how much contingency on a conversion before it's just a construction bet

Question

Underwriting a 300,000 gsf 1970s tower being marketed for conversion. Sponsor's numbers, roughly: acquisition $60/gsf so $18M, hard costs $230/gsf so $69M, soft at 18 percent of hard, total near $100M before closing costs. They show 65 percent efficiency, so 195,000 net rentable, 240 units averaging 810 sf. That's about $416k per unit all in.

At a 5.25 exit cap that needs roughly $21,800 of NOI per unit, call it $2,900 gross monthly, which is $3.58 per net sf. That's an aggressive rent for this submarket but not impossible in the best building.

My problem is the contingency. They carry 5 percent on hard costs and nothing on soft. On a conversion of a building nobody has opened up yet, 5 percent reads like a typo. I'd want 12 to 15, but at 15 percent the deal doesn't clear my hurdle at all, which tells me either my contingency is wrong or the deal is. What are people actually carrying, and which line items are the ones that blow through it?

Rena Expert · Accepted Answer

Split the contingency into two, because they behave differently. Design contingency covers the gap between a set of schematic drawings and a fully coordinated set, and on a conversion at pre-GMP stage it's commonly carried at 10 to 15 percent of hard cost, burning down as drawings complete. Construction contingency covers unknown conditions and change orders once a price is signed, typically 5 to 10 percent depending on how much destructive investigation happened before pricing. A sponsor showing a single 5 percent number on an unopened building is either carrying a design contingency inside the trade allowances without saying so, or they haven't got a design contingency. Ask them to show the estimate line by line and identify every allowance, because allowances are the place a soft budget hides.

The items that break conversion budgets in practice: curtain wall or window replacement, which can move a budget by tens of dollars per gross square foot on its own; vertical riser and shaft creation, which is where structural reinforcement and slab coring costs appear; egress, since a second stair or pressurization requirement discovered in code review is expensive and hard to design around; and hazardous material quantities that come back higher than the survey estimated. Note that only the last is truly unknowable. The first three are knowable with enough pre-purchase work, which is why the amount of investigation done before pricing should drive what contingency you demand.

You're also missing contingency on time. Add carry sensitivity, because a six month delay on a $70M construction loan is real interest plus extended general conditions plus a leasing season, and none of that touches the hard cost contingency line.

Your instinct that the deal fails at 15 percent is useful information about the deal. Plenty of these don't pencil, and the discipline is in saying so. Before you conclude, though, run the efficiency sensitivity @plumb raised. If 65 percent is aspirational, the rent needed at $3.58 climbs, and no contingency assumption saves it.

nataliehsu · Answer

The 65 percent efficiency is the number I'd chase before the contingency. On a deep-plate seventies tower with a big central core I've seen conversions land closer to 55 to 60 once corridors and the new risers get drawn properly. At 58 percent you lose about 20 units and the whole thing moves.

sonia_alvarez · Answer

Construction lenders I've listened to want their own contingency sized independently of the sponsor's, plus a funded interest reserve for the full term with a stress on the delay case. If the sponsor's 5 percent is what's in the loan model, ask whether the lender has already required a higher number and the equity is being shown a rosier version.

hannah_west29 · Answer

Ran your numbers at 58 percent efficiency. 174,000 net rentable, 215 units at 810 sf, and total cost barely moves because hard cost is driven by gross area, so cost per unit goes to about $465k. At the same 5.25 cap that's $24,400 NOI per unit, roughly $4.00 per net sf in rent. That's a different building in a different market. The efficiency assumption is worth more than the entire contingency argument.

bmoreno23 · Answer

What is efficiency measuring exactly? Rentable divided by gross, so the losses are corridors, stairs, elevators, and mechanical rooms?

raj_mehta14 · Answer

@ember that's it, and in an office building the core was sized for office elevatoring and office bathrooms, so you're carrying a core built for a use you're no longer putting there.

@sextant separate question, is there a completion guaranty and who signs it? If a single-purpose entity with no balance sheet is the guarantor, the cost overrun risk sits with the equity regardless of what the contingency line says.

dmitri_v6 · Answer

And read the unknown conditions clause in the GMP. Guaranteed maximum price on an existing building usually carves out concealed conditions, so the guarantee applies to everything except the risk you actually bought the building for. Some contracts cap the carve-out with a shared savings and shared overrun band. Which version they signed is worth more than the contingency percentage.

priya_n21 · Answer

Nobody has questioned the $3.58 rent. In a submarket where that's aggressive today, it has to be achievable in 20 months by 240 units all delivering at once.

how much contingency on a conversion before it's just a construction bet

8 replies