One cap rate for the whole building is where mixed-use gets mispriced

My lender's appraiser blended it and would not be moved. I brought a sum-of-parts analysis to the review call on a 9 over 3 last year and the reviewer's position was that there were no separable interests, so there was nothing to sum. I lost that argument and the loan sized off the blended number, about $180k below where I thought the building was.

What actually got me the deal was the seller coming down, not the appraisal changing. So my practical answer is: use sum of parts to decide what to pay, expect the blend when someone else's money is involved.

The one boiler detail in your post is doing more work than either method. A building where residential and commercial share HVAC, share a service panel, share a roof drain system, that's genuinely one asset and blending is honest. A building where the ground floor has separate systems, separate meters, separate entrances, that's two buildings stacked and the sum of parts starts making sense.

I've renovated both. Splitting shared mechanicals after the fact on a four story building ran me $95k and three months of unhappy tenants, and that was a comparatively easy one. Ask for the mechanical drawings before you decide which method describes the building.

The blend isn't wrong, it's just under-documented in what you're describing. A defensible blended rate is NOI-weighted from the component rates. If retail is 35% of NOI at a 7.75% market rate and residential is 65% at a 5.5%, your weighted rate is around 6.29% and you can show your work. What that appraisal did was pick a number in a range and call it blended, which is a different activity.

Done properly, the weighted blend and the sum of parts give you nearly the same value. The gap in your two documents is probably the broker's component rates being optimistic, not the method.

meridian, that's the cleanest way I've seen anyone put it. If both methods use the same component rates they have to agree, so the disagreement is always about the inputs.

On the management complexity deduction the broker took: from running a service business, I'd want to know if it's a deduction to value or an expense in the NOI, because doing both is double counting and doing neither understates the cost. Managing 14 apartments and 4 commercial tenants is two workflows. My commercial billing, CAM reconciliations, estoppels at sale, none of that exists on the residential side.

I'm following about 70% of this. If the weighted blend and the sum of parts are mathematically the same thing when inputs match, why does anyone argue about method at all? Is the argument actually just about which retail cap rate is real?

vellum, that's how I understood what meridian said and it's the part that helps me most. The method fight is a proxy for a comp fight.

Which means the practical work is finding three unanchored ground floor retail sales in that submarket, and in my county the recorder doesn't publish prices, so I've been paying a broker for that data. Budget for it, it's not free.

thatch is right that it's two workflows and I'll put a number on it. I manage a handful of mixed buildings and my commercial side takes roughly three times the hours per tenant per year of a residential tenant, most of it in reconciliation and lease administration rather than maintenance. Whether that shows up as a higher management fee in NOI or a deduction to value, pick one, but pick one.

Is there a version where the sum of parts is genuinely right because the components can be separated? Condo the ground floor out as its own unit and sell it to an owner-operator. I've seen main street buildings where the retail is a separate condo unit from the apartments above.

beacon, in the rural markets I look at nobody condos anything, and the buyer pool for a small town mixed building is about six people who all price it off gross rent multiples anyway. Cap rates barely come up. I've had a seller tell me the building was worth eight times gross because that's what his father sold one for.

The thing I'd add is insurance, which neither method touches and which changed my numbers more than the cap rate debate. My 8 over 2 has a restaurant bay and the carrier priced the whole building off the highest-hazard occupancy in it. My premium is roughly 40% above what a comparable straight apartment building down the street pays. That's a permanent expense difference between mixed and single use, and it belongs in the NOI before anyone picks a rate.

Carriers differ wildly on how they treat cooking occupancies. Get quotes before you're in contract, not after.

thicket, the insurance point is the one I wish someone had told me. My difference was smaller, maybe 15%, but the deductible structure was worse and I didn't notice until a water loss.

wren, on your actual question: the appraisals I've seen on small mixed buildings all blended, and none of them showed weighting. meridian's NOI-weighted version is what I now build myself before I make an offer, and I treat the appraisal as a lender's constraint rather than an opinion of value.

I'm running the same problem on a 12 over 3 in another thread. What decided it for me is that the two methods answer different questions. Weighted blend tells you what the building's worth to the mixed-use buyer pool today. Sum of parts tells you which half of your income is exposed if a sector turns. You need the second one for risk even if the first one sets the price.

Question for meridian or quarry. If retail is 35% of NOI, at what share does a lender start treating the whole loan as a retail loan rather than a mixed-use loan, and does that shift the appraisal method too? I've heard thresholds thrown around but never from a lender in writing.

anchor, I've asked that three times and gotten three different answers, which tells me it's a policy per lender rather than a rule. Ask them for it in writing early, that's the only version of the answer that's worth anything.

There's a valuation input above all of this that the appraisal probably ignored. If the district allows more density than what's built, the land under a 14 over 4 may be worth more than the income supports, and the income approach will never show you that. I've seen two-story main street buildings in newly upzoned districts sell well above any cap rate math because the buyer was pricing the site.

Worth checking the current district standards yourself at the planning counter, since these got rewritten in a lot of places recently and the appraisal is eighteen months old.

parcel, so a building can be worth more empty than occupied? That reorders a few things I thought I understood.

quill, yes, and it's the thing I've been sitting on for eight years with raw corners. Land value and improvement income are two separate valuations of the same parcel and you take the higher one. On an existing 14 over 4 it's rare for the site to win, because you'd have to demolish income, but in a district that just went to four stories by right with no parking minimum it's no longer impossible.

wren, your third possibility is closer to true than the room is admitting. Both methods are ways of arriving defensibly at something near the last few comparable trades. The methods matter because they tell you when the comps are wrong for your building, and a building with 35% retail NOI, a shared boiler and a restaurant bay is wrong for most comps.

Practical note from losing a deal on this exact thing. Blended appraisals hurt you most when the retail is above market rent, because the appraiser haircuts to market and you lose value at the lower blended rate on income you're actually collecting. Two of my four bays were above market on old leases and the appraisal wrote off about $9k of NOI I had in the bank.

So check whether your 4 bays are above or below market before you pick a side. If they're below market, the blend is your friend.

I've read maybe thirty mixed-use LP packages and the sponsors who do this credibly show all of it. NOI-weighted blend with the component rates sourced, a sum of parts as a cross-check, the management cost in the operating expenses rather than as a mystery deduction, and insurance quoted rather than trended. The ones who show a single cap rate and a sentence about mixed character usually can't tell you what share of NOI comes from the ground floor, which is the first question I ask.

wren, if that seller's package can't produce a component NOI split for you on request, that's the answer to whether either number means anything.