# Private Markets Fund Data: Common Problems and How to Fix Them > Why NAV oversight, Excel workflows and LP reporting break at growing firms managing $500M to $10B, and what to do about each. - Published: Nov 5, 2026 - Topic: Fund Operations - Source: https://aquata.ai/articles/private-markets-fund-data-problems --- Why fund data breaks first at firms that run lean, and what to actually do about it, whether or not you ever buy anything. ## Why is fund data so hard at firms that run lean? Nothing in private markets arrives clean, and nothing arrives the same way twice. Capital call notices come as PDFs. Portfolio company financials come as Excel attachments. NAV packs come from six administrators in six different layouts, on six different schedules. The biggest firms absorb all of that with a data team and a dedicated reporting function. A lean firm absorbs it with a controller, a fund accountant, and a shared drive only one person fully understands. The problem is not effort, and it is not skill. The problem is arithmetic. The reporting bar is set by the biggest managers, and it arrives in your data room unchanged, no matter how many people you have to clear it. ### What to know - **The bar does not scale down**: LP expectations are set by the biggest managers and arrive unchanged. They do not shrink because your team did. - **One missing layer**: NAV packs, spreadsheets, track record, LP reporting: they all break in the same place. No governed source underneath. - **The obvious moves assume a data team**: Hire, build, buy big, wait: each assumes people and budget you do not have. But fixes that cost discipline are in every chapter. - **What a real fix looks like**: It reads the PDFs, connects to your providers, and lets your accountants set the checks. If the plan starts with hire, it is not a fix. ## Six fund admins. One NAV. Zero confidence. > Every admin sends the pack differently. The oversight process is me, Excel, and a deadline. > > Fund controller, $2B private credit Each administrator sends the pack differently. Alter Domus does not format like Citco; Citco does not format like SEI. Checking their numbers means lining each pack against your own records and the underlying documents, every fund, every period, against a sign-off deadline. Six admins means six reconciliation routines, and at most firms the logic lives in one person’s head. Meanwhile the clock is speeding up. Evergreen and semi-liquid vehicles push valuations from quarterly toward monthly, sometimes daily. The UK FCA put private market valuation practices under review in March 2025, and the SEC Rule 2a-5 has made fair-value governance a documented obligation since 2020. Checks that barely fit a quarter now need to happen in days. ### Why the obvious fixes fail - **Hire**: An analyst without a system spends the month building spreadsheets, not reviewing numbers. - **Build in-house**: A checker that reads six admin formats is a software product you maintain forever. - **Buy enterprise**: Shadow-accounting suites are priced for firms running a full second set of books. - **Wait**: Every fund launch adds an admin, a format, a routine. The gap widens every vintage. ### What you can actually do Start with the workflow. - **Get oversight out of inboxes.**: A single tracked process beats NAV review scattered across email threads. - **Write the validation logic down as rules.**: Every controller carries checks in their head. Make them explicit rules with materiality thresholds. - **Score your administrators.**: Track error rates, timeliness, and responsiveness against SLAs. Admins prioritize clients who measure them. - **Then automate the comparison.**: Once the rules exist on paper, the case for a system that ingests every format writes itself. **NAV Oversight** Every admin’s pack in one model, checked by your rules, signed off with a trail. ## The firm runs on a file called FINAL_v15 > V15_FINAL_FINAL_DRAFT.xlsx isn’t actually the final version. I checked. > > Fund accountant, $1.5B buyout Excel became the firm’s plumbing because it works, for one person, one task, one deadline. Then the workbooks got copied, linked, and stacked: reconciliations feeding waterfalls feeding LP statements, held together by macros nobody maintains. Raymond Panko’s research, cited by EuSpRIG, found errors in 94% of spreadsheets studied, with a 5.2% cell error rate. A 2020 BBC report traced 15,841 lost Covid cases to an Excel row limit. The firm’s most important numbers travel through files with no version control, no audit trail, and no owner. At a big firm, a controls team contains this. At a lean firm, the person who built the waterfall tab is also the person who has to leave. The file breaks when they leave for good. ### Why the obvious fixes fail - **Hire**: The new hire inherits the workbooks and becomes the next person who can never leave. - **Build in-house**: Homegrown replacements trade a spreadsheet problem for a software-maintenance problem. - **Buy enterprise**: Enterprise platforms assume a team to migrate, model, and govern. - **Wait**: Every new fund, currency, and entity adds tabs. The shelf is already bending. ### What you can actually do Do not ban Excel. Your team is fast in it. That plan gets ignored by week two. - **Inventory the load-bearing workbooks.**: Which files produce numbers that reach LPs, auditors, or regulators? That list is your real risk register. - **Put light controls on that list only.**: Named owner, version discipline, documented source. Not enterprise governance for 400 spreadsheets. - **Move the system of record out of Excel, not the users.**: Excel becomes a window onto governed data. Analysts keep their pivot tables. The firm stops depending on v15 being final. **Central Data Layer** One governed source for every fund, entity, and investor. Excel stays as the sandbox, not the system of record. ## Three funds. Three IRR methods. One number LPs will believe. > An LP asked for net returns with and without one deal. It took three teams a week. > > CFO, $5B multi-strategy GP Growth multiplies math. New funds, new vintages, maybe an acquired team, and now there are three IRR conventions, two waterfall structures, and performance logic spread across spreadsheets that were each correct on their own terms. The cross-fund questions, exactly what LPs ask in diligence, have no single place to be answered. Your administrators cannot save you here. An admin’s job is fund-level NAV and investor allocations, not firm-level track record. The stakes are asymmetric: a restated performance number costs the credibility of every other number in the deck. MSCI’s 2024 research found subscription lines inflate IRR by roughly 100 basis points, and the SEC’s Marketing Rule FAQ now requires IRR with and without subscription-facility impact shown side by side. ### Why the obvious fixes fail - **Hire**: A performance analyst spends the first quarter reconciling conventions before producing one number. - **Build in-house**: Analytics built on inconsistent data hard-codes the inconsistencies. - **Buy enterprise**: Performance platforms are built for firms with performance teams, not a controller with a deadline. - **Wait**: Every vintage adds history. Restating conventions gets harder every year you do not. ### What you can actually do Five conventions diverge: cash-flow timing, what net nets out, subscription-line treatment, waterfall type, currency translation. Name all five. - **Pick one set of conventions and write them down.**: The document matters more than the choice: most disputes trace to two teams both being right under different assumptions. - **Centralize the inputs before the analytics.**: Cross-fund numbers built on scattered data hard-code the scatter. Admin packs and accounting data land in one place first. - **Make every number traceable.**: The LP question after what is the return is how did you get it. Weigh auditability first. **Performance Track Record** Admin and accounting data in one layer for IRR, TWR, and MOIC at fund, deal, and investor level. ## Four symptoms. One cause. NAV packs, spreadsheets, track record, LP reporting. They all break in the same place: no single trusted source of fund data underneath, and nobody on a three-person team whose job is to build one. Hold that thought, because it is where the fixes converge. ## LPs now shop the reporting, not just the returns > The 45-day quarterly letter used to be the standard. Now it’s the complaint. > > Head of Investor Relations, $3B credit GP LPs are writing fewer, bigger checks, and screening managers on how fast and how deep they report. The CFA Institute reported in September 2025 that only 10% of LPs think valuation reporting is functioning well. Adams Street’s 2025 survey found 88% of LPs plan to increase co-investment allocations. CSC’s 2025 private capital CFO survey found about three in four LPs want performance data daily or on demand. The hard part: what each LP wants is negotiated, not standard. Side letters, custom cuts, exposure look-throughs. Every answer is assembled by hand from PDFs, admin files, and the spreadsheets in Chapter 2, which is why the quarterly letter takes 45 days. ### Why the obvious fixes fail - **Hire**: An IR associate without connected data becomes a full-time report assembler. - **Build in-house**: An investor portal on scattered data ships stale numbers with better fonts. - **Buy enterprise**: Reporting suites assume the clean data already exists. Yours is the part that is missing. - **Wait**: Fundraising does not wait. The next data room opens with the same gaps and sharper questions. ### What you can actually do Transparency is negotiated and operational. Compete on the operational part. - **Adopt the ILPA templates now, not when forced.**: Every LP moved onto the standard format is a bespoke report you stop producing. Early adopters get diligence credit. - **Build answers once, reuse them.**: Most custom LP requests are variations on a dozen questions. Parameterize per LP to convert bespoke work into configuration. - **Close the gap between the report and the backup.**: The data behind the letter should match the letter, on demand: only achievable when reporting pulls from the same governed source as operations. **Self-Service LP Reporting** Fund data your IR team and your LPs can query without filing a ticket. ## What a fix looks like when no data team is coming Four problems, one root. What has to be true of any fix that works without a data team, and what it has looked like at firms that run lean. Each chapter has fixes that cost discipline rather than money, and they are worth doing. But the four problems share one root: fund data scattered across admin packs, PDFs, spreadsheets, and systems, with no governed layer that makes it one thing. Fix that layer and each symptom gets an order of magnitude easier. The trap is assuming the fix must be an enterprise data program. At a lean firm, it cannot be. ### The five constraints - **It reads the documents.**: Capital call notices, valuation PDFs, and Excel financials become structured data without re-keying. - **It already connects to your providers.**: Your administrators and vendors are an onboarding step, not an integration project. - **Your people set the checks.**: Fund accounting writes the validation rules; breaks surface before sign-off. - **Anyone can answer the question.**: IR pulls the custom cut without filing a ticket; Excel stays as the sandbox. - **There is no hire in the plan.**: If it assumes a data engineer or an enterprise budget, it was built for someone else. Start with one workflow and prove it in a quarter. The firms pulling ahead are not the ones with the biggest teams. They changed the arithmetic instead. ## Sources - Raymond Panko, EuSpRIG 2008: 94% of spreadsheets contain errors, 5.2% cell error rate: https://eusprig.org/wp-content/uploads/0802.3457.pdf - BBC News, 2020: 15,841 lost Covid cases traced to Excel row limit: https://www.bbc.co.uk/news/uk-54422505 - MSCI, 2024: subscription lines inflate IRR by approximately 100 basis points: https://www.msci.com/research-and-insights/blog-post/inflating-returns-with-subscription-lines-of-credit - McDermott Will & Emery, 2024: SEC Marketing Rule FAQ on IRR with and without subscription facilities: https://www.mcdermottlaw.com/insights/marketing-rule-faq-impact-of-subscription-lines-of-credit-on-presentation-of-net-irrs/ - CFA Institute, September 2025: 10% of LPs think valuation reporting is functioning well: https://www.cfainstitute.org/insights/articles/private-equity-markets-transparency-development - Adams Street Partners, 2025 Global Investor Survey: 88% of LPs plan to increase co-investment allocations: https://www.adamsstreetpartners.com/insights/2025-global-investor-survey/ - CSC, Future Private Capital CFO Survey, 2025: about three in four LPs want performance data daily or on demand: https://www.businesswire.com/news/home/20250430162117/en/Private-Market-Fund-Managers-Face-Data-Technology-and-Talent-Challenges-CSC-Study-Reveals-Key-Insights - UK FCA, March 2025: multi-firm review of private market valuation practices: https://www.fca.org.uk/publications/multi-firm-reviews/private-market-valuation-practices - SEC, 2020: Rule 2a-5 fair-value governance press release: https://www.sec.gov/newsroom/press-releases/2020-302