The raise: $175.6 million. The assets: two Sunbelt apartment communities totaling 612 units. Both properties came out of BREIT’s portfolio (Blackstone’s $120 billion-plus non-traded REIT platform).

BREIT-affiliated entities previously owned the buildings. A Blackstone affiliate now leases them from the DST, BREIT's operating partnership guarantees the rent, and the REIT's operating partnership may eventually acquire the properties through a tax-deferred transaction.

Subscribe now

Blackstone is on quite a few sides of this deal. Sorta like this:

That doesn't necessarily make it a bad investment, of course. In fact, for the right investor, a DST can solve a very real problem: how to sell an appreciated rental property without immediately paying capital gains taxes or spending the next decade dealing with tenants.

You are all big boys and girls and know there is no free lunch. Investors give up control, accept substantial illiquidity, and potentially pay several layers of fees.

And when the sponsor supplies the property, structures the investment, collects the fees, and potentially determines the exit, it's your job, dear investor, to understand exactly what you are buying (if you need a refresher, start here).

Let's look under the hood.

This article is for educational purposes only and isn’t tax, legal or investment advice. Talk to your own tax advisor before making any 1031, DST or 721 decision.

First, why would anyone buy a DST?

Imagine you've owned an apartment building for 20 years.

You bought it for $1 million. It's now worth $3 million. You've collected rent, depreciated the property, and accumulated a substantial unrealized gain (along with substantial brain damage from dealing with plumbing emergencies, roof leaks and your local legislators).

You're ready to retire from the landlord business.

Selling outright is simple, but expensive. Depending on your tax basis, depreciation history, state of residence, and income, a meaningful portion of your proceeds could go to taxes.

Section 1031 of the Internal Revenue Code offers an alternative: exchange your investment property for another qualifying real estate investment and defer recognition of the gain.

It isn’t the only deferral tool. We looked at Opportunity Zones here:

⏰ The problem is the clock.

You generally have 45 days to identify replacement property and 180 days to complete the exchange, subject to the applicable tax-return deadline. That's not much time to find, underwrite, negotiate, finance, and close on another building.

And after all that, congratulations! You're still a landlord.

Enter the Delaware Statutory Trust.

Under IRS Revenue Ruling 2004-86, a properly structured DST interest can qualify as an ownership interest in real estate for Section 1031 purposes.

Here’s how it works:

• A sponsor acquires properties, packages them into a trust, and sells fractional beneficial interests to investors completing 1031 exchanges.

• You buy an interest in a trust that already owns properties (instead of buying another building outright).

The properties are professionally managed, investors receive their share of income, and the tax deferral can continue. You get out of managing real estate without getting out of owning real estate. That's a pretty compelling proposition, am I right??

Why investors choose DSTs, and what they surrender

It’s not all unicorns and roses with DSTs. There are very clear benefits, but they come with some drawbacks:

Read more

But income statements tell you how much. They don’t tell you which loans, or why.

So for Part 2, I went into the loan books themselves (even more fun): every position in every schedule of investments, for five quarter-ends from June 2025 to June 2026.

Subscribe now

Three things for y’all (there are more, but I’m told good writers only highlight three at a time):

1. Existing borrowers are switching to PIK far more often than new borrowers are arriving with it. For every borrower that came in with PIK, about 2.5 existing borrowers switched from paying cash to paying in kind.

2. The switches (cash —>PIK) are still marked near par. The median switched loan was marked at 99.6 cents. But they drift lower afterward.

3. The overall PIK book is growing, and its marks are slipping. Debt with a PIK component rose from $23.7 billion to $28.4 billion at cost, while fair value fell from 96.8% to 94.3% of cost.

4. Perhaps the most important finding is in the footnote to this article. I know, I know, it’s the fourth thing, but I’ve long ago given up any aspirations of becoming a good writer (I just want you to get in the habit of reading the dang footnotes).

Of course, the backdrop for non-traded BDCs is the outflow of capital. In Q2, redemptions at non-traded BDCs outran new money for the first time since at least 2022, according to PitchBook. You all are highly intelligent people and know that PIK income cannot fund a cash redemption. We’ll discuss the implications below.

Missed Part 1? Start here:

Switches > new PIK

Let’s start with new PIK’ers.

The number of borrowers paying some interest in kind rose from 396 to 492 across the 15 funds in the year to June 2026. Most of that growth came from existing loans previously paying cash.

Over four quarters, 200 existing borrowers moved from cash-pay to PIK. Only 81 new borrowers arrived with PIK built in. Another 16 came back from the dead 🧟 non-accrual as restructured loans that now pay in kind.

Going the other way, 126 borrowers went back to paying cash (yay for them), 57 left the portfolios (sold? repaid? refinanced?) and 18 moved to non-accrual (the opposite of yay, - for the lenders).

Let’s zoom out to the big picture:

• about 11.5% of these funds’ loans, by cost, carried a PIK feature at June 30, 2026 (unchanged from a year earlier);

• but only about 5.4% of the loans to new borrowers over that year came with PIK attached from the start.

Lincoln International reported that up to 25% of middle-market debt was originated with a PIK provision (or “PIK toggle”) at the end of last year, according to the WSJ. The figure fell to 13.5% in Q2.

If Lincoln’s figures are measured on a comparable basis, the gap suggests that many borrowers have PIK provisions but don’t use them until later in the loan’s life. That makes perfect sense, doesn’t it? A PIK option is valuable to a borrower: imagine if your mortgage let you skip part of your monthly payment and add it to the balance instead.

Can we tell toggles from modifications?

The filings don’t make that easy. Personally, I would love to see the callout for which is which, but I don’t make the rules.

Here’s what Q2 2026 data tells us:

In that quarter, 51 loans across the 15-fund cohort started paying some interest in kind.

• In 17 of the loans (~37% of the dollars), the terms were reworked, which looks like a modification.

• Another 21 picked up a smaller step-up, usually 50 basis points, with the maturity unchanged. That’s what a PIK toggle typically costs the borrower, so these could be either.

• Of the rest, 13 kept their terms unchanged.

Some other fun reading for your enjoyment:

Marked near par

Logically, a borrower that stops paying all its interest in cash may be facing tighter cash flow than when the loan was made, regardless of whether this was a toggle or a modification.

Whelp, you wouldn’t necessarily know it from the marks:

Read more

Two years ago, CRE CLOs held the largest share of delinquent securitized multifamily debt.

Today, private-label CMBS holds half of it, despite accounting for just 12% of the outstanding balance.

Total delinquent balances climbed from $7.0B to $11.6B over the same period.

The distress is still there, and the CRE CLOs that look like they’ve cleaned up their act deserve an asterisk. (More on that below. Private markets do love an asterisk)

In this issue:

1️⃣ Private Credit: Apollo marks daily, the same loan gets different values depending on who holds it, and non-traded BDCs paid out more than they raised.
2️⃣ Private Equity: General Atlantic sells a continuation vehicle to a continuation vehicle 🫠
3️⃣ Commercial Real Estate: where multifamily distress is moving.

Accredited Insight delivers the LP’s perspective on private credit, private equity, and CRE, drawing on hundreds of deals reviewed, and thousands of conversations. Paid subscribers gain access to our database of over 40 case studies and articles on everything from evergreen funds to due diligence (the kind of analysis that tells you what the GP pitch deck left out).

Subscribe now

1️⃣ Private Credit

More marks, same information.

Apollo has started rolling out daily valuations across its roughly $850B credit business. Not everyone is thrilled. At the CAIS summit this week, Monroe’s Ted Koenig called daily marks on illiquid assets “an illusion of precision, not accuracy.”

Here’s the problem: a private loan doesn’t produce new information every day. The borrower sends financials quarterly, so between those reporting dates, a daily mark is a model’s output (and the model belongs to the manager).

You get more numbers, not more inputs. And most certainly not more precision. The analogy is a weather forecast that only updates once a week: you can refresh the page all you want, the result won’t change much.

Also worth mentioning: most of Apollo’s daily-valued book is investment grade. Which is not exactly the part of private credit keeping everyone up at night.

On that note, I wanted to share two new academic papers that explain why the who matters more than the how often.

Andrade and Lie looked at 4,655 loans held by multiple BDC sponsors and compared marks on the same loan, on the same date (you’ll find an amazing chart at the very end of the post, or by clicking here)

BDCs managed by the same sponsor mark a shared loan almost identically, with an average difference of just 0.026 cents. Across different sponsors, the gap is 1.13 cents (sounds like peanuts, but it’s more than 40x wider).

We’ve already seen this, of course. Remember Medallia?

What helps predict the gap? PIK (are you shocked, or what?)

Alexandra Dogaru (Berkeley Haas) reaches a similar conclusion from a different direction. Funds more reliant on PIK income mark identical loans higher than their co-holders, even when those loans pay in cash.

Her explanation comes down to incentives. PIK accrues into the income used to calculate incentive fees before a dollar is collected. Meanwhile, markdowns elsewhere shrink NAV and the fee base.

‼️Pay attention: the effect is statistically indistinguishable from zero in audited year-end reports (LOL, LMAO even). Apparently, the marks behave differently when someone else checks the homework.

Andrade and Lie have another finding that’s worthy of at least one❗️: sponsors that mark more conservatively are the ones whose funds have to buy back shares at NAV. Turns out, when your optimistic valuations will cost you in the price you pay investors who want out, you get more disciplined..

Those same non-traded BDCs hold 4.2 percentage points more of their assets in cash and quoted loans than their listed siblings (which makes sense because unlike their listed siblings, they have to pony up the cash to buy shares of exiting investors quarterly).

So what’s the stick that incentivizes a disciplined mark? Having to pay it, and having an auditor check it. And by the way, publishing the number daily does neither.

The flows flipped

For the first time since at least 2022, more money left non-traded BDCs than came in. PitchBook tracked 27 funds in Q2: $3.3B of new shares against $8.1B paid to exiting shareholders.

Fundraising fell to a multiyear low, from $13.4B a year earlier.

The 12 largest funds paid $7.6B against roughly $19.1B requested. BCRED alone received about $6.7B in requests, paid $3.2B, and took in $457M of new money.

Here’s why everyone heads for the exit at once:

Q3 requests eased a bit: OCIC at 16.8%, Apollo Debt Solutions at 14.7%, HLEND at 11.5%, and BCRED around 10%. The outlier is Blue Owl Technology Income ⁠Corp., or OTIC, where investors sought to withdraw $1.1 billion, equal to 39% of shares, up ​from 38.1% in the prior quarter (Reuters).

2️⃣ Private Equity

A continuation vehicle for a continuation vehicle

In 2021, General Atlantic moved four companies into a $3B+ continuation fund led by Ardian and HarbourVest. One of them, almost certainly Argus Media, is now headed for a second continuation vehicle.

GA filed for “GA Continuity II (AM)” on October 7, and Secondaries Investor reports Leonard Green’s Sage fund as lead buyer. The 2021 secondary buyers now get to choose: sell to Leonard Green or roll again.

The secondary market has developed a new business model: buy a continuation vehicle, hold the assets, and eventually sell them to another continuation vehicle. The industry calls these trophy assets.

Buyout firms are now raising dedicated funds to buy each other’s continuation vehicles. Leonard Green’s Sage closed at $3.6B. Hg just raised €1B at first close of Orion, targeting €1.75B, its first fund to buy other sponsors’ software CVs. GTCR, New Mountain, and Golub have similar vehicles.

How secondaries buyers make money on the discount:

None of this makes a CV a bad deal, of course. Sometimes a good asset deserves more time. But PitchBook found that most LPs choose to cash out of continuation vehicles rather than roll. The reasons often include fund life, investment mandates, and the transaction process (not necessarily a lack of confidence in the asset).

Which creates an interesting conundrum: the GP says the asset is too good to sell. Existing LPs mostly choose to sell. And often, the buyer is another buyout fund with a dedicated pool of capital for buying assets from other buyout funds.

3️⃣ Commercial Real Estate

Trepp tracks $636B of securitized multifamily debt. In August, $11.6B was delinquent. Of that, $5.8B sat in private-label CMBS, a 7.69% delinquency rate.

Freddie Mac carries 82% of the balance but has a 0.80% rate. CRE CLOs carry $41B at 3.88%.

Two years ago, CRE CLOs were the problem child, accounting for 40% of delinquent debt at a 6.05% rate. Today, they account for just 14%.

So, did CRE CLOs clean up their act?

Partly, the loan mix changed. CLO debt on pre-1980 buildings fell from $16.2B to $6.8B, while private-label debt on those same vintages rose from $20.4B to $29.1B.

But there is another wrinkle. CRE CLO managers can pull defaulted loans out of the pool, and they have reasons to do so. If a deal fails its over-collateralization test, cash that would otherwise go to equity and junior noteholders gets diverted to pay down senior notes.

A delinquent loan leaving the CLO can therefore change the reported numbers and the deal’s cash-flow dynamics. It does not, however, magically make the loan perform.

So when we read delinquency data, a falling rate can reflect better credit performance, a changing loan mix, or a change in what’s still being counted. Those are three very different stories.

The broader trend hasn’t turned, either. Trepp’s overall CMBS delinquency rate hit 8.02% in September, the highest since late 2020, while CMBS multifamily rose to 8.04%.

That’s it for this week.

If someone forwarded this to you, subscribe. If you’re the one who forwards it, thank you.

Share

-Leyla

P.S. Here’s the promised chart:

P.P.S. New here?

Here’s where you’ll find the full archive (somewhat organized):

On its own, that says little about credit quality. Payment-in-kind interest isn’t inherently bad. Many loans are written with a PIK component from day one, and some amendments that switch a borrower to PIK may have merit. A rising PIK number is not proof of distress in and of itself.

Subscribe now

But in private credit, LPs see very little. They don’t get borrower financials, the marks come from the managers, and a loan can sit near par long after the business behind it starts to struggle. Against that backdrop, PIK is one of the few signals that is consistent, observable and comparable across funds. It ain’t a crystal ball, but it’s about the only thing we get to see..

Here are two good case studies on the subject:

So I pulled eight quarters of income data for the 15 largest non-traded BDCs (and wore out my Ctrl+F keys searching for “PIK”)

PIK is rising while cash interest is flat

The cohort’s PIK interest rose from $186M in Q3 2024 to $314M in Q4 2025, dipped 12.8% in Q1 2026 (more on that below), then recovered to $292M in Q2.

Meanwhile, quarterly cash interest income has hovered between $5.35 billion and $5.40 billion for three consecutive quarters.

To be fair, cash interest income isn’t a perfect measure of portfolio performance, either. It depends on portfolio size, loan repayments, new originations and base interest rates.

But when we compare the two over a full year, the gap is clear. Over the 12 months to June 2026, PIK income grew faster than cash interest at 10 of the 15 funds.

PIK’s share of interest income is rising

You may have noted the blue column chart above is in absolute dollars. And dollar growth can simply mean bigger books. To control for that, we need to look at PIK as a share of total interest income.

PIK made up 5.16% of the cohort’s interest income in Q2, up from 4.56% in Q3 2024. We do have a behemoth in our sample: Blackstone’s BCRED holds about a third of the cohort’s NAV. Without BCRED, the rest of the group stands at 4.86%, the highest in the eight quarters I tracked.

‼️The non-traded group still runs below listed BDCs, where PIK represented 8.2% of interest income as of Q1 2026, according to PitchBook.

The Q1 dip, and why it isn’t good news

Q1 warrants a closer look, because it shows the indicator moving in the “right” direction for the wrong reason.

Read more

Mellon had spent the late 1970s and early 1980s lending into the oil patch, and into real estate ventures in Texas, Louisiana and Colorado that oil money had inflated (among other lending decisions that would come to haunt them). When energy prices collapsed, the loans blew up. By the end of 1987, Mellon had booked a record loss of $844 million.

The board went looking for outside help and brought in Frank Cahouet, a former Fannie Mae president who had run Crocker National in San Francisco. He arrived in June 1987 and promptly cut about 3,000 jobs. Then he proposed moving the bad loans out of the bank entirely, a structure now widely cited as the first good bank/bad bank split.

Subscribe now

In July 1995 Grant Street paid its last proceeds to bondholders and liquidated. Mellon, freed of the drag of non-performing assets, reinvented itself as a money manager and eventually became half of BNY Mellon.

The end.

Leyla, why TF the heck are you telling us about a bank?

Because I love a good banking story, ofc. Aside from my love of banks and good stories, there is a VERY IMPORTANT POINT, so pay attention. What I’m about to describe is a solution we’ll see more of, according to my crystal ball.

To be perfectly frank with y’all, a hedge fund side pocket is the better analogy (more on that below). But I like banks better than hedge funds, so the banks go first.

Fast-forward 38 years. The same structural idea has migrated from commercial banking into private equity, minus the price discovery.

Disclosure: This article is provided for educational and informational purposes only and should not be construed as investment, legal, tax, or financial advice. It is based on publicly available information, including Partners Group’s press releases and published media reports, as of October 9, 2026. The proposed restructuring of Partners Group Global Value SICAV is subject to shareholder approval, and its final terms have not been made public; details described here, including the reported allocation between sub-portfolios and redemption limits, may differ from the terms ultimately adopted. References to “bad banks,” “side pockets,” and academic research are analogies and frameworks for analysis, not claims about Partners Group’s intentions, conduct, or the future performance of any fund. The views expressed are solely those of the author. The author holds no position in Partners Group or its funds. Readers should conduct their own independent research and consult qualified professionals before making any investment decision.

Meanwhile, in Switzerland

The setting is different. Mellon held a distressed loan book in a banking crisis; Partners Group runs an evergreen fund facing elevated redemptions. What’s similar is the fix: split the old from the new, and release each into the wild.

On June 3, 2026, Partners Group notified investors in its flagship Luxembourg evergreen vehicle, Partners Group Global Value SICAV, that it would cap quarterly redemptions at 5% of Net Asset Value (NAV). Redemption requests had hit ~9.8% of NAV for the second quarter (most of the requests came from investors in Asia and Australia).

The next day, the firm said redemption pressure that started in private credit had spilled over into private equity. It flagged a second, Delaware-domiciled evergreen vehicle where tender requests were running at about 6% of NAV, above its 5% offer. The firm didn’t name it, but the description fits the Partners Group Private Equity (Master Fund), which I wrote about the same week:

On October 2, 2026, Partners Group presented its structural fix. Subject to a shareholder vote, the €6.6 billion SICAV would be reorganized into a Luxembourg umbrella structure containing two separate sub-portfolios, split by when the assets were acquired:

• Distributing Fund: the legacy, older assets, which will be realized and sold down over time to meet redemption requests.

• Compounding Fund: holds the newer investments, open to fresh subscription inflows, backed by an additional balance-sheet commitment from Partners Group alongside new institutional capital.

The Financial Times reports that existing investors would see roughly 75% of their exposure land in the Distributing Fund and 25% in the Compounding Fund, each with its own 5% quarterly cap. The FT’s source described the older assets as the weaker performers.

Partners Group frames the move as setting the strategy up to keep compounding returns. Nobody is using the words “bad bank.” Yet the architecture is unmistakable.

A “Bad Bank” Without a Market Price

The motive behind Partners Group’s restructuring isn’t all that different from Mellon’s in 1988: wall off the legacy book so the clean vehicle can raise new money. That’s the good bank’s whole reason to exist, and Partners Group says the Compounding Fund has already drawn interest from new institutional investors.

This structure, however is materially different from the good bank/bad bank split in several ways:

Read more

If you want to invest in actual physical data centers, Blue Owl has a vehicle for you.

It’s called Blue Owl Digital Infrastructure Trust, or ODIT. It’s a perpetual-life, non-traded REIT (~$2 billion of NAV), available to accredited investors.

Subscribe now

The pitch is easy to understand. Per its Q2 10-Q, as of June 30, 2026:

• 12 data centers totaling 235 MW

• 99.7% leased

Per the same filing, Class I shares returned 5.93% in the first half of 2026, and 12.24% annualized since inception.

Not bad, eh?
(contrary to rumors out there, I’m decidedly not Canadian, btw).

So here’s what we’ll cover today:

• The assets. Data centers are the new sexy thing, but at the end of the day, a data center is a box with contractual cash flows. We’ll cover the unglamorous stuff: cap rates, yield, rent increases, NNN vs gross leases.

• GAAP losses vs. NAV increases. Your favorite part: accounting nuances. (There ought to be some “GAAP and NAV walk into a bar” jokes).

Back to that ~6% YTD return: over the same six months, ODIT reported a GAAP net loss of $73.3 million.

Both numbers come from the same filing, and both are accurate. But only one of them sets the price you buy and redeem at, and the fees you pay. That’s NAV.

So where did the 6% come from? Let’s follow the money, starting with how the portfolio got here.

Disclosure: This case study is provided for educational and informational purposes only and should not be construed as investment, legal, tax, or financial advice. The views expressed are solely those of the author. All examples are illustrative in nature and not guarantees of future outcomes. Readers should conduct their own independent research and consult with qualified professionals before making any investment or financial decisions.

Where the Portfolio Came From

Before we get to the returns, we need to take a look at what ODIT actually bought.

Per the 10-K, ODIT’s first 11 buildings arrived as a package on December 1, 2025, from the BODI I Funds, which are managed by an affiliate of ODIT’s own adviser.

The transaction was about $3.29 billion gross. ODIT assumed $1.8 billion of debt, paid $1.48 billion in cash from its day-one offering, and issued about $42 million in shares and OP units, net of working-capital adjustments.

In other words, new money entering a Blue Owl evergreen vehicle bought assets from an existing Blue Owl vehicle.

Importantly, each side obtained a fairness opinion. BODI I’s advisory committee and ODIT’s board, including its independent trustees, approved the deal.

Many of these buildings predate ODIT by years. Per industry press, IPI Partners bought the Infomart and T5 sites in 2018–2019 and rebranded them as STACK Infrastructure. Blue Owl acquired IPI’s business in January 2025. Eight of the 11 legacy buildings were built by others between 2006 and 2015.

These weren’t shiny new buildings, and the leases they came with tell an interesting story.

Read more

The developer is a unit of Blue Owl Capital. The notice is about power: the project’s natural gas supply has been tangled up in permit denials. According to the Wall Street Journal, Oracle’s lease is “hell-or-high-water”: it can’t be terminated, and Oracle must pay rent whether or not the campus has power. The force majeure provision can delay the start of full rent payments by up to three years, after which the lease extends automatically.

Subscribe now

Oracle’s shares fell more than 3% that day. Blue Owl’s shares also declined. And the $18 billion of construction debt behind the project was already being quoted below 90 cents on the dollar.

Both companies moved quickly to calm things down. Oracle said force majeure notices are “commonplace in developments of this scale.” Blue Owl said the notice “does not change the financial commitments” to the project.

The Financial Times reported a key detail: even if rent is delayed, Oracle still owes a “carry cost” that covers interest to the lenders and a set return to Blue Owl’s equity. The WSJ described it differently: Blue Owl’s roughly 9% return during construction comes out of the money raised to build the project, and the financing set aside funds for up to three years of delays before lease payments begin.

Oracle, of course, is not the borrower [insert shocked Pikachu emoji]

The $18 billion construction loan sits with four companies called Red Chiles (there are Green Chiles in this story, too. Keep reading). The power plant’s financing sits somewhere else. The equity sits somewhere else again. And some of those pieces ended up in funds that individual investors can buy.

So who actually owns the risk?

I don’t have a dog in this fight. Build all the data centers you want, and if you want to invest in AI, more power to you (pun very much intended). I would actually like this project to succeed: at the moment, I slightly prefer Claude, but heck, Anthropic needs strong competition.

I was curious, and wanted to see what I could find. And find, I did.

How I went looking

I won’t be the first to tell you that data-center financing is complicated. These projects sit at the intersection of real estate, massive capex and hype. And your girl is a sucker for punishment. In other words, we are a match made in heaven.

I started where any reasonable person would: searching disclosed fund holdings for “Project Jupiter” and “Oracle.” That gave me a big fat zero relevant results, plus two unrelated deals that happened to share the codename. (Apparently every dealmaker’s favorite planet is Jupiter).

Now, remember, aside from imaginary things (future demand and pro-forma revenues) the project involves a very real tract of land. And land, of all things, gets recorded at the county. So I went to the Doña Ana County Clerk’s records and found Ordinance 367-2025, recorded September 22, 2025, as instrument #2520581.

The document was a goldmine of entity names:

• Yucca Growth Infrastructure, LLC.

• Red Chiles A, B, C and D, LLC.

• Green Chile Ventures LLC.

None of them say “Oracle,” “Blue Owl” or “Jupiter” on the tin, which explains why my first search found nothing.

I then traced those names through state business registries and SEC filings, with much better results.

Today I’ll tell you the story of this project. It’s incomplete, and I’ll tell you why. But I’ll show you what I found and exactly where I found it.

Three numbers that aren’t the same number

Let’s go through the numbers. We’ll start with the headline: a $165 billion data center (NYT likened it to “the fever dream of a modern pharaoh”). That number comes from the bond ordinance, and it isn’t $165 billion of financing.

It’s a ceiling on industrial revenue bonds, a standard New Mexico tax-incentive tool. The county takes title to the property, which takes it off the property-tax rolls, and leases it back to the developers for 30 years.

The county gets payments in lieu of taxes: $360 million over 30 years, or $12 million a year.

So the $165 billion is a tax structure. The ordinance splits it three ways: up to $15 billion for Yucca (power), $25 billion for Red Chiles (the buildings) and $125 billion for Green Chile (the tenant's equipment), and each series is bought by the company's own affiliate.

The actual project financing is much smaller:

• About $18 billion of construction debt, from more than 20 banks, priced at SOFR plus 2.5% (FT).

• About $3 billion of Blue Owl equity, according to Reuters.

Follow the names

Now we are going to play Hansel and Gretel (I play the part of Gretel, not the Witch). We are going to use the LLC name breadcrumbs to find our way. And yes, I'm deliberately using breadcrumbs, not pebbles. I'll tell you why at the end.

Start with the oldest. BorderPlex Manager, LLC was formed in Delaware on November 15, 2024:

Read more

Now, on to the part you won’t like: Fitch’s most recent private credit default rate:

Nearly half of what Fitch calls a default is a borrower that stopped paying cash interest (and switched to PIK, cause we all love not paying in cash, amirite?)

Subscribe now

In this issue:

1️⃣ Private Credit: redemption pressure is becoming K-shaped, while AI becomes the credit market's biggest borrower.
2️⃣ Private Equity: Partners Group considers splitting its evergreen fund into good assets and old assets.
3️⃣ Commercial Real Estate: 2021 bridge loans meet a higher-for-longer rate path, and an AAA tranche gets hosed.
4️⃣ Regulatory: hide your children. "Responsible retailization" is upon us.

Accredited Insight delivers the LP’s perspective on private credit, private equity, and CRE, drawing on hundreds of deals reviewed, and thousands of conversations. Paid subscribers gain access to our database of over 40 case studies and articles on everything from evergreen funds to due diligence (the kind of analysis that tells you what the GP pitch deck left out).

Subscribe now

1️⃣ Private Credit

Is three quarters of data enough to call it?

Across the nine largest non-traded private credit BDCs, only two show requests below 5% in Q3:

One important caveat: those headline percentages overstate new exit demand. When a fund prorates redemptions, the same investor can keep submitting requests quarter after quarter.

So the queue becomes self-sustaining: a capped fund with persistent outflows creates its own next-quarter redemption requests. The line doesn't really disappear until requests fall below the cap.

➡️ Which makes Oaktree's and Goldman's funds interesting outliers (the upper part of the K-shaped evergreen PC recovery). Is the difference better communication around liquidity? A different investor base? Something else? Meaningfully better portfolios? (I'm asking, the comment section is yours).

The AI buildout is now the credit market's biggest borrower

Apollo's latest deck has an interesting numAber: hyperscalers have issued nearly $250 billion of investment-grade debt in 2026 so far, more than the cumulative total of the prior decade.

AI-related issuers accounted for 54% of net IG issuance through August, up from 34% last year and effectively zero in 2024.

2️⃣ Private Equity

Partners Group is considering splitting its capped evergreen fund into good assets and old assets, according to the FT.

Global Value SICAV has been capping redemptions since June, when requests hit roughly 9.8% of NAV.

The proposed solution, subject to shareholder approval: turn the €6.6 billion vehicle into an umbrella with two portfolios:

• A “distributing” fund holding the older, weaker-performing assets and meeting redemptions by selling them.

• A “compounding” fund holding newer investments and targeting higher returns.

Existing investors would start at roughly 75% distributing, 25% compounding. The 5% cap is expected to apply to both. New money goes only into the compounding side.

The logic is straightforward: don't force a mature fund to sell its most liquid, best-performing holdings to satisfy redemptions and leave everyone else holding the leftovers (what I call “the bagholder problem”).

It creates a clean vehicle for new subscriptions while preserving the existing redemption queue.

➡️ This is worth watching, IMO. It’s an interesting solution to a common problem for mature evergreen PE vehicles.

Here are two case studies (both different funds from the one discussed above):

3️⃣ Commercial Real Estate

On September 16, the FOMC raised the funds rate 25 bps to 3.75–4.00%, the first hike since July 2023. Core PCE has remained above 3% throughout the year, and the median dot still showed one more increase before December. Every floating-rate borrower waiting for relief got a little less relieved.

👉The 2021 bridge loans are doing exactly what 2021 bridge loans were expected to do (and regular readers will find none of this surprising, see the date on this article):

Rate caps are expiring into a hiking cycle, and in-place income isn't close to the underwriting. The WSJ reports that Ashcroft Capital's five-property Georgia and Texas portfolio averaged Q1 rents of $1,423 against the $1,953 the lender projected. Income covers 57% of the payment.

And then another SASB AAA tranche gets in trouble:

From Bloomberg: Centre Square, the twin towers at 1500 Market Street in Philadelphia, was appraised at $471 million in 2019. Last month, a judge approved its sale for $70 million. (That’s an 85% decline, for those of you who are not good with doing math this early in the morning).

The $368 million CMBS deal behind it was a single-asset, single-borrower transaction that closed weeks before Covid, when the complex was 93% leased. By the end of June, occupancy was 28%. The borrower couldn’t refinance at the 2022 maturity.

The formerly AAA class is expected to recover roughly 44 cents, while seven junior tranches are wiped out. PIMCO, the largest holder, is looking at a loss north of $35 million.

This is the third AAA hit since the financial crisis, after 1740 Broadway and the Palisades Center mall. All three were SASBs.

4️⃣ Regulatory Updates

Meanwhile, the SEC proposes to widen the door.

On September 30, the Commission voted out a package it calls “responsible retailization” (formerly known as “democratization”).

The proposal would allow advisers to charge performance fees to registered funds, capped at 20%. It would also modernize the interval fund framework so repurchases can better match a portfolio’s liquidity profile, and create a rules-based path for multi-class closed-end funds.

➡️ Separately, the SEC is taking comments on new routes to accredited status: a FINRA-built exam, plus the CPA, CFA, CFP and Series 79/86/87.

The proposed exam would run roughly 75 multiple-choice questions over two hours, cost about $100, be open to anyone 18 or older, and stay valid for ten years. The largest section (20–28% of the questions) covers investment risk, explicitly including redemption restrictions and the impact of fees on net returns.

I’ll spare you the full extent of my skepticism here, - and I say that as someone who put the word Accredited in the title of a newsletter. Widening the pool is good for business, where I’m concerned.

On that note: unlike the funds, I don't check your investor status at the door. Subscribe whether or not you're accredited.

Subscribe now

What I’ll say instead is what’s missing. If the SEC wants more retail money in private funds, the package should come with expanded disclosure requirements for any fund sold to them.

‼️ Let's crowdsource the comment file. What would you add? ‼️

Standardized disclosure should include:

• standardized fee presentation (including underlying funds)

• valuation methodology, in writing (including underlying funds) - a girl can dream

• quarterly fund financials (I’m talking to you, Regulation D vehicles)

• redemption history: requests submitted versus requests filled, by quarter, going back at least three years

• all-in leverage, including fund-level credit facilities and subscription lines, with their effect on reported IRR

• full structural map — every SPV, holdco and feeder in the chain, what sits in each, and what each is obligated to; all off-balance-sheet commitments (guarantees, capital calls, backstops, unfunded obligations); and any securitization of fund assets, stated plainly, including what was transferred, what was retained, and whether the fund holds residual risk. (Pretty please? With a bow on top?🎀)

• realized versus unrealized returns, separately stated

• related-party and affiliate transactions, including any fees paid to the sponsor outside the management fee

• the GP’s own commitment, in dollars, and whether it was funded in cash or waived fees

• for credit funds: the share of income accrued as PIK rather than received in cash

What am I missing? Reply or leave a comment. I’ll compile the list and share it before the comment window closes.

-Leyla

P.S. I care. Do you?

The retailization train is coming regardless of what you and I think. I started this newsletter because I found the information asymmetry abhorrent. This is our shot to make the playing field flatter and less opaque. Let's use it.

P.P.S. New here?

Here’s where you’ll find the full archive (somewhat organized):

RREEF Property Trust, a $203 million daily-NAV REIT advised by an affiliate of DWS, just gave us an answer. On September 15, its board approved a plan to sell every property and wind the fund down, subject to a shareholder vote.

The company cited “heightened redemption activity” and “the challenges of attracting new capital.” New share sales, redemptions, and dividend reinvestment were all suspended immediately.

Subscribe now

The redemptions are only part of the story. For two years, almost no outside capital came in, while investors asked to take $223.8 million out. The fund sold properties, borrowed, and even took a $15 million investment from its own sponsor. All the while, it paid out more than its properties earned after interest and fees.

When an evergreen fund runs out of new money, it has three options: gate redemptions and wait (see Starwood REIT), list on an exchange, or liquidate. RREEF chose to liquidate.

👉 Here’s what happened at Starwood, which took option one:

👉 And here’s a real estate case study with a different twist:

Disclosure: This case study is provided for educational and informational purposes only and should not be construed as investment, legal, tax, or financial advice. The views expressed are solely those of the author. All examples are illustrative in nature and not guarantees of future outcomes. Readers should conduct their own independent research and consult with qualified professionals before making any investment or financial decisions.

Here’s what we’ll look at today:

1. Redemptions vs. new money: how much investors asked for, and how much they actually got back

2. Distributions: what the properties earned vs. what the fund paid out

3. The portfolio: what RREEF owns and what it paid for each property

4. Yield on cost: how to calculate it, and which properties are pulling their weight

5. The sales: what the fund sold, at what price, and the two properties to watch

Let’s jump in.

Invest in real estate? You’ll like this series - oldie but goodie:

The Money Stopped Coming In

For the last eight quarters, RREEF raised almost nothing from outside investors. Excluding dividend reinvestment and one sponsor purchase, it sold $12.7M of new shares in two years.

In Q1 2026, DWS itself bought $15M of shares. That was nearly all of that quarter’s new money, and it helped fund the exits.

Over the same period, investors asked to redeem $223.8M. Redemption requests peaked at $44.8M in Q2 2025. That quarter, investors received just 26% of what they asked for. By mid-2026 the redemption queue had mostly cleared, but new capital never came back. Seven months after the sponsor's purchase, the board voted to liquidate.

The headline yield vs. the rent

The press release highlights a 6.35% since-inception annualized return (Class I) and distribution rates of 5.1%–6.9%.

What the properties actually produced in H1 2026:

Read more

I called it “leverage on leverage” and walked through the mechanics: GP sets up an SPV → pledges the portfolio → borrows against it → and in some cases uses the proceeds to pay distributions to LPs who don’t know the money came from a loan.

Before we get in the weeds: this article is about fund-level NAV borrowing initiated by the GP. LPs can also borrow against their own fund stakes. That’s a different animal: it’s the LP’s decision, and it’s the LP’s risk.

Subscribe now

The issue with GP-initiated NAV facilities is that someone else is adding leverage to your equity position.

Twenty months later, the market is bigger, the SEC is paying attention, and (in an unexpected turn of events) LP pushback is actually working. Sort of.

Let’s catch up.

Disclosure: this is educational content, not investment advice. The author has no business relationship with any fund or lender mentioned here. Do your own due diligence (and read the darn footnotes).

Who’s Lending?

The largest single player remains 17Capital, the London-based specialist now majority-owned by Oaktree Capital. In March 2026, 17Capital closed its Credit Fund 2 at $7.5 billion (the largest NAV loan fundraise ever recorded).

That’s almost three times the size of its 2022 predecessor, which raised $2.9 billion. As of mid-2026, 17Capital has already deployed over $2 billion from Credit Fund 2 alone.

➡️ But the more interesting story is the bank vs. alternative lender segmentation. Banks now account for 60% of reported NAV lending volume, up from just 27% in 2024 (Rede Partners).

And banks are going after larger deals: the median bank deal size tripled in a single year, from €150 million to €500 million, while the alternative lender median held flat at €200 million.

The product has bifurcated: banks for the mega-cap books, specialist credit funds for everyone else.

Did LP Pushback Actually Work?

Read more

The struggle is real across the board: from one-off syndicated deals to publicly listed REITs and everything in between. And while funds and REITs that own actual properties have some flexibility (you know, sell properties, raise additional capital, refinance), real estate fund-of-funds have a different problem.

Subscribe now

They don’t control the assets underneath them.

We’ve seen this with Bluerock (BPRE):

Today, we're looking at another fund caught in the same pickle: Harrison Street Real Estate Fund (VCMIX).

But I’m less interested in enumerating all of VCMIX problems. The fund’s financial statements do a pretty good job of that themselves.

What’s more interesting is what happens next.

In May 2026, Nasdaq Fund Secondaries, LODAS Markets, and Harrison Street announced a collaboration to build a secondary-auction environment for interval-fund shares. Harrison Street said VCMIX would be the first manager to authorize transfers.

The development is interesting, and not just for VCMIX, but for the many other semi-liquid vehicles that are pro-rating redemptions. Remember, an interval fund has a mechanism for investors to get out (quarterly repurchase offers, at NAV), but it’s typically limited, and once the fund is pro-rating redemptions, there are few options for investors who want immediate liquidity.

And VCMIX’s repurchase history tells us there are plenty of investors who would like to leave:

The question is: at what price would someone else be willing to take their place?

That’s what we’re going to try to figure out.

Important disclosure. This case study is independent analysis published for educational purposes only and does not constitute investment advice, a recommendation, or a solicitation to buy or sell any security. All information is sourced from SEC filings (N-CSR and N-CSRS), the fund's prospectus, and publicly available company materials. Return estimates are hypothetical and for illustrative purposes only; they are not projections or guarantees. The author has no position in VCMIX and no business relationship with Harrison Street, Colliers, Nasdaq Fund Secondaries, or LODAS Markets.

The Fund

Harrison Street Real Estate Fund (VCMIX) is a $1.4 billion interval fund that invests in private and public real estate through a fund-of-funds structure:

• roughly 80% in institutional private real estate vehicles,

• 20% in public REITs and preferred stock.

It’s managed by Harrison Street Private Wealth, a division of Harrison Street Asset Management, which sits under the Colliers International umbrella. The fund charges 0.95% on NAV, plus sub-adviser fees, plus the underlying private fund fees (0.87%) that don’t show up in the headline expense ratio (1.82%). All-in, investors are paying roughly 2.7%.

The Shrink

Four years ago, VCMIX had $3.2 billion in net assets. Today it has $1.5 billion. The 54% decline is a two-part story: capital flight and shrinking valuation.

1. Capital Flight

Let’s start with capital flight: I’ve written before that once returns slow down (or turn negative), investors run for the exits. We’re seeing this play out with VCMIX.

Speaking of capital flight, here’s another CRE case study:

Every single quarter for the past three years, investor demand to sell has overwhelmed the offer. The fill rate has been declining, and by April 2026 it dropped to 17%. Nearly 20 million shares were tendered (about 30% of the fund's outstanding shares) and roughly $380 million worth went unfilled.

Inflows haven’t dried up entirely (color me shocked): the fund still attracted $147 million during the 12 months ended 3/31/26 (in addition to $9 million in reinvestments).

2. NAV Decline

Now let’s look at the valuation side of the story:

Read more

Today, I’m going to show you a close sibling of that practice, used in new-vintage PE evergreens, that can have a similar effect: massive early IRRs.

Subscribe now

Once you see it, you can’t unsee it. And once you understand the math behind it, you’ll see the trade.

Yes, I said “trade.”

No, I’m not advocating it. As always, none of this is investment advice. This is hypothetical, and today I’m not naming names.

For this hypothetical trade to work, two things need to happen. We’ll get to both later in the article. And because fund GPs are smart people, they’re doing what they can to minimize the number of wiseguys trying to execute the trade. I’ll show you what they’re doing, too.

And by the way: none of this is to say that evergreen PE funds aren’t worth owning. The point here is to understand what’s driving the returns you see in the first few quarters, and whether they reflect actual value creation or an accounting head start (spoiler: it’s often the latter).

The Setup

Imagine you are a large alternative asset manager. You run several flagship private equity funds (call them Fund I through Fund XIII) each one a traditional drawdown vehicle with a long, distinguished track record.

Now imagine you want to launch a new product: an evergreen fund, continuously offered, available to a broader investor base. Let’s say it’s a private equity fund. One that comes with a smooth upward-sloping NAV/share chart, direct investments in companies (rather than stakes in other funds), and supposedly great diversification. Wealth advisers and their HNW clients love those like cats love catnip.

You have a small problem: your new fund has no portfolio and no track record.

Enter the Seed Portfolio

Here’s how it works.

Read more

I have a gem of a chart for you today.

If you spend any time on X, you’ll know HVAC roll-ups are the butt of all PE-themed jokes. Laugh all you want. They’re still near the top of the PE deal leaderboard.

Subscribe now

In this issue:

1️⃣ Regulatory: the SEC proposed opening private markets to retail investors.
2️⃣ Private Equity: 63% of portfolio companies held 4+ years. Secondaries hit a record $121B in H1.
3️⃣ Private Credit: aggressive sponsors pay more to borrow.
4️⃣ Commercial Real Estate: CRE transaction volume is recovering, pricing is up 8.6% YoY, and one awesome chart.

Accredited Insight delivers the LP’s perspective on private credit, private equity, and CRE, drawing on hundreds of deals reviewed, and thousands of conversations. Paid subscribers gain access to our database of over 40 case studies and articles on everything from evergreen funds to due diligence (the kind of analysis that tells you what the GP pitch deck left out).

Subscribe now

1️⃣ Higher Thresholds, Wider Doors

The SEC wants to open private markets to retail investors (a familiar theme, no?)

A proposed rule titled “Enhancing Retail Exposure to Private Markets” landed at the White House Office of Management and Budget on September 1(RegInfo). It would do two things:

1. amend both the Investment Advisers Act and Investment Company Act to let retail investors access private markets through registered investment companies, and

2. expand the universe of clients who can be charged performance fees (this is the interesting part).

From the SEC: “Exposure to the full dynamism of our markets — both public and private — should not be reserved for wealthy insiders.”

And there’s an amusing bit of timing here.

Back in June, the SEC raised the qualified-client thresholds under Rule 205-3. This rule determines who can be charged performance fees by registered investment advisers.

The AUM test went from $1.1 million to $1.4 million; the net-worth test from $2.2 million to $2.7 million. Existing contracts are generally grandfathered, but new subscriptions and investments after June 29 have to meet the new numbers.

My first thought was: “Wait a second. Isn’t there a fund-of-funds workaround?”

(One can invest in a fund-of-funds without meeting the qualifications for the underlying funds, even when those underlying funds are only open to qualified clients).

Thankfully, the SEC appears to have thought of that, too. The new thresholds apply indirectly to certain fund-of-funds structures, including secondary funds. (White & Case)

Let’s play this one out a few steps. What happens when there is a flood of new money into any given asset class?

I’ll tell you what my crystal ball says 🔮 More semi-liquid (“evergreen”) vehicles. More ’40 Act structures chasing the same assets.

2️⃣ Private Equity: More Activity, Less Liquidity

274 days. (How’s that for a hook?)

That’s the median time it took to close a North American PE deal in Q2 2026, according to With Intelligence by S&P Global. (For those who struggle to divide by 30, that’s about 9 months)

Deals are getting done (despite slower fundraising, see chart below), but the exit bottleneck is not easing up.

More than 63% of active North American portfolio companies have now been held for more than four years. Median holding periods exceed five years across every major sector except industrials, at 4.9 years.

And when exits don't happen, someone eventually has to create liquidity. Enter secondaries.

H1 secondary transaction volume reached $121 billion, according to Evercore, the strongest first half on record. Full-year estimates are now $250–270 billion. Fundraising topped $50 billion in H1, led by Coller’s $17 billion and Partners Group’s $9 billion raises.

➡️ Let me connect some dots: remember how we speculated that opening the gates to allow more investors into private markets could create a surge in evergreen fund offerings?

What’s better to put into a fund than secondary stakes in stale primary funds?
(Buy them at a discount, mark them up, and generate beautiful, glorious unrealized gains..)

You really should read this, if you haven’t yet:

3️⃣ Sponsor Reputation Comes With a Yield Spread

How much more does it cost to borrow money when you’re that sponsor?

A new paper from Greg Nini at Drexel and Vincent Buccola at the University of Chicago Law School puts some data behind something credit investors have long suspected: sponsors develop reputations in the debt markets, and lenders price accordingly.

“The Sponsor Premium” uses leveraged-loan data from 2016–2025 to test whether a sponsor’s reputation for fair dealing (or the lack thereof) shows up in loan pricing. The paper comes with a naughty list of sponsors. (FT wrote about it here)

Whelp, turns out it does.

Across three different measures of sponsor reputation, portfolio companies backed by the most aggressive sponsors paid meaningfully higher yields than those backed by sponsors with more cooperative track records.

➡️ What the paper doesn’t ask is whether the more aggressive sponsors came out ahead (via more favorable outcomes in cases of default) to compensate for their higher borrowing cost.
If I had to bet, I’d say the aggressive sponsors weren’t doing it for nothin’.

4️⃣ Real Estate

Altus group published Q2 Transaction Analysis Report. Tons of good data:

• Q2 transaction dollar volume rose 11.3% QoQ, with trailing-four-quarter volume up 16.3% YoY.

• Pricing is recovering too: median transaction pricing rose 8.6% YoY to $131/SF, led by industrial (+13.2%). Multifamily and retail also posted solid gains, while hospitality was the only major sector with declining prices.

• Larger deals are taking an increasing share of transaction value, with investors favoring larger, higher-quality assets, durable income, and markets with strong fundamentals.

But the chart that caught me off-guard? This one:

I’m speechless.

Thanks for reading! As always, if you have any suggestions, reply to this email, leave a comment, or find me on socials (unhinged me on X, slightly more prim and proper me on LinkedIn). Have a great week!

-Leyla

Share

P.S. New here?

Here’s where you’ll find the full archive (somewhat organized):

The news created some waves, but this isn’t really a story about one lender getting one loan wrong. Blue Owl may have been acting entirely in good faith, and every individual mark may have been technically defensible under the applicable accounting framework.

That’s the problem.

Subscribe now

Whether Blue Owl got Loparex wrong is nearly beside the point (OBDC’s total position before the write-down was ~1.4% of total assets). The more interesting question is whether anyone outside Blue Owl could have known it was wrong.

And the most interesting question of them all is what that means for a $1.8 trillion private credit market.

There is no continuously observable price telling investors what a private loan is actually worth. Instead, there is an appraisal-like process: management estimates fair value, an independent valuation firm reviews the methodology, auditors test the process, and investors receive the number with a three-month delay.

A mark can be perfectly reasonable under accounting rules and still turn out to be spectacularly disconnected from economic reality.

Check out this Medallia chart:

And here’s the story on that loan:

The counter-argument, of course, is that public opinion can be spectacularly disconnected from economic reality too. But in the public markets, at least, investors have the option to trade on that information (or sell and cut their losses). That latter option is not available to investors in OBDC II, which also holds the Loparex loan.

Speaking of OBDC II:

The private credit ecosystem asks investors to treat internally generated estimates as if they were observable market prices. Unfortunately, we have very little way of knowing how wrong those estimates are (until something breaks).

What Happened

Back to issue at hand. Loparex makes very decidedly non-software things like engineered films and silicone release liners. That’s the backing you peel off a label or a bandage. (I too, had no idea what they were called).

In August 2019, Pamplona Capital Management acquired the company with roughly $600 million in debt, implying leverage of approximately 7.6x EBITDA. A bolt-on acquisition a few months later pushed leverage closer to 8x.

Enter our protagonist, Blue Owl, then Owl Rock Capital Corporation, who underwrote the second-lien loan at closing. This was a direct origination. Blue Owl held it exclusively across two BDCs: $112 million of the second lien on OBDC’s books; and $28 million on the books of its smaller sibling, OBDC II (4:1 investment ratio, that split will remain constant throughout the years).

A second tranche was added for the bolt-on. Combined par across both vehicles: approximately $172 million. For the next four years, the positions were carried near par.

The Restructuring

I’ll spare you the full saga of what Covid —> supply chain issues —> cost increases did to the financials of the borrower. Suffice it to say the company, saddled with substantial leverage, has not done well.

By January 2024, every party in the capital structure had retained restructuring counsel. In March’24, Loparex executed one of the year’s most aggressive liability management exercise (LME). This was an uptier transaction (when some creditors get priority over others).

From Octus: “Loparex completed a liability management exercise in 2024. In the first stage of the Loparex transaction, 75% of the first lien lenders provided $135 million of new money in the form of tranche A of a new superpriority term loan facility due February 2027.

They also swapped their $334 million of the $354 million U.S. dollar S+450 bps first lien loan due July 2026 and $76 million of the $194 million euro E+525 bps first lien loan due July 2026 for first-out with a total principal of $175 million, $280 million of S+450 bps second-out term loan and $120 million of S+450 bps third-out term loan.”

In plain English: the deal created a five-layer waterfall in repayment priority. The lenders who put up new money jumped to the front of the line. Everyone else got pushed down.

S&P downgraded the company to selective default, deeming the exchange tantamount to default because non-participating lenders received materially less than originally promised.

On April 12, 2024, S&P upgraded Loparex from SD to CCC+ with a negative outlook. The language in that report is not pretty, let’s put it that way. S&P explicitly stated that the capital structure remained “unsustainable” and assigned recovery ratings across the new tranches.

S&P’s recovery rating scale runs from ‘1’ (highest expected recovery, 90–100%) to ‘6’ (lowest, 0–10%). At origination in 2019, S&P assigned the second lien a ‘6.’ After the 2024 LME, S&P affirmed it.

Here’s the full post-restructuring picture:

• Tranche A (superpriority new money): ‘B’ rating, ‘1’ recovery (90–100%)

• First-out and second-out: ‘CCC+’ rating, ‘3’ recovery (50–70%)

• Third-out and existing first lien (non-participating): ‘CCC-’ rating, ‘6’ recovery (0–10%)

• Second lien: ‘CCC-’ rating, ‘6’ recovery (0–10%)

❓Pop quiz: which tranche did Blue Owl hold?
Yes, it was the very bottom one. $166 million of their $172 million total exposure (96%) sat in the second lien.

Monroe Capital

If you are doing mental math, you know we’re missing some $400 million dollars in loans (the company was purchased with ~$550 million in debt, Blue Owl’s exposure was around $172 million), so I tried looking for the “missing” money.

Monroe Capital Income Plus Corporation, a small BDC managed by Monroe Capital, held a tiny piece of Loparex’s original first-lien term loan (about $754,000 in par, acquired in April 2023.)

Pocket change compared with Blue Owl’s combined $166 million second lien.

But Monroe’s marks tell a story that Blue Owl’s don’t:

Read more

On August 30, Apollo announced a $9 billion investment in ONEOK, the Tulsa-based midstream energy company. The headlines called it an “equity investment.” Technically, it is.

But upon closer inspection, it’s neither fish nor fowl. It’s equity with debt-like features, but not preferred equity. It has debt-like features but not the lender’s priority in repayment. It’s a financial instrument engineered so that both sides get to call it what they need it to be.

ONEOK gets $9 billion of capital that counts as equity on its balance sheet, doesn’t dilute existing shareholders, and helps reduce its leverage.

Apollo gets a contractual cash-flow stream that can potentially be packaged, tranched and sold as investment-grade securities to insurance companies and other institutional investors.

In other words: equity for the borrower, debt-like cash flows for the investor, and potentially investment-grade bonds for the eventual buyer.

Today we’ll walk through two things:

1. how this deal was structured on the corporate side,

2. and how equity gets converted into rated bonds.

If you missed the CLO article, you’ll want to read that first. This transaction isn’t technically a CLO, but the mechanism (taking contractual cash flows and turning them into rated securities) is nearly the same, with one important difference: a CLO diversifies across hundreds of borrowers, while this structure depends entirely on a single company’s cash flows.

This is what shows up on insurance balance sheets:

This is a case study examining the structure of a publicly announced transaction. It is based on SEC filings, public press releases, and reporting by Bloomberg. The author holds no position in ONEOK, Apollo, or any securities discussed. Nothing in this article is investment, legal, or financial advice.

What ONEOK is doing with the money

ONEOK is using the $9 billion for two things.

First, it’s acquiring Brazos Midstream’s natural gas gathering and processing assets in the Permian Basin for $4.4 billion. This nearly doubles ONEOK’s processing capacity in the region.

Second, it’s using roughly $5 billion to extinguish existing senior debt through tender offers, make-whole calls and term-loan repayment.

Net result: ONEOK gets bigger while reducing leverage. Pro forma 2027 leverage is expected to fall to roughly 3.25x debt-to-EBITDA, from around 4x.

And it does this without issuing common stock.

‼️ If ONEOK had issued $9 billion of new bonds, leverage would have gone the wrong direction. If it had issued $9 billion of common equity, existing shareholders would have been significantly diluted (current market cap is roughly $60 billion). Apollo’s structured equity lets ONEOK thread the needle.

The structure

This is the dense part (it’s your cue to get your coffee).

Read more

Well, CLOs are like that for me lately (no, I’m not shopping one, thanks for asking).

Subscribe now

Last fall, when First Brands filed for bankruptcy, I wrote about CLO mechanics (the fun stuff: the tranching, the waterfall, the difference between BSL and middle-market collateral).

If you missed that piece, start here:

Today’s piece is the companion. We’ll tackle a different question: why they work (and where they don’t).

‼️ CLOs, CDOs (collateralized debt obligations), CFOs (collateralized fund obligations) and a slew of other structured financial products are all the same species. No, they are not the same (think chihuahua vs. German shepherd: you won’t mistake one for another, but they are both dogs). Once you understand the mechanics, you'll be able to grasp newfangled wonders of finance, like the recent Apollo equity structuring (stay tuned, more on this in a separate post). And maybe you’ll start seeing them everywhere too…

📚 In the interest of keeping you awake, this article is fairly high-level. If you suffer from insomnia (or are dying to learn more about OC ratios for mezzanine tranches) I’ve included further reading at the end. Let’s get into it..

A Very Quick CLO Refresher

A typical broadly syndicated loan CLO issues six to seven rated debt tranches plus unrated equity. The AAA tranche (the largest slice, up to ~64% of the deal) gets paid first and benefits from substantial subordination beneath it.

That subordination means more than a third of the underlying loan portfolio would need to default with zero recovery before the AAA tranche loses a dollar.

Now look at the bottom of the stack (the dark blue equity slice in the chart above). The equity tranche has no guaranteed return. It collects whatever cash is left after every debt tranche above it has been paid (and after management fees, trustee expenses, etc).

In good years, that residual can deliver 15–25% cash-on-cash distributions. In bad years, overcollateralization tests fail, the waterfall redirects cash upward to senior debt tranches, and equity distributions stop entirely.

The median unlevered IRR for CLO equity vintages from 2003 through 2023 is approximately 11%. But median numbers only tell part of the story: deals priced at wide loan spreads (2008–2009, 2020) significantly outperformed, while tight-spread vintages (2014, 2021) delivered lower returns.

We’ve published case studies on several funds that hold CLO debt and equity on the books:

One important caveat: everything above describes broadly syndicated loan (BSL) CLOs. Private credit CLOs (sometimes called middle-market CLOs) are a slightly different animal. The underlying loans are to smaller, often unrated borrowers, with less liquidity, less transparency, and historically higher default rates.

To compensate, private credit CLOs are built with thicker cushions at the bottom: the AAA tranche is a smaller share of the deal, and the equity and junior mezzanine slices are proportionally larger. Spreads are roughly 200 bps wider than BSL equivalents. More on that here.

What Protects Senior Tranches

So how do you take a pool of below-investment-grade loans and turn them into AAA-rated securities?

Read more

Every other week, we send a quick digest on what’s catching our eye in private markets.

Before we get into it, here’s the chart of the day:

Non-listed BDC fundraising fell to $2 billion in Q2 2026 (an 82% drop from a year ago and the lowest quarterly total since 2020). Redemptions hit 12.4% of NAV, the highest level Stanger has ever recorded. Sponsors fulfilled 38% of requests and returned $5.9 billion (against a record $15.4 billion in requests), but net outflows still came in at roughly $3.8 billion for the quarter and $5.6 billion for H1. Aggregate NAV declined 3.1% quarter-over-quarter to approximately $122.4 billion.

Subscribe now

In this issue:

1️⃣ Private Credit: BDC fundraising hits a five-year low. But trapped investors are not willing to take a haircut to get out (yet?)

2️⃣ Commercial Real Estate: Apollo injects $1 billion into SREIT’s affordable housing portfolio. We’ll take a closer look at the deal terms below.

3️⃣ Private Equity: who’s actually investing in continuation funds? Plus, Vanguard makes a $4 billion bet on advisor infrastructure.

Accredited Insight delivers the LP’s perspective on private credit, private equity, and CRE, drawing on hundreds of deals reviewed, and thousands of conversations. Paid subscribers gain access to our database of over 40 case studies and articles on everything from evergreen funds to due diligence (the kind of analysis that tells you what the GP pitch deck left out).

Subscribe now

1️⃣ Private Credit

1. 26% discount was a no-go for trapped investors

Cox Capital Partners offered to purchase up to $90 million of shares in five non-traded BDCs managed by HPS, Apollo, Ares, and Blue Owl at an average 26% discount to NAV (I wrote about the Blue Owl fund here).

The result? Less than $5 million in orders. Apparently, investors want liquidity, they just don’t want to pay 26% for it. (Yet??)

Cox is now extending its tender offers to interval funds managed by Cliffwater and Variant Investments. Variant, for context, saw requests to redeem roughly 50% of its shares last quarter.

Meanwhile, on the publicly traded BDC side, the discount between NAV and share prices is starting to narrow. The S&P BDC index is up 6% QTD.

But the underlying pressure hasn’t exactly disappeared: defaults are up (you’ll find the chart here), fundraising continues to crater, while the largest public BDCs have seen their portfolios contract for three consecutive quarters, as repayments continue to outpace new originations.

2. BlackRock is shopping what’s left of TCP Capital

BlackRock is sounding out buyers for the $671 million loan portfolio remaining in TCP Capital Corp.

KBW has reportedly begun pitching the portfolio to firms including Ares, although talks are still early, Bloomberg reports. TCPC is also considering other options, including reinvesting, returning capital to shareholders, or merging with another entity.

This follows TCPC’s sale of $523 million of loans to a Pantheon-backed vehicle earlier this month. That was an interesting transaction, read the full story here:

2️⃣ Commercial Real Estate

1. SREIT’s $1 billion structured lifeline

Apollo is investing $1.02 billion in a joint venture with Starwood Real Estate Income Trust, the $22.5 billion non-traded REIT that suspended redemptions earlier this year.

The basic idea is simple:

• Apollo gets 41.5% of a portfolio of ~120 affordable-housing properties,

• SREIT keeps 58.5% and uses the proceeds to pay down debt.

But, as always, the interesting stuff is in the fine print (did I get you in the habit of opening SEC filings yet?)

Apollo’s investment is rated investment-grade and carries a guaranteed minimum yield that increases over time. SREIT owes those payments regardless of portfolio performance.

SREIT can buy Apollo out between years 5 and 12. If it does so in years 5–10, Apollo’s IRR is capped at 7%. The economics get increasingly expensive with a longer wait (higher minimum payments, make-whole contributions, and contingent obligations). If SREIT can’t meet those obligations, Apollo gets governance rights over the portfolio.

So the trade is pretty straightforward:

SREIT gets liquidity and time. Apollo gets a protected return.
(What do existing investors get? You tell me in comments please)

Economically, Apollo’s JV investment functions as preferred equity:

And here’s an example of a recapitalization in a one-off multifamily deal:

3️⃣ Private Equity

1. Who’s actually in continuation funds?

As continuation funds become a bigger story in PE, Preqin's latest data, as of April 2026, gives us a pretty interesting look at who's actually putting money into these vehicles:

Public pension funds (19.1%), foundations and endowments (18.8%), and private-sector pension funds (18.1%) collectively account for 56% of CV investors. Fund-of-funds managers represent another 16.6%.

You see private wealth? At just 2.5%? Of course, there's an asterisk: many fund-of-funds managers are themselves targeting wealth investors, so retail exposure is higher than this chart suggests.

👉 Here’s a fund with CVs on the books:

👉 And here’s your primer on CVs:

2. Vanguard buys Altruist for $4 billion

Last, but most certainly not least: Vanguard is buying Altruist for roughly $4 billion (WSJ). The price is more than 2x Altruist’s $1.9 billion valuation from April 2025. (Umm. Maybe not the value acquisition.)

Altruist is an AI-forward wealth technology and custody platform for independent financial advisors. It provides the plumbing behind an advisor’s practice: self-clearing brokerage, account opening, trading, portfolio management, billing, reporting, and all the other glamorous infrastructure that nobody thinks about until it breaks.

It competes directly with Schwab and Fidelity in the RIA custody space.

CEO Salim Ramji is pushing Vanguard beyond index funds and deeper into wealth-advice infrastructure.

Vanguard already has partnerships with Blackstone and Wellington to bring private-market products to the wealth channel.

This is my personal opinion, take it for what it’s worth, but I just can’t fathom how this product will set the world on fire:

But products are only half the equation. You also need the pipes (now you own the product and the distribution channel). Altruist gives Vanguard those pipes, and, importantly, lays them directly underneath the independent advisors.

Altruist will operate as a standalone business, retaining its leadership and brand.

Thanks for reading! As always, if you have any suggestions, reply to this email, leave a comment, or find me on socials (unhinged me on X, slightly more prim and proper me on LinkedIn). Have a great week!

-Leyla

Share

P.S. New here?

Here’s where you’ll find the full archive (somewhat organized):

In today’s guest post , who writes investigative research on insurance, structured credit, and private equity, provides a roadmap to the private asset categories that end up on insurance balance sheets, why they’re there, and what the labels actually mean. If you’ve followed the private credit and insurance conversation and found the terminology confusing, this is your field guide.

-Leyla

Subscribe to Rod's Substack

Overview

Out of the ashes of the Savings and Loan crisis a book was published with the title – The Best Way to Rob a Bank is to Own One, by the great chronicler of the S&L crisis, William Black. I think the same can be said about today’s insurance industry.

I have published many articles focusing on insurance and the Private Equity (PE)/ Private Credit (PC) ecosystem. Those articles focused on specific insurers and/or asset types.

The goal of this article is to demystify the many ways insurance companies are exposed to the PE/PC ecosystem and to describe the growing influence of private asset managers over the insurance industry.

There are many risky asset types that insurance companies invest in. This is not new. What is new is the more aggressive push to invest in illiquid private assets originated by an affiliated PE/PC sponsor. The investment categories are many and often in the public discourse are incorrectly defined and / or conflated.

This article provides an overview of 2 trends:

1) PE ownership of insurance companies, and

2) Insurance companies increasingly investing in Private Assets.

These trends are inextricably linked. The primary goal of PE/PC and insurance tie ups are the ability to use insurance company balance sheets to absorb PE/PC originated assets, increase Assets Under Management (AUM) and perpetuate the fee generation machine.

I first focus on the broader trends and then dive into specific asset categories and risks that PE linked insurance companies are exposed to. This article doesn’t cover all risky asset classes, but rather those most connected to PE/PC sponsors. Asset classes like Commercial Mortgage-Backed Securities (CMBS) and Commercial Real Estate Mortgages, merit attention but are beyond the scope of this article.

As an example of the latter, Apollo dissolved their Commercial Mortgage REIT and “sold” $9B of its commercial mortgages to Athene at a price near 100, when the REIT was trading at a deep discount to NAV. To me, this raised questions about whether Apollo was using its insurance balance sheet to support an affiliated vehicle. This illustrates the fact that PE/PC sponsors can influence the assets of their captive insurer far beyond the private assets label.

How and why Private Asset sponsors control insurance companies

Before jumping in, it’s important to understand the flow from the PE/PC sponsor to the insurance companies. Apollo was among the first to invest in, control and scale an insurance company (Athene).

The first goal is for the sponsor to control the investment decisions at the insurance company. This can be achieved in several ways. Most straight forward is for the PE/PC sponsor to own the insurance company. Once they own the insurance company, the control is complete. Once owned, the PE/PC sponsor can appoint themselves as the Investment Manager (IM) for the insurance company. This role is generally formalized under an Investment Manager Agreement or IMA.

A PE/PC sponsor doesn’t need to have complete ownership to benefit from the above relationship. Ownership levels run the gamut from 0 to 100%. The key is to be appointed Investment Manager. Ownership gives more control, but even absent ownership, the PE/PC role as IM can achieve some of the same goals and presents similar risks.

See below for the basic flow of the relationship.

Private asset sponsors want control of insurance companies for 5 reasons: 1) the holy grail of permanent capital, 2) efficient leverage, 3) ability to commit a large balance sheet for deals that wouldn’t be possible with a traditional fund structure, 4) ability to grow Assets Under Management and therefore fees, 5) weak regulatory oversight.

Trends in Private Assets and insurance

As noted above, two trends are occurring in parallel. Growing control of insurance balance sheets by private equity or private credit sponsors and insurance companies increasingly investing in private assets.

The chart below (source: McKinsey) illustrates the explosive growth in private capital-controlled insurance companies. While McKinsey doesn’t identify the specific insurers, it’s likely that Apollo, KKR and Brookfield alone comprise 50% of the assets. Notably, I don’t believe McKinsey includes MassMutual since they own the asset manager, rather than Vice Versa (Barings) and it’s likely they wouldn’t consider Barings as a private asset manager. Adding MassMutual alone would bring the total close to $2T.

Turning to the specific assets, the chart below shows that the increase in private placements held by insurance companies closely tracks the growth shown in the above chart. It’s important to note that private placements are not the same as Private Letter Ratings (PLR) - a topic that has gained a lot of attention. While private placements often have private letter ratings, they can also have public ratings. Therefore, the categories are related, but not the same.

A late 2025 Fitch report revealed that PE affiliate insurers have 6X the exposure to private assets (as a percentage of asset) relative to non-PE affiliate insurers.^[1]

Read more

Private credit is a catch-all phrase that includes virtually all non-bank lending. Real estate private credit is one of the buckets, with its own unique wrinkles:

The vast majority of such funds file a Regulation D form, which exempts them from filing financials with the SEC and makes them totally opaque.

For a while, I wanted to do a case study on one. I just didn’t have a good candidate. Well, folks, I finally found one.

Bonus: the fund is new (less than a year old), which makes the financial statements somewhat easy to decipher (ok, ok… they are never easy, this set happens to be a little less convoluted than others).

Here’s another real estate debt case study, also full of interesting things:

Disclosure: This case study is provided for educational and informational purposes only and should not be construed as investment, legal, tax, or financial advice. The views expressed are solely those of the author. All examples are illustrative in nature and not guarantees of future outcomes. Readers should conduct their own independent research and consult with qualified professionals before making any investment or financial decisions.

What’s in the fund

As of December 31, 2025, the fund had $50.3 million in net assets across 70 positions. So, what can we learn from the financial statements? Quite a bit.

1. Start with the balance sheet

“Why do we start with the balance sheet, Leyla?” Because it’s the first statement in the annual report.
(In this case, you only have to flip through 16 pages to get there.)

And because reading financial statements is hard enough as is, so we’ll choose the path of least resistance.

A few things jump out:

1. Cost and value of investments

The fund held $52,408,408 of investments at cost and $52,449,483 at fair value (are your eyes bleeding yet?) The gap is only $41,075 in net unrealized appreciation (it’s from the RMBS sleeve, not the loans).

The loans are short-term, with a weighted-average term of 13.7 months, and are carried at par. Which brings us to a very important point for investors:

Read more

(No idea why I decided to pursue a degree in English my first go-around in college, but that’s beside the point. I came to my senses later in life, when faced with the prospect of living in abject poverty surrounded by books.)

Enough about college majors, back to math problems.

(The stock rallied 11% on the news, but that’s a separate story.)

Subscribe now

👉 I kid you not, today’s math problem was solved by reading the footnotes to appendixes. So read the dang financial statements (and footnotes, and appendixes, and exhibit 99.1, while you are at it)

To solve today’s math problem, we first need to identify what exactly was sold. And no, “loans” is not the right answer.

Step 1: CLO

On May 27, 2026, TCP took $535.8 million of loans on its books (TCP’s assets, not the borrowings of the fund) and packaged them into a securitization (a CLO):

• TCP sold the senior, investment-grade-style slices of the pool to outside CLO debt investors, raising $405.9 million.

• It kept Class D tranche, and the bottom slice (the “equity” or first-loss piece) for itself.

• That retained slice was worth roughly $130 million at par, against a $536 million pool.

For those not yet sufficiently caffeinated to do the math, that’s about 3x debt-to-equity leverage on that specific slice of the portfolio, layered inside a fund that was already carrying its own leverage (1.38x net debt-to-equity) at the company level.

None of this is unusual, I’m simply showing you the plumbing inside. TCP owned and consolidated the entire structure on its own balance sheet for about ten weeks before announcing the Pantheon transaction.

The first case study on this fund:

Step 2: What got sold to Pantheon and at what price

By June 30, 2026, the loan pool inside this structure (plus some additional contributed positions) was marked at $523 million. We know there is $406 million of CLO debt sitting on top of it. Add the roughly $74 million TCP separately posted to cash-collateralize the vehicle's unfunded commitments, and TCP's total equity claim comes to:

Read more