Greystone 2025: The Funding Stack Improved, but the Real Test Moved to Asset Quality and Parent Cash

Getting To Know The Company

At first glance Greystone looks like a straightforward debt story: 132 bridge loans backed by multifamily and seniors housing properties, a $3.9 billion portfolio, an Aa3.il rating, a 1.63 interest coverage ratio, and an additional note raise in February 2026. That is too simple. In practice, this is a bond-only public issuer sitting above a multi-layered credit platform, where most of the assets are financed through short bank lines and CLO structures, while the Israeli public investor owns unsecured debt at the parent-company level.

What is working today is clear enough. The liability structure became more durable: funding agreements fell from $1.43 billion to $0.99 billion, CLO liabilities rose from $1.17 billion to $1.61 billion, and in February 2026 the company also expanded its Israeli note series with roughly $188 million of net proceeds. The book itself still reflects relatively conservative underwriting, with a weighted average stabilized LTV of 66.4%, an average loan term of 1.8 years, and 50% of the portfolio carrying no CapEx reserve at origination.

The problem is that the better liability structure did not translate into better economics in 2025. Total investment income fell from $409.4 million to $344.4 million, net investment income fell to $112.1 million, and real estate and collateral related expenses jumped to $22.9 million from $5.8 million a year earlier. At the parent-company layer the picture is even sharper: only $1.46 million of cash remained at year end, operating cash flow was negative, and the company paid $53.3 million of distributions to its shareholder even though the year-end trust deed disclosure said there was no remaining distributable balance as of December 31, 2025.

That is why 2026 now reads less like a growth year and more like a proof year for bondholders. The market no longer needs proof that Greystone can raise debt. It needs proof that the loan book can return to cleaner earnings, that collateral-protection costs normalize, and that what looks like a wide capital cushion actually translates into real flexibility at the issuer level. It also helps to remember the actionability constraint here: this is a bond-only listed vehicle, with no listed common equity, so the public read runs through covenant quality, debt-layer protection, and cash access, not through equity upside.

Greystone's Economic Map

These two charts frame the story early. On one hand, this is a focused book rather than a broad commercial real estate credit machine. On the other, the relatively conservative underwriting and the low CapEx burden explain why the portfolio still reports a reasonable LTV profile even in a weaker earnings year.

Events And Triggers

Trigger one: 2025 was a sharp transition year away from short bank funding toward CLO structures. In May 2025 the company issued FL4-CLO at roughly $900 million, and in October 2025 it issued HC4-CLO at roughly $450 million. In the fourth quarter it also refinanced its largest healthcare funding line, about $300 million, and lowered the spread to SOFR plus 2.25%, around 75 basis points below the prior level. This matters because the company replaced part of its short and more flexible funding with longer and more structured financing.

Trigger two: the loan book did not grow in 2025, it turned over. New originations were about $1.45 billion, while repayments were about $1.5 billion. So the year did not produce numeric growth in the asset base. It produced a change in funding mix and liquidity quality.

Trigger three: after the balance-sheet date, on February 10, 2026, the company expanded Series A by NIS 583.54 million principal amount, about $186.14 million, and received net proceeds of about $187.97 million. At the same time it executed an additional forward contract designated as a fair-value hedge of the NIS-denominated principal. This is more than just another debt raise. It is a direct oxygen injection into the parent-company layer, exactly where 2025 had ended with very little free cash.

Trigger four: three days before publishing the annual report, on March 19, 2026, the board approved a $10.59 million dividend to the sole shareholder. The shareholder told the company it intended to use the distribution to increase the subordinated shareholder loan. That looks like a small detail, but it sits at the center of the story. It shows that even after a weaker profitability year, management still treats the equity and debt layers as one connected structure rather than as a fully separate public-credit silo.

Trigger five: until January 10, 2027 the fund retains a right of first look on qualifying investment opportunities. After that date the broader Greystone group may change the allocation policy after consultation and six months' notice. Management says it does not expect a change at this stage, but the very existence of that date means the market should read 2026 as a year that also tests sourcing continuity, not just earnings.

This chart clarifies what the balance-sheet headline hides. The real question is not whether total debt fell much. It barely did. The real question is who sits closest to the collateral, who benefits from security, and who depends on the residual layer.

Efficiency, Profitability And Competition

The core 2025 message is better funding structure alongside weaker profitability. That is not an accounting paradox. It is the result of two forces working together: lower yield on the book and a much higher cost of protecting troubled collateral.

Total investment income fell 15.9% to $344.4 million. Most of the damage came from interest income on investments, which declined from $374.0 million to $306.1 million. Management's explanation is clear: the weighted average variable rate collected on the portfolio fell by 96 basis points, and the average portfolio balance also declined year over year. Even though interest expense fell to $192.2 million, net investment income still dropped 27.0% to $112.1 million, and net increase in net assets resulting from operations fell to only $74.2 million.

That chart matters because it shows 2025 was not an economic breakout year. It was a liability-management year. Fee and other income did rise to $31.0 million because higher repayments generated more exit fees, but that is not the kind of income stream investors should treat as a durable base.

What really weighed on earnings

The most unusual line in the report is not necessarily the net loss on investments. It is the jump in real estate and collateral related expenses. That line rose to $22.9 million from $5.8 million in 2024. In the accounting note the company effectively breaks that number into two distinct mechanisms: $12.51 million of operating shortfalls and capital improvements funded to preserve the value of troubled properties, plus another $10.41 million of protective advances for taxes, insurance, and essential maintenance that were expensed once recovery from the borrower was no longer viewed as reasonably collectible.

That matters because it means the 2025 damage did not come only from spread compression. Part of it came from having to spend real money to keep collateral alive. That is no longer just a credit-market question. It is a book-quality question.

Credit quality: better, but not clean

This is where first-read impressions can mislead. The broad watchlist actually improved. Unpaid principal on the watchlist fell from $701 million to $484 million, and the share of the book on watchlist fell from 17.5% to 12.4%. But loans more than 60 days past due moved the other way: the balance rose from $160 million to $176 million, and the share of the portfolio rose from 4.0% to 4.5%.

That is not a contradiction. It is a risk hierarchy. Fewer loans may require broad monitoring even while a somewhat larger subset pushes deeper into trouble. The delinquent-loan list also shows that risk is not concentrated in one single position that could overwhelm the company: the largest 60-plus-day exposure is $44.8 million, or 1.14% of the book, and two of the delinquent loans, totaling $25.56 million, were already repaid after the balance-sheet date. That reduces tail-risk concentration, but it does not remove the cost signal.

Where the moat is, and where the pressure remains

Greystone does have a real operating moat. Agency licenses are limited, and the report describes the broader Greystone platform as one of only four firms that hold the relevant Fannie Mae, Freddie Mac, and FHA/HUD approvals across programs. That is a meaningful advantage in sourcing borrowers, underwriting deals, and offering not just bridge financing but also a realistic route into permanent financing.

But that same moat creates dependency. The model works only if the borrower can actually refinance out once the asset has been stabilized or repositioned. When that happens, Greystone earns exit fees and recycles the book. When it does not, the company ends up paying taxes, insurance, and repairs to protect collateral. So 2025 reads like a year in which the licensing moat remained intact, but its conversion into clean profitability weakened.

Cash Flow, Debt And Capital Structure

The right cash lens here is all-in cash flexibility. Not how much the business produces before strategic uses, but how much flexibility is left after counting collateral reserves, debt service, distributions, and transfers between layers. That is the relevant framework for a bond-only issuer.

On a consolidated basis, liquidity improved, but much of it is trapped

At the consolidated level operating cash flow was positive at $208.1 million. That is a good number. But it did not translate into a much larger free cash pile. Ordinary cash fell from $112.0 million to $72.8 million, while restricted cash rose from $118.1 million to $261.8 million, mainly after the HC4-CLO issuance. In other words, cash plus restricted cash rose to $334.6 million, but much of that improvement sat in ring-fenced reserves rather than in free corporate cash.

The $460 million liquidity figure used in the investor presentation also needs to be read correctly: it includes $73 million of cash, $262 million of restricted cash, and $125 million of unused lines. That is important, but it is not all equivalent to cash the parent can freely distribute or even freely redeploy.

At the parent-company level, this is where the story lives

At the standalone parent-company level, the picture is much tighter. The separate financial statements show year-end cash of only $1.462 million, down from $70.297 million at the start of the year. Operating cash flow was negative $2.362 million, and financing cash flow was negative $66.446 million.

The reason is not weak operating performance at the parent alone. It is the transfer structure. During 2025 the company received only $9.17 million of net distributions from subsidiaries, but paid $53.255 million of distributions to the shareholder and repaid another $13.191 million of the shareholder loan. That is exactly where the consolidated read can mislead.

Covenants are wide, but they need to be read correctly

The company ended 2025 in compliance with all covenants. Interest coverage was 1.63 versus a 1.25 minimum. The equity-to-assets ratio was 33% versus 20% or 22.2%, depending on the clause. Covenant equity stood at $1.419 billion.

But one of the report's most important subtleties sits right here: the trust deed's definition of equity includes perpetual shareholder loans. So the covenant cushion is real, but it is not built only on ordinary accounting net assets. Consolidated net assets were $870.8 million, while the shareholder loan stood at $548.3 million. That is not a mistake. It is simply the capital architecture, and investors need to read it that way.

What improved in debt, and what did not

The good news is that the company reduced its reliance on warehouse-style bank funding, increased the CLO share of the stack, and kept net debt to assets at 58.6% versus 61.2% in 2024 and 67.0% in 2023. Interest coverage also remained above 1.6 over the last three years.

But the other side also needs to be stated plainly: Series A is unsecured. It is structurally behind the secured debt sitting close to the assets. The presentation shows $637 million of unencumbered assets, and that does provide some cushion. Still, Israeli noteholders do not rely on the same collateral layer as the banks and the CLO structures.

Another important detail is the CLO cash-flow architecture itself. As long as each structure remains in compliance, the company can receive residual cash. If one stops complying, cash in that CLO goes first to pay principal and interest on the CLO notes and the residual stream to the company is halted until compliance is restored. All CLOs were compliant at year-end 2025, but this is a reminder that not every dollar earned in the book is a dollar that automatically moves upstairs.

Outlook

The four non-obvious findings that carry from this cycle into 2026 are these:

1. 2026 is not a proof year for capital-markets access. The company already proved that. It is a proof year for earnings quality and collateral quality.

2. The February 2026 note expansion solved an immediate liquidity problem at the parent, but it did not erase the fact that the parent layer had been run almost down to zero by the end of 2025.

3. A lower broad watchlist is not enough if collateral-protection costs remain high and hard delinquency keeps moving up.

4. The book still benefits from a supportive sourcing framework through January 10, 2027 because of the right-of-first-look arrangement, but that window is not permanent.

What has to happen inside the book

The first requirement is normalization in the cost of dealing with troubled loans. If real estate and collateral related expenses stay around $20 million or more, new CLO issuance will not be enough to return the story to a cleaner place. Over the next 2 to 4 quarters the market will want to see lower collateral-protection expense, stable or lower over-60-day delinquency, and continued resolution work without simply rolling problems forward.

What has to happen in funding

The company already did most of the heavy work in the liability stack. It now has to show that better financing is actually flowing through to earnings rather than only to the appearance of balance-sheet strength. If bank-line spreads do decline, and if the mix between warehouse funding, CLOs, and unsecured notes remains balanced, 2026 can look like a stabilization year. If not, the market will begin to ask whether the company simply exchanged one problem for another.

What has to happen at the parent layer

This is where the test is sharper. The note series was expanded and the company already tied that move to a new FX hedge. But what will determine the read over the next several months is not just the size of the issuance. It is what the company does with that cash. If the parent layer is again used for aggressive distributions or shareholder-loan rearrangement without meaningfully better upstream cash generation from below, noteholder confidence will be tested.

What has to happen in seniors housing

There is also a real execution test here. Beginning April 1, 2026 Greystone takes healthcare asset management back in-house after operating under a contractual relationship with Monticello. That may improve control and coordination, but it also adds operational burden at the same time HC4-CLO and the broader healthcare book need to prove they are more than just a clever funding move.

Risks

The first risk is borrower exit risk. The entire model depends on borrowers refinancing out once assets are stabilized or repositioned. If the U.S. permanent-financing environment tightens, if asset performance falls short of agency standards, or if the borrower fails to execute its business plan, Greystone remains stuck with a loan that was supposed to be a bridge rather than an end state.

The second risk is credit-quality deterioration beneath the headline of watchlist improvement. The broad watchlist fell, but hard delinquency rose, and the collateral-protection expense line already shows that this is not just a yellow-flag monitoring story. It is also a real cash-use story.

The third risk is structural subordination for Israeli noteholders. Series A is unsecured, effectively junior to secured debt at the collateral level, and dependent on a mix of unencumbered assets, upstream cash from subsidiaries, and the continued presence of a subordinated shareholder loan that absorbs losses first. That structure can work well, but it needs to be read as its own credit layer rather than as a simple consolidated balance sheet.

The fourth risk is CLO test risk and restricted-liquidity risk. All structures were in compliance at year-end 2025, but if one falls out of compliance the residual cash stream from that structure stops until it cures. In the same way, restricted cash and unused lines are not the same thing as free corporate cash.

The fifth risk is licensing and regulatory dependence. Parts of the platform rely on agency servicing and origination approvals and on Greystone Servicing Company as the entity behind almost all bridge and mezzanine loan origination. Losing one of those approvals or materially impairing that servicing platform would directly damage the model.

The sixth risk is concentration that is not extreme, but still real. Texas accounts for 16% of the portfolio and New York for 15%. Seniors housing accounts for 24.1% of the book. This is not a one-state or one-asset blowup profile, but it does leave meaningful concentration points.

Conclusion

Greystone enters 2026 with a better liability structure and a renewed ability to access the Israeli debt market, but 2025 did not prove that the business itself became stronger. Profitability weakened, the cost of protecting collateral surged, and the issuer's own cash position was run down almost to zero. So the short-to-medium-term market reading will be determined less by the size of the balance sheet and more by credit quality, upstream cash flow, and distribution discipline.

Current thesis: Greystone improved the funding stack, but the 2026 test moved from raising debt to proving that the book still converts relatively conservative underwriting into cleaner earnings and cash that is accessible to noteholders.

What truly changed is the center of gravity. A year ago the focus could still sit on establishing the Israeli note layer and bringing the vehicle to market. Now the question is no longer whether the platform exists. It is how much of the value created inside that platform remains available once secured debt, reserves, CLO structures, and the shareholder-loan layer are taken into account.

The strongest counter-thesis is that this reading is too conservative. One can argue that in the same year spreads on the asset book came under pressure, the company still refinanced funding, issued two new CLOs, maintained a 66.4% stabilized LTV, held total liquidity of $460 million, stayed compliant with every covenant, and successfully expanded the Israeli note. That is a serious counterargument, which is exactly why 2026 has to decide between the two.

What could change the market's reading over the next few months is fairly clear: lower collateral-protection expense, stable or lower hard delinquency, proof that the February raise is not quickly consumed by distributions or internal capital reshuffling, and a smooth in-house transition of healthcare asset management.

---