asking

Read the Track Record section of the ILPA Due Diligence Questionnaire 2.0 closely and you will notice that it asks for far more than a table of returns. Question 14.5 wants three to five investments with a TVPI below 1.0x and an account of what went wrong, while question 14.6 asks you to compare LP returns, both with and without the use of a credit facility, against the most appropriate private and public market benchmarks for your last two predecessor funds and to describe the consistency and dispersion of those returns. The LP wants the numbers, the basis they were built on, and evidence that you understand what they do and do not show.

The APMP’s distinction between compliance and responsiveness is a useful lens here. A compliant answer fills in the cells the questionnaire asks for, and that keeps you in the process; a responsive answer anticipates the question the reader will ask next and answers it first, which in a performance section means stating the basis of every figure, naming the benchmark and the method, and owning the weaker deals before anyone has to ask. Our ILPA DDQ guide walks through the rest of the questionnaire.

The metrics, defined properly

Many LP follow-ups on performance start with two metrics being read as if they answered the same question. The definitions below follow the ILPA Performance Template definitions and the 2020 GIPS standards for firms from CFA Institute, which for closed-end funds reporting money-weighted returns call for a since-inception IRR alongside TVPI, DPI, RVPI and the PIC multiple.

Gross and net IRR deserve the most care, because their spread is how an LP reads your fee load and carry. Under the ILPA Performance Template, which ILPA says should be used on a go-forward basis for funds commencing operations on or after 1 January 2026, fund-level net IRR is calculated on the cash flows between the fund and its fee-paying investors and should be shown both with and without the subscription facility, while ILPA’s Performance Template guidance lists the fund-level gross IRR and TVPI as optional metrics. If your firm is an SEC-registered adviser, the Marketing Rule adds a harder edge for anything that counts as an advertisement: gross performance must come with net performance at least as prominent, over the same period and on the same methodology, and the SEC staff FAQ takes the view that comparing a gross IRR without the subscription line only against a net IRR with it would violate the rule. Whether a particular DDQ response counts as an advertisement depends on the facts of how it is requested and used, and target or projected returns count as hypothetical performance with conditions of their own, so it is worth settling the treatment with your compliance team before the first questionnaire of the season goes out.

MOIC and TVPI drift apart because of the denominator, since MOIC divides by the capital that went into deals and TVPI by everything LPs wired in, including fees and expenses, so the two never measure quite the same thing even on the same gross or net basis. TVPI splits cleanly into DPI and RVPI, which is the split an LP reaches for to see how much of your multiple is cash and how much is a valuation. PME turns an absolute return into a relative one, exactly what question 14.6 is after; CFA Institute’s Enterprising Investor primer on PME explains how the Long-Nickels and Kaplan-Schoar versions differ (Direct Alpha is a third common method), and in a DDQ the rule is to name the index, the method and why you chose them.

Venture and private credit managers face the same logic with a different emphasis: a young venture fund’s TVPI can sit almost entirely in RVPI, so the valuation policy carries more of the answer, while private credit funds can be asked about yield and realised losses as well.

One illustrative fund, from cash flows to answer

Every figure here is illustrative and simplified, built to show how the metrics relate. Take Fund II, with $500m of commitments, reporting as of 30 June 2026. LPs have paid in $400m, which gives a PIC multiple of 0.80x, and of that $340m went into investments while $60m covered management fees and fund expenses. The fund has distributed $260m to LPs and holds a NAV of $420m, net of accrued carried interest.

• Net TVPI: ($260m + $420m) divided by $400m = 1.70x
• DPI: $260m divided by $400m = 0.65x
• RVPI: $420m divided by $400m = 1.05x, and 0.65x plus 1.05x gives back the 1.70x TVPI
• Gross MOIC: the investments have produced $285m of proceeds and are valued at $480m before accrued carry, so ($285m + $480m) divided by $340m = 2.25x
• IRR: on the dated LP cash flows (calls from March 2020 to June 2023, distributions from December 2023), net IRR comes to 12.4%, while gross IRR on the investment cash flows comes to 19.2%

The $25m gap between gross proceeds and net distributions is carry already paid, and the $60m gap between gross and net value is carry accrued, which together with fees explains why a 2.25x gross MOIC becomes a 1.70x net TVPI and a 19.2% gross IRR becomes 12.4% net. That bridge, written in two plain sentences, is one of the most useful things you can add to a performance answer.

Now suppose, as the opening email implied, that Fund II had funded its first two capital calls, $200m in total, from a subscription facility and called that money from LPs eighteen months later, together with $10m of interest. Net IRR with the facility rises to 13.8%, because LP capital was at work for less time, while net TVPI slips to 1.66x, because the interest made paid-in capital larger. Those are the two figures from the opening email, each correct, and each useless to an LP without its label.

Timing matters even without leverage. In an illustrative case, two deals that each return 1.8x produce an IRR of roughly 15.8% if the money comes back in four years and roughly 8.8% if it takes seven, which is why LPs read IRR and multiples together, and why a high IRR paired with a modest multiple can prompt a question about early partial exits or recapitalisations.

The follow-up questions LPs send next

These are the follow-ups worth closing off in advance.

1. “Is that with or without the credit facility?” Give both and label both, in line with ILPA DDQ question 14.6 and the ILPA template. For GIPS-compliant firms, the figure without the line can be left out only when its principal was repaid within 120 days using capital called from investors and none of it funded distributions.
2. “What sits between gross and net?” Offer the bridge before they ask for it, naming fees, fund expenses and carry, as in the Fund II example.
3. “How much of the TVPI is real money?” Show DPI and RVPI side by side and point to the valuation policy behind the NAV.
4. “Which benchmark, and why that one?” Name the private benchmark and its quartile basis, the public index used for PME, and the method, then say why they suit the strategy.
5. “What went wrong on the deals below 1.0x?” Question 14.5 asks for this directly, so write the lessons learned once, have them approved, and keep them current.
6. “Is anything excluded?” State whether the figures cover fee-paying LPs only, whether the GP commitment and co-investment vehicles are in or out, and why any predecessor deals are missing, as question 14.4 asks.

Writing the answer itself

A strong performance answer has a recognisable anatomy: the headline figures with their basis, the as-of date, a methodology note, one line of context, and the disclaimer your compliance team has approved. Our guide to the anatomy of a DDQ answer covers that structure in general terms; applied to Fund II, an illustrative version reads like this.

[Illustrative figures] As of 30 June 2026, Fund II has generated a net IRR of 12.4% and a net TVPI of 1.70x excluding the effect of its subscription facility (13.8% and 1.66x including it), a DPI of 0.65x and an RVPI of 1.05x, on paid-in capital of $400m (0.80x of commitments). Over the same period, gross IRR calculated on investment cash flows without the subscription facility, before management fees, fund expenses and carried interest, is 19.2%, with a gross MOIC of 2.25x. Net figures reflect the aggregate cash flows of fee-paying limited partners and exclude the GP commitment. Unrealised investments are valued under the firm’s valuation policy, provided in the data room. Figures are unaudited. Past performance is not indicative of future results.

Every number carries its basis, every figure shares one as-of date, and nothing invites a clarifying email, which is the standard to hold across every section of a DDQ response. If a questionnaire asks for a date your administrator has not yet closed, say which quarter you are reporting and why, rather than mixing a fresh IRR with last quarter’s multiples in the same answer.

Keeping DDQs, RFPs and pitch decks in step

In a single fundraise the same Fund II net IRR can appear in the ILPA DDQ, a consultant’s own questionnaire, a pension RFP, the pitch deck, the PPM and the quarterly report. It stays consistent only if there is one approved version, which in practice means one performance block per fund per quarter, owned by finance, signed off by compliance, and refreshed on a calendar tied to the administrator’s quarter-end close. When the new block is approved, the old one is retired rather than deleted, so you can still explain to an LP why the number in this year’s DDQ differs from last year’s. The proposal world points the same way: in our Proposal Win Rate Report, a survey of 100+ bid teams working on RFPs, 65% of high-win teams (51%+ win rate) had formal review or governance, against 42% of low-win teams (0 to 25%).

Where an approved content library fits

AutoRFP.ai is the accuracy-first, AI-native platform for RFPs, security questionnaires, and DDQs: every answer is written from content your team has approved, and shows the sources behind it. For the performance section, that means the quarterly performance block, the gross-to-net bridge, the benchmark rationale and the approved disclaimers sit in one content library, so when an ILPA DDQ, a consultant questionnaire or a pension RFP asks for Fund II’s net IRR, the draft is written from the version your team signed off, with the source shown so a reviewer can check the figure and its as-of date before anything goes out. A Trust Score on each answer helps reviewers decide where to look first.

Your finance team and administrator still own the numbers, because nothing in the content library calculates returns; its job is to make sure the approved wording, with its basis and as-of date intact, is what gets used every time the question comes round. Across the platform, 63% of AI-generated answers are submitted with no edits, and because every DDQ and RFP is drafted from the same library, they start from the same approved figure. For the wider workflow, see our AI for DDQs guide, our roundup of the best DDQ software and the DDQ response software page.

See your performance section drafted from approved content

Bring last quarter’s DDQ and your approved performance language, and we will show you how AutoRFP.ai drafts the track record section from it, with every source visible. Book a demo.