Debt-only on platforms netted 8.9 percent. An extension clause nearly ate it.

I'm still at the stage where the words trip me up. When you say a loan against a property, are you actually the lender on record, or are you buying a piece of something the platform holds? Because in the two offerings I've read the borrower's name is on a note to an entity I've never heard of, and my money buys units in that entity. First position keeps getting used for both and I can't tell which one I'd be.

vellum has the right end of it. tender, in the three platforms you used, were you a member of a special purpose entity that held the note, or did you buy a participation interest in a loan the platform's affiliate originated? Those behave differently the day the platform itself has trouble, and the difference lives in the operating agreement rather than the deal page.

Also curious whether you ever saw the servicing agreement. Not the summary of it. The document.

The extension clause is the part I'd push on. Seven extra months at the original coupon on an 11 month note means your annualized on that position dropped hard even though you were paid in full. Did the forbearance require consent from participants, or could the servicer grant it unilaterally? Every one of these I've read gives the servicer sole discretion, which is fine right up until it isn't.

fathom, two of the three were member interests in an SPE that held the note, one was a participation in a loan the affiliate originated and retained a slice of. The participation one I liked less on paper and it performed best, which I refuse to draw any conclusion from. I got the servicing agreement on two platforms by asking investor relations directly. The third sent me a summary and said the full agreement was available on request to accredited holders, which I read as no.

canton, unilateral. Servicer sole discretion, no participant vote, notice after the fact. On that position the effective annualized came out around 6.5 percent because interest accrued at the original rate through the forbearance and I lost the redeployment. The book absorbed it because it was 4 percent of capital.

The honest cost of the ladder is the idle cash. Money comes back on a Tuesday and the next offering I actually want might be three weeks out. I averaged something like 11 days uninvested per turn.

11 days per turn across 21 positions is roughly 230 days of dead capital, so call it 6 or 7 percent of your invested time earning whatever your cash account pays. That is most of your 80bp gap right there, and it's structural rather than bad luck.

The number I'd want next to your 8.9 is what short treasuries paid over the same 18 months, whatever that was in your window, because the spread is what you actually got paid for taking sponsor and collateral risk. If the spread was 350bp for first-loss-adjacent exposure to small bridge loans, that's a real decision. If it was 150bp, it's a hobby.

What was the LTV basis on the positions? Stated against as-is appraisal, or against as-completed value on a business plan that hadn't happened yet? A 70 percent number means two different things depending on which, and the offerings that quote as-completed rarely say so in the same sentence.

On the one that went sideways, what did the collateral appraise at going in and what did the servicer think it was worth when they chose forbearance over foreclosure? If those two numbers were close you got lucky on timing. If the second was well below, the forbearance was the servicer avoiding a writedown.

How did you land on 8 percent per sponsor? I'm trying to set caps before I fund anything and every number I pick feels arbitrary. Did you also cap per platform, given that platform failure hits every position on it at once?

I've done the contractor work on plenty of these deals, and the thing that decides whether a bridge loan pays is whether the draw inspections are real. On the projects my team has worked where the lender sent someone out before releasing money, the schedule held. Where the lender took photos by email, it didn't.

ledger's cash drag point is the one I'd underestimated. If you can only really deploy every three weeks, then a 6 month note isn't a 6 month note, it's a 6.7 month note at the same dollar return. Does that change how you'd rank a 12 month position at 9 against a 6 month at 9.5, tender?

I'll disagree with the shape of this, though the result is fine. 21 positions at $1.5k to $3k across three platforms is a lot of separate documents, separate tax reporting, and separate maturity dates for a book under $50k. I did something close to it for two years and moved most of it into two pooled debt funds instead, took roughly 70bp less on the stated rate, and got back a weekend a quarter plus consolidated reporting.

The tradeoff is real. In the pooled fund I can't decline the deal I don't like, and the manager's incentive on a troubled loan is not mine. tender's version gives him that veto. I just think the veto is worth less than people assume when you're looking at a deal page and a set of numbers the sponsor picked.

Are debt positions like these generally open to non-accredited investors, or was tender accredited across all three? I keep finding platforms where the interesting-looking loans sit behind a gate and the open ones are pooled products.

tender, run the counterfactual. You had one borrower stop paying out of 21 and full recovery. What does the book look like if two stop paying and one of them recovers 65 cents on principal? That's not a stress scenario, that's an ordinary vintage for small bridge paper.

I'll do it since I'm curious. Take his $46k, weighted stated 9.7, and assume 8 percent of capital in the two bad positions. One recovers whole at the original coupon after a long forbearance, one recovers 65 percent of principal and no interest. Loss of principal is about 1.3k, plus roughly 250 of foregone interest on that slug, plus the drag from the first one. Blended lands somewhere near 6.2 to 6.5 percent over the same 18 months, still positive.

Which is the actual argument for the ladder, sextant. It isn't that 8.9 is great. It's that at 21 positions the bad ones move the answer by a couple hundred basis points rather than deciding it. At five positions the same two defaults put him near zero.

The thing I'm taking from this is to stop reading the rate first. The rate was 9.7 and the answer was 8.9 and the difference was clauses and calendar.

Worth saying that raw land loans on these platforms behave nothing like the bridge loans tender is describing. No cash flow, no improvement, and if the borrower walks the collateral is a parcel that might take a year to sell in a thin market. I've looked at a lot of them and funded none.

tender, one clause to look for next round, since you're already reading the right documents. In a couple of these SPE agreements the manager can amend the note terms, including maturity, without participant consent, and separately can advance protective payments and recover them ahead of participants. The second one is the quiet reordering. You can be paid in full on paper and still sit behind a servicer advance that grew for seven months.

Ask for both the servicing agreement and the SPE operating agreement before funding. The platform that would only give you a summary told you something.

Genuine question, not a challenge. 8.9 percent net, no tenants, no toilets, no calls. Why would anyone with $46k buy a small rental instead? I know the answer involves borrowed money and appreciation, I just want to hear someone who did the debt version say it.

The mechanics I've never seen anyone spell out. When a position pays off, what actually happens? Does the money appear in a cash balance on the platform, do you get an email, is there a statement showing interest separately from principal? I have avoided funding anything partly because I don't know what the ordinary week looks like.

In the small rural markets I follow, the appraisal on a bridge loan is often supported by two comps from eleven miles away and a year old. The rate on those offerings is usually the highest on the platform, which tells you what the servicer thinks too.

Reading this mostly convinced me to keep waiting. 8.9 percent for 18 months of clause reading is fine, and I'd rather hold cash and buy a parcel outright when the price is wrong in my favor. Different appetite, no argument with the execution.