Debt-only positions on lending platforms can net around 8.9 percent, and an extension clause can nearly eat the gap.

The terminology trips a lot of people up. When a platform says a loan is against a property, the real question is whether the buyer is the lender on record, or buying a piece of something the platform holds. In many offerings the borrower's name is on a note to an entity the investor has never heard of, and the money buys units in that entity. First position gets used loosely for both, so it's worth pinning down exactly which one applies before committing.

Vellum has the right end of it. Tender, on platforms like these, the key question is whether an investor is a member of a special purpose entity that holds the note, or buying 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 worth asking whether anyone has actually seen 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, in a book built that way the positions usually come out as member interests in an SPE that holds the note, with the occasional participation in a loan the affiliate originated and retained a slice of. The participation reads worse on paper and often performs better, which is not a conclusion worth drawing anything from. Servicing agreements can generally be had by asking investor relations directly. A platform that sends a summary and says the full agreement is available on request to accredited holders is telling you no. Canton, unilateral. Servicer sole discretion, no participant vote, notice after the fact. On a position like that the effective annualized lands near 6.5 percent, because interest accrues at the original rate through the forbearance and the redeployment window is lost. A book absorbs it when the position is 4 percent of capital. The honest cost of the ladder is the idle cash. Money comes back on a Tuesday and the next offering worth taking might be three weeks out. Something like 11 days uninvested per turn is a realistic average.

Eleven days per turn across twenty one positions works out to roughly 230 days of dead capital, call it 6 or 7 percent of invested time earning whatever a cash account pays. That accounts for most of an 80 basis point gap, and it is structural rather than bad luck. The number worth putting next to an 8.9 percent return is what short treasuries paid over the same eighteen months, because the spread is what actually gets paid for taking sponsor and collateral risk. A 350 basis point spread for first loss adjacent exposure to small bridge loans is a real decision. A 150 basis point spread is 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 offerings that quote as completed rarely say so in the same sentence. On the position 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, that's timing working out. 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?

The thing that decides whether a bridge loan pays is whether the draw inspections are real. On projects where the lender sends someone out before releasing money, the schedule tends to hold. Where the lender accepts photos by email, it usually does not, and that is the gap an extension clause ends up covering.

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? For small bridge paper that is an ordinary vintage, well short of a stress scenario.

Happy to run it. 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. An 8.9 on its own is a modest number. What matters is 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 lesson in this one 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. There is no cash flow and no improvement, and if the borrower walks the collateral is a parcel that might take a year to sell in a thin market. Careful lenders review a lot of them and fund very few.

Tender, one clause to look for next round, since you are already reading the right documents. In some 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 reordering that goes unnoticed. 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 provide a summary told you something.

A genuine question rather than a challenge. 8.9 percent net, with no tenants and no calls. Why would anyone with $46k buy a small rental instead? The answer presumably involves borrowed money and appreciation, but it's worth hearing it from someone who ran the debt version.

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 small rural markets, the appraisal behind 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.