31 percent of a bridge fund's tape matures inside nine months, all still marked performing

Vellum has the right end of it, and the document language is where the next two days should go. Three things worth pulling verbatim. First, the definition of performing, and specifically whether an interest reserve funded out of loan proceeds counts as the borrower paying. A loan can look current for nine months on money the fund lent it. Second, the incentive fee crystallization language. If the promote is paid on distributed income annually with no clawback against later principal losses, the manager gets paid on the interest that shows up and the fund eats the write-down two years later. Third, the gate. Five percent of NAV per quarter is the cap on total redemptions, not any single investor's share of it, so read whether unfilled requests roll forward with priority or go back to the end of the line each quarter. Those two designs produce very different exits when everyone asks at once. The extension amendments on those twelve loans are worth pulling too. An extension with a fresh appraisal, a principal curtailment and a rate bump is a lender in control. An extension that only moves the maturity date is a lender hoping.

The subscription line, when sized around fifteen percent of NAV, is the number worth starting with, because it does two things to a headline yield. It funds new loans before capital calls settle, which lifts the yield on invested equity, and it sits senior to investors in the fund's own capital stack. Worth asking what the line is priced at, whether the borrowing base excludes loans past original maturity, and what happens to distributions if the bank cuts the advance rate. A lender to the fund that shrinks the line during a stress window forces the manager to hold cash instead of paying investors. Then the valuation policy. The actual written policy, not the summary in a pitch deck, and who marks the loans. If it's the manager with an annual third-party review of a sample, the marks are an opinion until an exit proves otherwise. Also worth checking whether servicing and any special servicing fee goes to an affiliate of the manager, because a workload created by extensions can be a revenue line for the same people deciding to extend. That's disclosed in most documents and almost nobody reads that page. And the auditor's opinion on the last two years, including any change in auditor. A switch inside an otherwise quiet multi-year record is worth an explanation.

62 percent of principal in one metro cluster with as-stabilized LTVs is one bet twice. If values in that cluster came off ten percent, a stated 32 percent cushion on a rehab loan is largely gone before the property is finished, and the exit becomes a sale into the same softness the borrower was trying to sell into. What any lender or investor should demand before wiring anything is the tape recut three ways. As-is value at origination rather than stabilized. Sponsor cash equity behind each loan, actual dollars in, not paper equity from a marked-up appraisal. And origination vintage by quarter, so it's clear how much of the book was written at 2021 and 2022 basis versus last year. A book originated in the last eighteen months at conservative as-is values behaves differently from the same LTV written at the peak, and one line of weighted average hides that completely. On a wait-or-subscribe question, say a quarter of income given up is roughly $5,200 on $250,000 at 8.4 percent. That's a cheap price for seeing whether 31 percent of the book actually pays off. Debt aggregates have beaten equity indices this year, 5.5 gross against 4, and that gap is exactly why so much capital has closed into these funds. Money arriving fast is not the same as underwriting getting better. If the next tape shows the maturity wall refinanced out at par with new appraisals, that's a good sign. If it shows twenty extensions instead of twelve, that tells a different story.