Eight NAV REITs Declared August Distributions in the Same Week. We Read All Eight Filings — and the 'Net Distribution' Column Does Not Mean the Same Thing in Any Two of Them

Eight filings, one week, one format

NAV REITs — the perpetual-life, non-traded vehicles that replaced the old closed-end non-traded REIT — publish their distributions monthly on Form 8-K. Between August 27 and September 1, 2026, eight of the largest did it in the same six-day window.

They all use the same disclosure shape. One gross distribution, identical for every share class. Then a deduction. Then a net figure per class.

Here is every class of all eight, as filed, with the drag we computed:

Read that table the way most people will, and you reach an obvious conclusion: Invesco is the cheapest and Brookfield is the most expensive.

That conclusion is not supported by these filings, and the reason is in the second column.

The column that breaks the comparison

Look at what each sponsor actually subtracts.

Blackstone, Starwood, J.P. Morgan, Invesco and North Haven deduct one thing: the stockholder servicing fee. That is the ongoing payment to the broker-dealer that sold you the shares. It is a distribution cost, not an investment-management cost.

There is also a class that does not appear in the table at all. Blackstone's 8-K ends with a sentence worth more than it looks: "Class C currently has no distribution amount presented as it is generally an accumulating share class whereby its share of income will accrete into its NAV." A BREIT Class C holder receives no cash distribution; the income is retained in the share price. Brookfield also has a Class C, and it does distribute ($0.0465 net). Two sponsors, the same letter, two different products — which is the whole argument of this article, compressed into one character.

Apollo and Brookfield deduct two things: a management fee and a stockholder servicing fee. Their tables have four columns instead of three.

RREEF deducts a third thing entirely, labelled "Dealer Manager and Distribution Fees."

This is why the single most eye-catching number in the dataset is misleading:

Apollo's Class I is not losing a fifth of its distribution to a fee that Blackstone's Class I avoids. Both pay a management fee. Apollo shows it in this table; Blackstone charges it at the fund level, so it is already inside the gross figure before the table starts.

The practical consequence is blunt: a "net distribution" of $0.0557 from Blackstone and $0.0900 from Apollo are not measurements of the same thing, and any table that ranks sponsors by that column is ranking their disclosure choices.

We flag this specifically because it is the kind of error that survives. Both numbers are correct. Both come straight from an SEC filing. Nothing about either looks wrong. What differs is the label, and a label does not trigger anyone's alarm.

The eighth REIT breaks the pattern, and it is the most useful one here

Added September 4, 2026. North Haven Net REIT (SEC CIK 1999784) declared its August distribution on August 31, 2026, inside the same six-day window, and we missed it in the original set. Adding it does more than extend the table, because North Haven does the one thing the other seven do not.

The gross distribution is not one number. It is four, ranging from $0.1171 to $0.1569 — a 1.34x spread before a single fee is deducted.

That changes how the headline gap should be read. North Haven's best-to-worst net spread is 53.22%, which would place it second only to Brookfield in the table above. But most of that gap is not a fee. Split it and it multiplies out cleanly: Class E's gross is 33.99% bigger than Class S's ($0.1569 against $0.1171), and on top of that Class S loses 12.55% of its own gross to the servicing fee, which lifts the gap by a further 14.36%. 1.3399 × 1.1436 = 1.5322. Roughly three quarters of North Haven's class gap is distribution policy and one quarter is fee — and the filing does not separate them for you. (Our arithmetic, from the filed figures.)

Correction, September 18, 2026: this paragraph previously said "only 12.55 percentage points of that gap is the stockholder servicing fee". 12.55% is the fee as a share of Class S's own gross, which is a different quantity from its contribution to the class gap; the two are not interchangeable and the sentence mixed them.

This is the same trap as the cross-sponsor comparison, one level down: a "class gap" is only a fee measurement when the gross is held constant. For the other seven REITs it is, so their gaps are fee gaps. For North Haven it is not, so its gap is a fee gap plus a distribution-policy gap, and the filing does not separate them for you.

It is worth saying that North Haven is also the most forthcoming of the eight on the cost that never appears in any of these 8-Ks: the upfront selling commission. Its 10-Q for the quarter ended June 30, 2026 reports an inception-to-date total return for Class S of 7.62% excluding upfront selling commissions, and adds in a footnote that "[i]nclusive of the maximum upfront selling commissions, the inception-to-date total return for Class S shares and Class F-S shares are 5.7% and 5.6%, respectively." That is up to 3.5% of the transaction price, charged once at the door, on top of every ongoing fee in the table above — and the distribution 8-K, by construction, cannot show it.

Correction, September 18, 2026: this passage previously described the same numbers as "the commission drag as a return number, taking a 7.62% gross to 5.7%", with no source. 7.62% is not a gross figure: it is the Class S total return measured without the upfront commission, against 5.7% with it, and it lives in the 10-Q rather than in the distribution 8-K this article is built on.

What the data does support

Strip out the cross-sponsor comparison and there is still a large, solid finding, because within a single REIT the table is internally consistent — same deduction set, same month, same assets, and for seven of the eight the same gross.

Three patterns hold across all eight, and these are comparable because they are structural rather than numerical:

1. There is always a class that keeps 100%. It is Class I, Class E, Class N, Class L, Class Z or Class M-I depending on the sponsor. It is the class sold without an intermediary taking an ongoing cut — the adviser's own class, the institutional class, or the fee-based-advisor class.
2. The classes that lose the most are the ones sold through commission-based intermediaries — Class S, Class T, Class A. In every single one of the eight, the worst class is one of those.
3. The gap is not small. The narrowest of the eight, Invesco, still hands its Class I holder 15.67% more income per share than its Class T holder. The widest, Brookfield, hands over 54.83% more.

None of this is hidden and none of it is improper. Servicing fees are how commission-based distribution gets paid, they are disclosed in every prospectus, and an investor who reached the product through a broker received something for that fee. The point is that the cost is per-class, permanent for as long as you hold, and invisible in the headline distribution rate that gets quoted.

The J.P. Morgan waiver, which is genuinely unusual

One filing contains something the other seven do not. J.P. Morgan's 8-K states that the Class D stockholder servicing fee was waived for August 2026, and then goes further: the NAV attributable to current holders of Class D shares "will not be included in the computation of stockholder servicing fees charged on Class D shares in perpetuity."

That is why JPMREIT's Class D shows 100% in the table above while Blackstone's Class D shows 94.43% and Starwood's 94.68%. It is a real, permanent difference for that specific cohort of holders — and it applies to current holders, not to whoever buys Class D next year.

We note it because it is the one place in this dataset where a class-level number differs for a reason that is not structural, and a reader comparing Class D across sponsors would otherwise draw the wrong conclusion from it.

How to use this if you own one of these

The arithmetic that matters to you is not the sponsor ranking. It is one line in one filing.

1. Find out which class you actually hold. It is on your statement. Most people sold through an intermediary hold Class S, T or A.
2. Open that REIT's latest distribution 8-K on EDGAR and read the row for your class. The gross and the net are both printed.
3. Divide net by gross. That is the share of the declared distribution you actually receive, this month, for as long as you hold that class.
4. Ask whether a 100% class exists and whether you can convert into it. These programmes generally cap the servicing fee at a percentage of what you originally paid and convert the shares to a class with no servicing fee once the cap is hit — but the cap, the trigger and the destination class are set by each prospectus, not by a rule, so the only answer that counts is the one in your REIT's. Whether the conversion has happened, and when it will, is a question for the sponsor with a specific answer.

How this was checked. Every figure here comes from the sponsor's own Form 8-K, listed by accession in the dataset. On September 18, 2026 the whole table was re-verified by parsing the nine filings and comparing each class row to ours: 56 of 56 match on gross, net and the share of gross kept, and every net equals its gross minus the deductions the filing itemises. No class present in any of the nine filings is missing from our table. Three things in the surrounding text did not survive that pass and are corrected above, each with a note: the decomposition of North Haven's class gap, the labelling of its 7.62% / 5.7% return figures, and a count of sponsors in the summary block.

The one comparison to avoid is the one this article exists to prevent: do not rank these eight by the net column. It is the same trap as comparing two funds' "expense ratios" when one of them capitalises a cost the other expenses. The numbers are right; they answer different questions.

Sources

Every figure is taken from a Form 8-K filed with the SEC between August 27 and September 1, 2026, and every deduction was checked against the printed net for each class of each REIT before publication. All eight reconcile to the cent.

• Blackstone Real Estate Income Trust, Inc. — CIK 0001662972, accession 0001662972-26-000130, Item 7.01, filed August 31, 2026.
• Starwood Real Estate Income Trust, Inc. — CIK 0001711929, accession 0001193125-26-375795, Item 8.01, filed August 31, 2026.
• J.P. Morgan Real Estate Income Trust, Inc. — CIK 0001893262, accession 0001628280-26-059501, Item 8.01, filed August 31, 2026.
• Invesco Real Estate Income Trust Inc. — CIK 0001756761, accession 0001756761-26-000122, Item 8.01, filed August 31, 2026.
• Apollo Realty Income Solutions, Inc. — CIK 0001882850, accession 0001193125-26-376677, Item 7.01, filed August 31, 2026.
• Brookfield Real Estate Income Trust Inc. — CIK 0001713407, accession 0001713407-26-000090, Item 7.01, filed August 28, 2026. Covered in detail in our Brookfield share-class analysis.
• RREEF Property Trust, Inc. — CIK 0001542447, accession 0001542447-26-000108, Item 8.01, filed September 1, 2026.

The share-of-gross percentages, the highest-to-lowest ratios and the cross-sponsor comparability analysis were computed and written by CrowdfundedWealth from the filed figures.