Three platforms, one bet, which I only saw afterward

The line about writing down the one assumption the deal dies without is going straight into my notes, and I haven't even funded anything yet. In your eight, was that sentence the refinance, or the exit cap? Feels like those are the same sentence wearing different clothes.

I own small multifamily directly and I did a version of this too, just with fewer zeros. Bought three buildings in an eighteen month stretch, all the same vintage, same submarket, same plan of pushing rents after light renovation. When insurance and taxes jumped, all three jumped together.

What I'd push back on slightly, thatch. Vintage concentration might have hurt you more than asset type. Eight positions inside two years means every one of them was underwritten in a similar rate and price environment, and no amount of spreading across property types fixes that. Money going in over six or seven years would have looked different even if you'd bought sunbelt multifamily the whole way through.

cairn's point about vintage is the sharper version of what happened here. Concentration in these portfolios shows up along axes most allocators don't track, and platform count is the weakest one.

The axes worth writing into a tracker: sponsor, property type, metro, business plan, capital structure at the property level, and the year the deal was underwritten. thatch's book was spread on the first axis and concentrated on the other five. Seven of eight sharing floating rate property debt with a cap purchased for the initial term is a single exposure, because the mechanism is identical. When the index rises past the strike, the cap pays, and when the cap expires the sponsor has to buy new protection at the price the market charges that day. That cost was not in the original budget in most 2021 and 2022 underwriting, which is why distributions across an entire vintage stopped in the same two quarters rather than in a spread-out way.

On capital calls, the dilution outcome thatch describes is ordinary rather than punitive. Most operating agreements let the sponsor issue new units at a valuation the sponsor sets, or convert the funded portion of the call into a preferred position ahead of non-participating members. Which of those applies is in the agreement, and it should be read before funding rather than at the moment the call arrives, because the practical choice at that point is fund or accept the stated consequence.

One practical test: for each position, write the single assumption whose failure ends the deal, then sort the list. Anything appearing more than twice is a sizing decision, not a diversification one. Terms and structures vary by offering, so confirm each one in the current governing documents.

The phrase rate cap ran out of term stopped me, and I'm mostly reading here since all of this is new to me. Does that mean the protection just expires on a date regardless of whether the loan is still outstanding? That seems like an obvious mismatch to leave in a deal.

rigging, that's how it works, and the reason is cost. A cap for the full loan term is much more expensive than one for the initial period, so buying the short one makes the year one numbers look better. It's the same instinct as an adjustable mortgage that's cheap for three years.

thatch, I came here wanting income without a second job and posts like this keep pulling me back toward the least exciting thing I can find. Eight positions and you still ended up with one bet is the part I'll remember.