The direction of travel in private markets is clear: retailisation, evergreen structures, semi-liquid funds, all of it is pushing toward more frequent valuation points and NAVs that investors can actually transact on. That creates a real challenge for the valuation profession. Reporting and review cycles compress, valuation updates come more often, and the volume of work scales with the frequency, not just the complexity. Doing all of that well, at that pace and at that scale, is genuinely hard.
The obvious answer is technology, faster infrastructure, more automation, tighter cycles, and on the surface that's exactly right. What gets much less attention is a quieter problem sitting underneath the whole conversation, one that faster technology doesn't solve on its own. It just makes it easier to miss.
Why doesn't faster technology close the data gap?
Public market data updates daily. Pricing feeds, index levels, credit spreads, all of it refreshes on a schedule technology can keep up with easily. Company-specific information doesn't move at the same speed, and it never will. A private company's financials, its operating metrics, the qualitative context a valuer actually needs to make a defensible judgement call, that information updates quarterly if you're lucky, sometimes less often, and no amount of infrastructure investment changes the reporting cadence of the underlying company.
That gap matters more, not less, as valuation frequency increases. A monthly valuation built on quarterly company data has a known staleness window, which is something everyone in the process should keep in mind. A daily or near-daily valuation built on the same quarterly company data has the same staleness window, just dressed up in a much more current-looking wrapper. The market data around it refreshes daily. The number that actually drives the valuation doesn't.
Credit is a good example of why this matters more than it looks. A meaningful share of day-to-day value movement comes from market yields, and those do move as fast as the feeds tracking them. But some of the sharpest repricings, the moments where a valuation genuinely needs to shift, are usually triggered by something in the underlying credit picture, a rating action, a covenant issue, a change in the borrower's own numbers. That's exactly the data that updates slowest. It's also exactly the data behind the biggest swings. Faster cadence doesn't change that, it just means the slow-moving piece sits quietly underneath a much faster-looking process. Credit's a clear example, but the same structural gap shows up in real estate appraisal cycles and infrastructure asset reviews.
What does faster valuation technology actually solve?
The more frequently a valuation gets produced, the more current it looks, and the easier it becomes for everyone downstream, investors, boards, auditors, to assume the whole thing is fresh. Nobody's being deliberately misled. It's just that a daily NAV blurs which inputs are actually current and which ones are still running on last quarter's numbers.
The honest framing is that faster valuation technology solves a real problem and masks a different one. It solves for capacity and speed, whether a team can physically produce valuations at the pace investors and regulators now expect. It doesn't solve the information problem, which is that some of what feeds the model will always lag behind the schedule the model itself now runs on. The mistake is treating the capacity problem and the information problem as one and the same, assuming faster technology closes both gaps at once. That's where the real challenge in this next phase actually sits.
What should a valuation function be able to say?
What should change isn't the ambition toward faster valuation cycles. It's how explicit the process is about which inputs are current and which aren't. A valuation function that can say, clearly and on the record, exactly which data points refreshed today and which are still running on last quarter's disclosure, is in a fundamentally stronger position than one that's simply gotten faster at producing an output without being able to say the same thing. That's a governance question as much as a technology one.
Regulators are already circling this territory, if not quite from this angle. The CSSF, in its 2026 self-assessment questionnaire, now asks directly whether a fund's valuation policies and procedures address stale valuations. That's the right question, but it's worth being precise about what it actually covers. A stale valuation is an output that didn't get updated when it should have, a process failure. What this piece is pointing at is different, and arguably harder to spot: a valuation that updates perfectly on schedule, every time, built on inputs that were always going to lag behind that schedule. Not a process failure. A structural one. Any valuation policy serious about stale valuations should be asking both questions, not just the first.
This is exactly where architecture matters. A platform built to be transparent by design, timestamped inputs, source data linked to every output, a clear line between what updated today and what didn't, gives everyone in the process the same picture at the same time, so nobody downstream has to guess what's current and what isn't.
References
- CSSF, "UCI reports foreseen by Circular CSSF 21/790 for year-ends 31 January 2026, 28 February 2026, 31 March 2026 and 30 April 2026 now available on eDesk and information on main updates", communiqué of 10 February 2026 — announces the Self-Assessment Questionnaire released on 9 February 2026 and, in the Valuation section, the sub-question "addressing whether the written valuation policies and procedures include stale valuations".
- Circular CSSF 21/790 of 22 December 2021, on the reports to be submitted by undertakings for collective investment — the instrument under which the Self-Assessment Questionnaire, Separate Report and Management Letter are issued per financial year-end.
- CSSF, Feedback Report — Thematic review: Valuation framework for less liquid and illiquid assets, 4 June 2026 — supervisory expectations on valuation policies, procedures and independent validation for assets without observable prices.
- IFRS 13 Fair Value Measurement, paragraphs 72–90 — the fair value hierarchy and the distinction between observable market inputs (Levels 1–2) and unobservable entity-specific inputs (Level 3).
Related pages
- AIFM & Fund Managers — the valuation data layer behind fund-level NAV and reporting cycles.
- Credit Spreads & Interest Rates — the market-data side that does refresh daily, and where it comes from.
- Introduction to Loan Valuation and the Private Debt Market — the market this piece is describing, from the credit side.
- PIK Loans and PIK Toggle: Private Credit Risk Monitoring — why borrower-level signals, not market yields, drive the sharpest repricings.
- IFRS 13 — Measurement Techniques and Measurement Inputs — how the standard classifies observable versus unobservable inputs.
- Calibration in Cost of Capital — Friday Digest — what to revisit at each valuation date when there is no market price to check against.
- Facts & Figures — coverage, sources and update cadence of the underlying database.
