Every quarter, the investor report for a real estate fund gets pulled together from property manager data, fund accounting, and more spreadsheets than anyone would like to admit. Then, once it goes out, the supplemental requests start coming in. An investment consultant reviewing the fund for a pension plan sends over their own template. Another investor wants NAV broken out by property type or a metric not normally reported on.
I spent a few years in investor relations working on CRE investor reporting for several funds, and I also work on supplemental requests. Across both, the pattern hasn’t changed: it’s the same underlying data every time, just reshaped for whoever is asking.
It’s natural to see this as a formatting problem. Every LP wants something a little different, so the fix must be one standard template that everyone accepts. I thought about it that way for a long time too, but the more investor reporting I do, the more I think the format is only the symptom.
What it costs the GP
The time is the part everyone sees. Every custom request means pulling the data, reshaping it, checking it, and sending it out, and that work grows with each new investor rather than with the size of the portfolio.
What’s easier to overlook is consistency. The quarterly report itself is assembled from property manager data and spreadsheets. Supplemental reports then tend to get built one at a time, often by different people and at different points in the quarter, so they don’t always start from the same numbers.
Here’s something I’ve seen more than once. One LP asked for occupancy by region and another asked for occupancy by property type, and both breakdowns should have tied back to the 94% in the quarterly report.
But one was built from the quarter-end rent roll and the other from a rent roll pulled after close. One included a property still in lease-up, while the other left it out. One report showed 94%, and the other showed 92.5%.
No one did anything wrong. The two reports were just built separately and drifted apart. The problem is that the first person to notice is usually the LP or their consultant, and at that point you’re explaining a gap instead of building confidence.
What it costs the LP
Even with all that custom work, LPs often still can’t compare one manager with another because numbers that look the same on the page are frequently defined differently underneath.
IRR is a good example. A lot of funds use a subscription credit facility, borrowing against LPs’ uncalled commitments instead of calling capital right away. Because LP money goes in later and is invested for less time, the IRR comes out higher even if the deals performed exactly the same. That means two funds with similar results can report quite different IRRs depending on how they use the facility and whether they show returns with or without it.
Leverage is another one. Some managers measure loan-to-value against current appraised value and others against cost, and some count preferred equity as debt while others don’t. On a $100 million property with a $55 million loan and $15 million of preferred equity, leverage is 55% one way and 70% the other, and both are defensible.
When an LP is invested with ten or twenty managers, those differences add up, and pulling everything into one clear view of the portfolio becomes a project of its own.
It happens inside firms too
This isn’t only a problem between managers. It often shows up inside a single firm as well. Funds launched years apart tend to keep the definitions of whoever set them up, so one fund might calculate leverage one way and another fund a different way, with both being perfectly reasonable. Nobody decided to be inconsistent; the approaches just never got compared.
For an LP invested in more than one of the firm’s funds, that’s confusing, and even the firm’s own comparisons across its funds end up needing a footnote.
The industry is working on it
None of this is new. NCREIF and PREA have spent years building reporting standards and definitions for real estate funds, ILPA publishes reporting templates, and GIPS sets out how performance should be calculated.
The newest effort is REDI, the Real Estate Data Initiative. It was put together by a group of large LPs and consultants and gives managers a common data model and template for the fund, asset, and portfolio data investors ask for most. What I find notable is that it’s being driven by the people who receive the data.
The goal behind all of these is the same: one report with shared definitions that every LP can receive and compare.
Adoption is still the hard part, though. These standards are voluntary, and large LPs and consultants still have the leverage to ask for their own formats, so for now most GPs are working toward a standard while still answering custom requests.
Consistent definitions don’t guarantee consistent numbers
This is the part that’s easy to miss. A GP can have a clear, well-documented definition of NOI and still end up reporting numbers that aren’t consistent.
It helps to think of a metric as a recipe. The definition is the recipe, for example, NOI equals revenue minus operating expenses, excluding capital spending. The ingredients come from property managers, though, and each of them has their own chart of accounts and their own way of coding things.
Take two similar multifamily properties in the same fund, each run by a different property manager. Every time a unit turns over, there’s painting, flooring, and sometimes appliances to replace before the next tenant moves in. One property manager capitalizes those turn costs, so they stay out of NOI, while the other expenses them, so they reduce NOI.
The recipe is the same, but the ingredients aren’t, so the second property looks less profitable than the first even if they’re performing about the same. That difference then carries into the fund total and anything built on it, from margins to valuations.
Occupancy works the same way. Both properties might report 95%, but if one property manager counts a unit as occupied once the lease is signed and the other only once the tenant has moved in and started paying, the LP is looking at two identical numbers that mean different things.
Now picture a fund with 30 properties across several property managers on different systems, each with their own approach to turn costs, capitalization thresholds, management fees, bad debt, free rent, and one-time items. Each choice is small and reasonable on its own, but together they produce a fund-level number that’s hard for anyone to fully explain.
That’s why I don’t think the industry is short on standards. A template defines what a number should mean, but someone still has to get the property-level data there first.
Where AI Fits in CRE Investor Reporting
It wouldn’t be a 2026 blog post without a section on AI, so here it is and I’ll keep it practical. AI can take a lot of the manual work out of investor reporting, but only if the data underneath is in good shape.
As I wrote in my previous blog post, AI Investor Reporting in CRE: A Proven System That Works, AI is only as reliable as what you give it. If you point it at inconsistent data you’ll get inconsistent numbers back, just faster and with more confidence.
Point it at one cleanly mapped dataset, and it becomes much more useful. An AI agent can take a consultant’s template, fill it from the same numbers as the quarterly report, and run the tie-out, so the team spends its time reviewing requests instead of rebuilding them.
That’s the approach we take with CREx OS. The CREx team keeps each property manager’s data clean and mapped to one chart of accounts month to month, and AI agents work on that trusted record, so the data gets fixed first and the automation sits on top of it.
You don’t need the whole industry on board
A GP doesn’t have to wait for every LP and manager to adopt a standard. Most of the time savings come from getting your own inputs consistent, and none of it requires industry-wide agreement to get started.
Once that’s done, a request for leverage on cost instead of value isn’t a project anymore. It’s just a different calculation on numbers you already trust, and if you decide to adopt something like REDI later, most of the hard work is already done.
Before next quarter’s reporting cycle, it’s worth working through a few things:
- Lock one dataset each quarter. Once the quarter closes, the quarterly report and every supplemental request should come from the same finalized numbers, not a rent roll pulled after close or a spreadsheet saved on someone’s desktop.
- Map property-level data to one standard, so turn costs are treated the same at every property, and NOI and occupancy mean the same thing regardless of who manages it.
- Label every metric with its basis, for example, “leased occupancy” or “LTV on appraised value” rather than just “occupancy” or “LTV.”
- Add a definitions page to your investor reports so LPs can see how each number is built and understand any differences between reports.
- Set internal defaults so that when an LP doesn’t specify, your team knows which version to use.
- Tie every supplemental request back to the quarterly report before it goes out, so any gap gets caught internally rather than by an LP.
LPs will keep asking for their own formats, and that probably won’t change any time soon. But if the numbers underneath are consistent, every version you send will tell the same story.
If your supplemental requests are still getting built from different datasets than your quarterly report, that’s worth fixing before next quarter. CREx OS gives funds and asset managers one trusted record to report from every time. Let’s talk