How much of the REIT-to-private cap rate gap survives once you actually adjust the numerator?

Question

Working through a supplemental for a mid-cap industrial name and the headline story (REITs implying 6.3% against private appraisals around 5.1%) keeps shrinking as I strip things out.

Starting point: annualized cash NOI from the quarterly supplemental, backing out straight-line rent and above/below market lease amortization. Then I capitalize corporate G&A at the same rate, because the private comp set has no public-company overhead. Then I have to do something with the third-party management platform, the development pipeline at cost, and about $180m of JV interests carried at book.

By the time I finish, the 120bp gap looks more like 55 to 70bp depending on whether I capitalize all of G&A or only the portion that would survive a takeout. And the private side of the comparison is appraisal-based, so it lags.

For those of you who actually run this as a deployment screen rather than a talking point: what do you do with the platform and the development book, and do you adjust the debt to market? Secured debt struck years ago at low coupons is carried at par on the balance sheet, and marking it down widens the implied discount to NAV without changing anything about the real estate.

Rena Expert · Accepted Answer

The adjustment set you have is close to what a takeout buyer would run, and the shrinkage you are seeing is real. Most published versions of the gap use unadjusted EBITDA-style NOI over enterprise value, which flatters the REIT side by 40 to 80bp on the same portfolio.

On the platform and development book, the cleanest treatment is to value them separately and remove them from both the numerator and the enterprise value, rather than trying to fold them into a cap rate. Third-party management fee streams get a multiple, not a cap rate, and the multiple you pick swings the answer more than anything else in your model. Development at cost is worth cost only if the pipeline pencils at today's exit assumptions, which for industrial started in a 4-handle basis and is now leasing into a slower market. Haircut it or exclude it.

Mark the debt. If you do not, you are comparing a levered public entity carrying below-market fixed debt at par against unlevered private appraisals, and the mismatch shows up as a discount that no acquirer can actually capture. Marking secured debt to market usually cuts implied NAV discounts by a meaningful chunk on names that termed out early.

The whole exercise rests on the private appraisal side being current, and it usually is not. Appraisal cap rates move in steps with a lag, so part of the gap you are measuring is timing rather than mispricing.

Where this bites in deployment: a 55bp adjusted gap can be arbitraged by the REIT issuing equity and buying assets, and it can also sit there for eight quarters while rates chop. Sizing the position to the gap alone assumes convergence has a schedule. It does not.

sonia_alvarez · Answer

The debt mark is the part most write-ups skip. I pulled a fixed-rate schedule off a net lease name and the weighted average coupon was under 4% with a five year average remaining term. Carrying that at par while the private comps are unlevered appraisals produces a discount that only exists on paper. Mark it and a chunk of the gap goes away.

hannah_west29 · Answer

@fathom I ran a version of this and got a wider answer than you, so one of us has a bad input. I capitalized only cash G&A net of the fee income it supports, which cut the drag roughly in half, and I left JVs at book on the argument that the pro rata NOI is already in my numerator. If you are also including pro rata JV NOI and then adding book value of the JV interests to the asset side, you are double counting. Which side did you put them on?

paulgrant17 · Answer

Naive question from someone reading over your shoulders: is the cash NOI in the supplemental already net of recurring capex, or do you have to reserve for that separately?

How much of the REIT-to-private cap rate gap survives once you actually adjust the numerator?

4 replies