Reading three competing term sheets on a 12 unit refinance, and why the lender paid option deserves the most scrutiny

Your DSCR arithmetic is right, which means their NOI is not yours. Agency underwriting will typically haircut your number: management fee imputed at 4 or 5% even if you self manage, replacement reserves set by their schedule rather than your 250, market vacancy rather than actual, and any income line they consider non recurring stripped out. Run it at 5% management and their reserve number and see if 1.05M pops out at 1.25. The likely result is that you land within a few thousand. Ask the broker for the lender's underwriting summary, the actual grid. If he cannot get you that, that is information too.

The comparison you built is missing the parts that live outside the rate, and each of these three sheet types has one. On C, the 1% exit fee is 11.5k on 1.15M and it is payable whenever you leave, including at maturity. That is not in your five year cost comparison because C is a three year term, so you are also buying a refinancing event in year three at whatever rates are then. The lender-paid comp is inside the 7.40%. Debt funds price a rate and pay the broker out of the spread, so the enthusiasm is honest in the sense that he is not hiding a fee, and self interested in the sense that his comp there is usually larger than a point. Ask him this instead of asking what he is paid: what would the rate be on this same loan if you came to the fund direct. That question has an answer he can get, and the gap is his comp. On B, yield maintenance with a lockout period is a different animal from step down. If there is a lockout you cannot prepay at all during it, at any price. Check whether the credit union sheet says defeasance or yield maintenance and whether there is a floor rate in the calculation. A YM floor at the note rate means you effectively cannot exit early in a falling rate environment without paying real money. On A, the step down is clean but check the assumption language. Agency small balance loans are often assumable with lender approval and a fee, and if you might sell in year four that is worth actual dollars to a buyer. Last thing, go find the broker fee agreement you signed. The standard form says the fee is earned on any lender introduced during the engagement and for 12 months after. Which means if you walk from all three and close with the credit union yourself in eight months, you may still owe the point. Whether that holds up depends on your state, and that is a lawyer question, but read the clause tonight.

The credit union being fifteen minutes away is worth more than your model can show, especially in a small market. By the third loan, a local lender will often do things for a borrower it knows that no agency shop would consider. That does not make B the right answer at 980k proceeds. It does mean you should not burn them by going dark after they wrote a sheet.

The gap between your 1.51 and their 1.25 is almost certainly underwritten NOI, as @wendy_park48 says, and there's usually a second constraint stacked behind DSCR that binds first: loan to value against the lender's appraised value, and on small balance product a per unit or minimum loan floor. Run 1.05M against a plausible appraisal at their cap rate. If they're capping your underwritten NOI at, say, 7% and getting 1.5M in value, 1.05M is exactly 70% LTV and that's your binding constraint, with DSCR sitting slack above it. Ask which of the two is binding. The answer determines whether improving operations raises proceeds or does nothing.

On the comp question, the mechanism is worth being precise about. Borrower-paid comp is a fee you write. Lender-paid comp is priced into the rate, so it costs you every month for the life of the loan rather than once at closing. On a three year interest only loan at 1.15M, even 100 basis points of embedded comp is roughly 11.5k a year. That's the arithmetic that makes "costs you nothing" wrong as a description, without making the broker dishonest. @raj_mehta14's version of the question, what the direct rate would be, gets you the number without an accusation.

How hard to push: as hard as you want, in writing, once. A broker who won't put comp structure in an email is telling you about the rest of the relationship. Confirm all of it with the lenders directly before you rate lock, since term sheets are indicative and the binding numbers appear in the commitment letter.

I'm new to this and mostly reading, but the thing that jumped out is that A and C aren't the same product and your 90k comparison treats them like they are. One amortizes and one doesn't. Over five years A pays down principal and C doesn't pay down anything. Is that in the 90k or is that on top of it?

@bmoreno23 that's fair, and it usually isn't in that kind of comparison. Ledger, if your 90k is pure interest and fees, add roughly 60k to 70k of principal amortization on A over five years, and then subtract the fact that C hands you 100k more today that you can put to work. Whether that trade is good depends entirely on what the extra 100k does. If it sits in the operating account it's expensive money. If it's the down payment on the next building it might be the cheapest capital you'll see this year.

That's the actual decision in front of you, and no term sheet answers it.

Worth asking the agency lender what capital expense reserve they will require at closing on top of the ongoing 250 a unit. On a 12 unit with any deferred roof or boiler work they can escrow a lump sum up front, and that can eat 30k of proceeds after everyone has already agreed on the number. Get the property condition report ordered early so it does not show up two weeks before your bridge matures.