How many deals before an LP book stops being a bet on one sponsor?

Eight deals in one asset class in one region is one bet with eight invoices. The correlation that kills you is the rate environment and the insurance market, and those hit every southeast multifamily deal in the same quarter.

Geography matters less than debt structure in my reading. Two sponsors in the same metro, one with fixed agency debt and one with floating bridge debt and a cap expiring, are in genuinely different situations.

Sorry to ask the basic thing, but why can't you just put 400k across two funds instead of eight deals and let them do the picking?

rigging, then you're picking two managers instead of eight deals and you don't see the assets before your money's committed. It moves the diligence rather than reducing it.

The argument for going deeper with fewer sponsors is that your second and third deal with the same operator is the only diligence that's actually cheap. You've seen their reporting, you've seen how they handled a bad month, you know whether the distribution shows up on the day they said. That's information nobody outside the LP base has.

Counting positions assumes each 50k is equally risky. It isn't. A stabilized cash-flowing asset with 55 percent fixed-rate debt and a heavy reposition with a bridge loan don't belong in the same count.

Something nobody's said about deploying over three years: the vintage spread is doing more for you than the sponsor spread. Eight deals bought in the same twelve months at the same basis is a single entry point. Spreading purchase timing across three years of pricing is diversification you get almost for free.

thicket's point is the strongest one here. Run it as a grid instead of a count. Sponsor, vintage year, asset class, metro, debt type. Eight positions across five variables means you can't cover all of them, so pick which two you refuse to concentrate in. Mine would be vintage and debt structure, because those are the ones that took deals down in the last cycle and no amount of sponsor quality fixed them.

Or don't put all 400k in equity. A slice in debt positions behind the same kind of assets changes your loss profile more than adding a ninth sponsor does.

From what I see on the management side, sponsor quality shows up in the boring operational stuff. Do they staff the site properly, do they pay vendors on time, do they let deferred maintenance accumulate to make the current year's numbers look better. You can't get any of that from a deck, and you can get a lot of it by asking the property management company who worked the asset before.

Four sponsors at two deals each is what I've drifted into and I wouldn't call it a compromise. The second check is where you finally get honest reporting, because by then they want a third one from you. First-time LP with a sponsor, you get the polished quarterly. Third deal in, you get a phone call when something breaks.

meridian, the flaw in the eight-position plan is that you'll do your best underwriting on deals one and two and your worst on seven and eight, because by then you're filling a target number and the good deals aren't open when your capital is free. Deploying to a schedule is what makes people take the deal that's available rather than the deal that's good. I'd rather hold cash for four quarters than hit a count.

Also read what each operating agreement says about the manager's right to bring in additional capital ahead of you. Two deals with the same sponsor can have completely different preferred-equity carve-outs, and I've seen one where the manager could admit new capital senior to the existing LPs with no consent requirement at all. Your diversification count means nothing if three of the eight can subordinate you at the manager's discretion.

I showed up expecting one number back and instead I've got five variables and a warning about preferred equity jumping ahead of me. Genuinely less clear than when I started reading, which I suspect is the correct outcome.