The Transparency Gap: Why Private Real Estate Cannot Answer What Public Markets Answer Daily

Two investors hold the same kind of asset: a portfolio of leased commercial buildings in the United States. The first holds shares in a listed REIT. On any morning she can see a price, and within 40 days of quarter end she can open a filing in which occupancy, the lease expiration schedule and the debt maturity ladder each sit in a defined place, in a defined format, on a defined date. She did not have to ask for any of it. A trade body will shortly add her REIT’s figures to every other’s and publish a sector total, because the fields are the same in every filing.

The second holds a limited partnership interest in a private fund. His answer to the same three questions arrives in a quarterly pack, some 45 to 60 days after the period closed, as a PDF. The occupancy in it was re-keyed from a property manager’s spreadsheet. The NAV was consolidated from an administrator’s capital account statement and a valuer’s draft. If he wants to know whether the valuer’s rent assumption agrees with the rent roll behind the occupancy figure, the pack has no place where that reconciliation is recorded, because nobody was required to perform it.

It is tempting to read this as a difference in speed, or in the honesty of the people involved. It is neither. The public clock is not dramatically faster in Europe, where a listed company’s half-yearly report may take three months under the FCA’s disclosure rules, and the private manager is usually as careful as the REIT’s finance team. The difference is where the number lives. The REIT’s occupancy is a field, derived from a system of record and filed once. The fund’s occupancy did not exist until someone assembled it, and it revises whenever the assembly is redone.

This paper is about that gap: what the private frameworks actually require, why the gap persists inside those requirements, what it did to valuations and liquidity in the last cycle, what institutional LPs now ask for in response, and what closing it would take. The Smart Bricks Research Lab has been working through the question from the primary standards, the regulators’ texts and the index publishers’ numbers, and the argument that follows rests on those.

Question	Listed REIT investor finds it in	Available	Private real estate LP finds it in	Available
NAV	Listed REIT investor finds it inEPRA NRV/NTA/NDV or reported book value, reconciled from IFRS equity	AvailableEach filing; price daily	Private real estate LP finds it inFund NAV in the quarterly pack, consolidated from administrator, PM and valuer inputs	Available45 to 60 days after quarter end, 120 at year end (ILPA v2.0)
Occupancy	Listed REIT investor finds it inFiling field; Nareit aggregates the sector (93.8%, Q2 2026)	AvailableWithin weeks of quarter end	Private real estate LP finds it inProperty manager submission re-keyed into the pack	AvailableSame pack, undated as-at
Lease expiry	Listed REIT investor finds it inLease expiration schedule in the 10-K/10-Q	Available40 to 45 days after quarter end	Private real estate LP finds it inRent roll extract, if included	AvailableSame pack, often omitted
Debt maturity	Listed REIT investor finds it inDebt schedule in the filing	Available40 to 45 days	Private real estate LP finds it inFund-level debt note; asset-level often absent	AvailableSame pack
Fees and carry	Listed REIT investor finds it inNot applicable at asset level; G&A disclosed	AvailableEach filing	Private real estate LP finds it inCapital account statement; ILPA template if adopted	AvailableSame pack; two thirds of LPs never validate (poll)
Performance	Listed REIT investor finds it inTotal return observable from price and dividends	AvailableDaily	Private real estate LP finds it inIRR/TVPI as stated by the GP; recomputable only with cash flow table (ILPA Performance Template, from Q1 2026)	AvailableQuarterly, with estimates restated

The public investor's answer is a field derived from a system of record and filed on a clock; the LP's answer is assembled from three maintained views and sent as a document. The difference is where the number lives, not how fast it travels.

Source: SEC Form 10-Q General Instruction A; Nareit T-Tracker Q2 2026; EPRA BPR Guidelines (September 2024); ILPA Quarterly Reporting Standards v1.1 and Reporting Template v2.0 guidance (January 2025); ILPA Performance Template (January 2025); INREV European ODCE Index; bfinance poll (September 2025, directional)

Read together, the frameworks that govern private fund reporting are precise about three things: when the pack is due, what fields it contains, and how often the underlying assets must be valued.

On timing, ILPA’s Quarterly Reporting Standards, revised in September 2016, set an expected delivery of 60 days after quarter end and a target of 45, with a further 30 days where audited financials are involved. The guidance accompanying ILPA’s Reporting Template version 2.0, published in January 2025, asks direct funds to deliver within 60 days of a quarter and 120 days of a fiscal year end, and applies to funds in their investment period during the first quarter of 2026 or commencing on or after 1 January 2026. The year-end clock lengthened between the two documents, from 90 days to 120. INREV’s Reporting Guidelines require at least one interim report in addition to the annual report and leave frequency and timing to each vehicle’s documentation. INREV’s own European ODCE index appears eight weeks after quarter end, following a collection window of up to six weeks in which contributing funds may submit estimates, restated in the next publication.

On valuation frequency, the floors are lower than most LPs assume. Under the Alternative Investment Fund Managers Directive, an annual valuation of the assets and an annual NAV per unit satisfy the law. The GIPS standards require an external real estate valuation at least every 12 months, and permit that interval to run to 36 months where client agreements stipulate. Funds in the NFI-ODCE index must comply with the NCREIF PREA Reporting Standards, which include annual audits, quarterly valuations and time-weighted returns; that quarterly cadence is a condition of index membership rather than a legal floor.

What none of these documents contains is a requirement about the production of the number between the deadline and the field. No framework sets a threshold for how complete the data behind a pack must be before release, how often a published figure may be restated, or how far the as-at date of a valuation input may sit from that of the NAV it feeds. The frameworks govern the edges of the process: the due date, the template and the appraisal calendar. They are silent on the middle, where the number is made.

• 40 days

  Form 10-Q deadline for large accelerated and accelerated filers after quarter end

  SEC

• 45 to 60 days

  ILPA targeted and expected delivery of the quarterly pack (2016); 60 days for direct funds and 120 at year end under the 2025 template

  ILPA

• 8 weeks

  Publication lag of the INREV European ODCE Index, on data that may be estimates restated next quarter

  INREV

• Once a year

  Minimum frequency at which an AIFM must value assets and calculate NAV per unit

  AIFMD Article 19(3)

• 12 months

  Maximum interval between external real estate valuations under GIPS 2020, extendable to 36 by client agreement

  CFA Institute

• Quarterly

  Valuation frequency required of NFI-ODCE constituent funds under NCREIF PREA Reporting Standards

  NCREIF

Every private framework specifies when the pack is due or how often the asset is valued. None specifies what must be reconciled before the number is released.

Source: SEC Form 10-Q; ILPA Quarterly Reporting Standards v1.1 and Reporting Template v2.0 guidance; INREV; Directive 2011/61/EU Article 19(3); GIPS Standards for Firms 2020; NCREIF NFI-ODCE constituent criteria

Consider what each investor’s number is made of.

The REIT’s occupancy figure is derived. Somewhere in the company’s property management system a lease has a commencement date, an expiry and an area; the occupancy field in the filing is a computation over those records, run once, signed by an officer and filed. If the underlying lease record changes, the next filing changes with it, and the difference is visible because both filings are public and both used the same definition. The number has a single source and a single derivation, and both sit outside the document that reports it.

The fund’s NAV is assembled. Three parties maintain three views of the same portfolio. The administrator maintains the capital account: commitments, calls, distributions, allocations of income and expense by investor. The property manager maintains the operating view: the rent roll, arrears, occupancy, the operating statement. The valuer maintains the mark, built from a rent roll and a set of assumptions that may or may not be the property manager’s. Each view is itself a maintained ledger, revising as late invoices, corrected allocations and re-run appraisals arrive. At quarter end someone in the manager’s finance team consolidates the three into a NAV and a pack.

Nothing in that process requires the three views to agree before the pack goes out. The valuer’s rent assumption is tested against the property manager’s rent roll only if someone chooses to test it, and an accrual that never reaches the administrator never reaches the NAV. The consolidation is the only moment when the three views meet, and the pack records its result rather than a reconciliation, because no reconciliation is required.

The frameworks confirm this by where they place their tolerances. Luxembourg’s CSSF Circular 24/856, in force since 1 January 2025, sets the NAV calculation error threshold for funds investing primarily in assets such as unlisted shares, buildings and loans at 1.00% of NAV, and permits vehicles for professional and well-informed investors to document a higher threshold that must not exceed 5%. That is a tolerance for the moment after publication: it decides whether an issued NAV must be corrected and investors compensated. Neither the circular nor ILPA, INREV or the NCREIF PREA standards define a tolerance for the moment before, when the components could still be tested against their sources.

The Ontario Securities Commission’s settlement with CI Investments in February 2016 shows the failure in its plainest form. Interest earned on cash collateral was not recorded in the NAV of certain funds; the omission was, in the regulator’s words, not promptly identified or corrected, and approximately CAD 156.1 million was returned to investors. The OSC found no dishonest conduct. An accrual sat in one ledger and never reached the calculation that depended on it, and no step in the process was designed to notice.

The structural contrast, then, is this. The REIT’s figure is one derivation from one system of record, filed once and visible thereafter. The fund’s figure is a view of three other views, each of which keeps moving after the pack is sent. The pack captures the assembly at the moment it stopped, and the LP cannot re-derive any line of it, because the derivation happened in a spreadsheet he does not hold, from sources he cannot see, to a tolerance nobody wrote down.

The consequence is not confined to reporting departments. In 2022 the two architectures told their investors opposite stories about the same asset class.

For the calendar year to 31 December 2022, the NFI-ODCE index of open-ended diversified core funds returned 7.47% gross of fees, of which 3.44% was income and 3.92% appreciation. Over the same twelve months the FTSE Nareit All Equity REITs index returned negative 24.95%. The listed market, repricing daily as rates rose, marked US commercial property down by roughly a quarter; the appraisal-based index, valued quarterly, marked the same kind of property up. In 2023 the positions began to reverse: the listed index returned 11.36% for the year while the ODCE index lost 4.83% in the fourth quarter alone. Neither side was lying. One set of numbers was derived from prices that existed; the other was assembled from appraisals that trailed them.

The private side’s answer to the lag is a queue. In 2024, with the index’s one-year gross return at negative 1.43%, its 25 constituent funds took in USD 6.5 billion of contributions against USD 15.9 billion of distributions and redemptions, a net outflow of USD 9.4 billion from funds holding USD 278.5 billion of gross real estate. Investors who wanted out could not sell at the appraised price, because there was no buyer at that price; they joined a line and waited for the fund to raise cash or for the appraisal to fall to a level at which it could.

Regulators have recognised the mechanism explicitly. After a series of suspensions among UK open-ended property funds, the FCA’s Policy Statement 19/24, in force from 30 September 2020, requires a fund investing in inherently illiquid assets to suspend dealing when its standing independent valuer expresses material uncertainty over at least 20% of scheme property. The FCA’s stated purpose was to reduce first-mover advantage: the position in which the investor who redeems first, at a stale NAV, is paid out at the expense of those who stay. A regulator writing that rule concedes that a lagging NAV creates a race, and chooses to stop the race rather than repair the lag.

There is academic evidence that LPs already read the interim mark as a claim rather than a fact. Brown, Gredil and Kaplan, studying buyout and venture funds, find that some underperforming managers inflate reported returns during fundraising, that those managers are then unlikely to raise a next fund, and that top performers if anything understate. The market punishes the inflation, which means LPs are already discounting the mark and looking for something beneath it. The question is what they can find.

• +7.47%

  NFI-ODCE gross total return, year to 31 December 2022

  NCREIF

• -24.95%

  FTSE Nareit All Equity REITs total return, calendar 2022

  Nareit

• USD 15.9bn vs 6.5bn

  NFI-ODCE distributions and redemptions against contributions, 2024; net investor cash flow -USD 9.4bn

  NCREIF

• 1% of NAV

  Luxembourg error tolerance for funds investing primarily in other assets including buildings; professional vehicles capped at 5%

  CSSF Circular 24/856

• 20%

  Share of scheme property under material uncertainty that forces a UK illiquid-asset fund to suspend dealing, a rule aimed at first-mover advantage

  FCA PS19/24

• CAD 156.1m

  Returned to investors after NAVs were understated because accrued interest never reached the calculation

  Ontario Securities Commission

In 2022 the appraisal-based index and the listed index told investors opposite stories about the same asset class. The private side's mechanisms for the lag are tolerances applied after publication, queues, and suspensions.

Source: NCREIF NFI-ODCE 4Q2022 and 4Q2024 results; Nareit REIT Industry Fact Sheet (December 2023); CSSF Circular 24/856; FCA PS19/24; OSC news release, 10 February 2016

A number that has to be assembled before it can be sent cannot be queried by the person who receives it.

Smart Bricks Research Lab

Set out plainly, institutional LPs are now asking for four things, and each of them is a request for a figure they can rebuild rather than a figure they are told.

A capital account they can see today. The LP’s position in a fund is a simple arithmetic object: commitment, less calls, plus distributions, adjusted for allocated income, expense and unrealised movement. Every input is a notice the LP has received or a figure the administrator holds. Yet the position is delivered as a quarterly statement, rebuilt each quarter end, and the administrator’s version and the manager’s are reconciled only when someone notices they differ. The ask is that the position be a query against the notices and the ledger, available on any day. The shortfall is that most managers can produce it only as a document, on the reporting calendar.

Performance they can recompute. ILPA’s Performance Template, released on 22 January 2025 for implementation from the first quarter of 2026, standardises gross and net IRR and TVPI or MOIC, requires each to be shown with and without the effect of fund-level subscription facilities, and ships the cash flow table alongside the metric. An IRR stated without its cash flows is a conclusion; with them, it is an argument the LP can check. The subscription-line split exists because Albertus and Denes found that such facilities raise IRR-based performance in US buyout funds by 6.1 percentage points while multiples slightly decline, an effect invisible in most packs until the template. The shortfall is a manager who reports the metric and treats the flows behind it as internal.

Fees and carry they can validate. A management fee is a percentage of a defined base under a defined clause; carried interest is the output of a waterfall written into the partnership agreement. Both should trace from the invoice line to the clause that authorises it. In practice the fee is calculated in a spreadsheet, the waterfall is implemented by someone’s reading of the agreement, and the LP receives the result. The ask is that each fee line carry the clause and the base it was computed from. The shortfall is a line that carries only a number.

Obligations they can see honoured. Side letters grant most-favoured-nation elections, reporting rights, excuse rights and fee terms that differ by investor. They are negotiated once, extracted once into a compendium, and reread when someone suspects a breach. The ask is that the fund be able to state, for each investor, what it has promised and when each promise falls due. The shortfall is that the compendium is a document, and the triggers live in the calendars of whoever negotiated the letter.

Two observations tie these together. First, the demand does not depend on a regulator. The SEC’s private fund adviser rules, adopted in August 2023, would have required quarterly statements of performance, fees and expenses; the Fifth Circuit vacated them on 5 June 2024, and ILPA’s templates carried substantially the same content into the market on a voluntary basis. LPs asked because they needed it, and managers are adopting because their LPs asked.

Second, in every case the shortfall is architectural rather than a matter of candour. The manager is not withholding the capital account, the cash flows, the fee base or the side letter register. The manager does not hold them in a form that can be queried, because each was assembled for a document rather than derived from a record. The figure the LP wants exists only in the sense that it could be assembled again, by the same people, on the same calendar, with the same absence of reconciliation.

What the LP asks for	The instrument that exists	Where the manager falls short	The test to run
Performance they can recompute	The instrument that existsILPA Performance Template: cash flow table, returns with and without subscription lines	Where the manager falls shortIRR stated without flows; subscription-line effect undisclosed (6.1pp uplift in buyout)	The test to runDoes the cash flow table travel with the metric
A NAV whose components they can trace	The instrument that existsINREV RG27 GAAP-to-INREV NAV reconciliation; CSSF 1% tolerance	Where the manager falls shortValuer draft unreconciled to rent roll; accrual missing; tolerance defined only after publication	The test to runWas every component reconciled to source before release, and to what written tolerance
Obligations honoured on time	The instrument that existsSide letters and LPA; MFN election windows	Where the manager falls shortObligations extracted once into a compendium and reread only after a breach	The test to runCan the fund say what it has promised each investor, in seconds

Each ask has an instrument behind it. The shortfall is architectural: the number is assembled at quarter end rather than derived from the record, so it cannot be queried, rebuilt or validated by the person receiving it.

Source: ILPA templates and guidance (2025); INREV Guidelines RG27; CSSF Circular 24/856; Albertus and Denes (2019)

The commitment decision is where the gap now shows up as a cost to the manager. LP diligence instruments have widened from strategy and track record to the operating infrastructure behind the numbers. ILPA’s Due Diligence Questionnaire, in its 2021 version, covers accounting, valuation policy and service providers alongside the investment questions; INREV’s 2023 questionnaire covers vehicle risk processes as well as strategy and projected performance. A manager claiming GIPS compliance submits to verification of how its composites are constructed and, on a first claim, must present at least five years of performance or the record since inception. Each asks the manager to show how a number was made, and no longer only what it was.

Independence of the mark is being codified for the same reason. In January 2022 RICS accepted all 13 recommendations of the Pereira Gray review of real estate investment valuations. For regulated-purpose valuations in the UK the responsible valuer now rotates after at most five years and the firm after ten, with a three-year break before either may return. Rotation rules are what a profession adopts when the people relying on its output have stopped taking that output on trust.

Lateness has meanwhile acquired a meaning. A pack that regularly arrives at the outer edge of the 60-day window, or a NAV restated more often than its peers’, is read by an operational due diligence team as evidence about the process. Chronic lateness says the consolidation is hard, and a hard consolidation is one in which the three views disagree.

Where managers do not supply recomputable numbers, LPs are building the capability themselves. Large allocators run public-market-equivalent comparisons on their own cash flow records, reconcile capital accounts against their own notice files, and engage fee validators to re-run waterfalls from the partnership agreement. That work is expensive, and its coverage degrades as the number of managers in a portfolio rises, which produces a quiet preference at re-up time for the manager whose numbers are cheaper to verify. An earlier note argued that provenance would become a selection criterion for the assets a firm buys; the claim here is narrower and one step later, that it is becoming a re-up criterion for the managers an LP keeps.

Private real estate will not trade daily, and its appraisals will remain quarterly at best. The part of the gap that can be closed is the part between an assembled number and a derived one, and closing it is a matter of where the fund’s figures come from rather than how often they are published.

Stated as tests rather than features, a record-derived reporting layer has to do five things. It derives each investor’s capital account from the commitment ledger and from every call and distribution notice actually issued, so that the position on any day is a computation over documents the LP also holds. It reconciles each component of NAV to its source before release, on an exception basis: the administrator’s capital account against the manager’s ledger, the rent roll against the leases, the accruals against the bank. It ties the valuer’s inputs to the rent roll at first pass and records a written rationale for any movement in the mark. It scopes each investor’s view to that investor’s position and rights, so that the side letter register is a filter on the same record rather than a separate document. And it puts a person between the derived figure and the release, so that nothing reaches an LP without a named approver.

There is a listed-market precedent for a NAV that is defined rather than described. EPRA’s Best Practices Recommendations set out three NAV metrics, NRV, NTA and NDV, each reconciled from IFRS equity attributable to owners, so that a reader can move from the audited balance sheet to the reported NAV by a published path. INREV already asks the private side for the same thing: RG27 requires a reconciliation from GAAP equity to INREV NAV. The derivation is therefore not a novel demand. What the layer adds is continuity, so that the reconciliation exists at every point in the quarter rather than once in the pack.

The ILPA Performance Template is the external anchor. If the cash flow table travels with the metric, a layer that already holds the flows produces the template as a by-product of normal operation rather than as a quarterly project. The same holds for the capital account statement, the fee schedule and the obligations register: each is a view over records the layer already maintains. The earlier note on the IC memo standard asked that every material number be traceable outside the document that reports it; this is the same standard applied to the fund’s own reporting.

The test a committee should put to its own team or to any vendor is short. Given her capital account statement, can an LP re-derive it from the notices she received and the allocations she was shown, and does the answer change if a single notice is corrected at source? If the answer arrives as a PDF, the gap has been decorated rather than closed.

The cost of closing the gap is mostly governance before it is software. The only externally supplied tolerance a manager can borrow is the CSSF’s 5% ceiling for professional vehicles, and it is a tolerance for errors already published. Everything tighter, and everything that applies before release, has to be written in-house: how far the administrator’s ledger may diverge from the manager’s before the pack is held, how large a movement in a valuer’s rent assumption requires a documented reason, which control totals must tie each cycle and to what precision. That document is the CFO’s to write and the committee’s to approve. Behind it sits an exception queue that never empties, because tenants restructure, allocations are corrected and appraisals are re-run; someone must own that queue as a named role, and a manager who cannot say who clears it has not budgeted for the layer.

The limits deserve stating as plainly as the argument. Derived numbers are still appraisal-based at the valuation layer. GIPS’s twelve-month external valuation floor and NCREIF PREA’s quarterly cadence set the ceiling on how current a private NAV can honestly be; a layer can make the components traceable and their as-at dates visible, and it cannot make the appraisal daily. Live visibility does not create liquidity. The 2024 ODCE figures, USD 15.9 billion out against USD 6.5 billion in, would not have been different had every constituent fund reported perfectly; better reporting would have told LPs sooner what they were holding, which is the only promise this paper makes. A daily REIT price is not the truth either: the listed index fell by a quarter in 2022 and recovered part of it the following year, and a committee that treats the public mark as the correct one has swapped one assumption for another.

Transparency about uncertainty can itself close a door. Under the FCA’s rules, a fund whose valuer honestly reports material uncertainty over a fifth of its assets must suspend, so a manager that surfaces uncertainty promptly may be gated while a slower one trades on. That is a real cost of candour, and it argues for the rule to be revisited rather than for the candour to be withheld. LP-side capacity is finite too: an allocator with a hundred manager relationships cannot consume a hundred live feeds, and the layer’s value to that allocator is that the pack it does read can be checked.

The private market will keep answering a quarter late, and revising the answer, for as long as its numbers are assembled for a document. The public market answers daily for a plain reason: its numbers were built to be asked. The private side can build its numbers the same way without acquiring a daily price, and the LPs now writing the templates have said, in the only language a market listens to, that they intend to hold managers to it.