Building a book of eight positions: which axis do you diversify on first

Sponsors first, and the reason is that sector risk shows up in the numbers where you can see it, while sponsor risk shows up in behavior where you can't. A bad multifamily vintage prints in the rent comps and you know what happened. A sponsor who commingles reserves across deals doesn't print anywhere until the wire doesn't come.

At $300k though your sponsor spread is expensive. Six or seven relationships means six or seven sets of references to call, six or seven sets of K-1s, and you'll be the smallest LP in every room. I'd do four sponsors and accept two checks each, which is roughly your second option, arrived at from the first axis.

Coming from the building side, the sector split you drew isn't as diversified as it looks. Small bay industrial and retail strip both live or die on the same small business tenant, and in a downturn those tenants go at the same time. Multifamily and self storage at least respond to different things. If you want real sector spread with four slots, I'd rethink which four.

solder's point is the one I'd chase. My list has the same problem, I built it thinking I had five markets and then noticed three of them are one employer.

Question rather than an answer, hollow. If you go seven sponsors, how do you actually keep track of seven reporting cadences? I've watched people on the service side get pulled into cleaning up an investor's records because nobody standardized anything and the K-1s arrived in April, June, and September. That's not a reason to concentrate, and it is a real cost you should price into the sponsor spread option.

The vintage axis is the one I'd defend, and nobody ever picks it. Every value-add deal I've been on the construction side of blew its capex budget by 15 to 30 percent, and the ones that survived it were the ones that bought cheap enough to absorb it. Entry basis is the whole cushion. If you space eight checks over 24 months you're only spanning two vintages anyway, which is barely a hedge. To actually spread vintage you'd need to space over five years, and then you're holding cash for years earning nothing. That's the reason it loses to the other two, not because it doesn't matter.

harrow, 15 to 30 percent over budget on capex, is that on the renovation line only or the whole project cost? I'm trying to get my own numbers honest and that spread would change what I hold in reserve.

I've drawn this exact grid four times and the thing that finally moved me off it was realizing the axes aren't independent. The four sponsors you've done real work on probably specialize, so choosing them already chose your sectors. And you can't time vintage, because you commit when a sponsor happens to have a deal open, not when you decide the cycle looks good. So sponsor selection eats the other two whether you plan it that way or not.

Which means my vote is sponsors, by default rather than by conviction.

At $300k I'd take four sponsors and be patient about it. Eight checks in 24 months is a pace, and pace is how people end up in the deal that was open rather than the deal that was good. I've held land for eleven years waiting on one road, so my tolerance here is not typical.

On your $1M question, yes it changes. At $1M you can be a $150k LP, and $150k gets your calls returned, gets you the actual rent roll instead of the summary, and sometimes gets you a look at the deal before the general list. That access is worth more than the eighth position.

orchard's last paragraph is the real answer to the $1M version and it goes the opposite way from the diversification logic. Concentration buys information. I'm on the active side and the deals I get shown first are the ones where I've already closed with the person twice. Same dynamic on the LP side. Six sponsors who each think of you as a $25k name will show you exactly what they show the mailing list.