[f1yght]

They're also difficult because modern office buildings just don't lend themselves to small apartments very well. One of the conversions in New York cut a giant hole through the middle of the building so they could have interior units that get light. Giant modern office buildings take a ton of effort to convert.

[RC_ITR]

This is actually one of the reasons SF is so well positioned to do this - our floorplates are notoriously small.

Take these new buildings for example:

SF Offices:

- Salesforce Tower: 1.4mn square feet over 61 floors (23k sqft/floor)

- 250 Howard: 734k sqft over 43 floors (17k sqft/floor)

- 350 Mission: 455 sqft over 30 floors (15k sqft/floor)

Office Elsewhere:

- 10 Hudson Yards (NYC): 1.8mn sqft over 52 floors (35k sqft/floor)

- BMO Tower (Chicago) - 1.5mn sqft over 51 floors (29k sqft/floor)

- Block 158 (Austin) - 720k sqft over 35 floors (20k sqft/floor with bottom floors c.3x larger than top floors)

Newly built large residential buildings in SF were usually 15-20k sqft/floor (e.g., The Avery), so there are actually very few buildings here that don't work for residential.

[alephnerd]

The square footage isn't the issue - it's the water+gas pipes along with windows and lighting. You'll need to do a floor by floor renovation of water+gas lines along with thinking about how to architect apartments in a way to get lighting.

All that is a significant investment in a city with very high labor prices (because most skilled trade types got priced out), so it might not be economical to do such a redesign at the moment.

[RC_ITR]

The alternatives are building a building from scratch while these offices sit empty, no? Certainly you must agree retrofits are cheaper than building an entire new building. Also fyi, SF isn’t even allowing gas in new buildings anymore.

I agree with your logic when CRE yields 1.5x the rent of residential, but the model needs to be updated now that it's no longer the case.

Is your base cases these offices sit empty forever? Get torn down?

I think that line of thinking is the one massively ignorant of the realities of real estate.

In a city with almost no residential rental vacancies? That's the end-game?

EDIT: For everyone direct-quoting:

The conversion alone might cost about $400 or $500 per usable square foot, Mr. Bernstein added, and would in many cases be more expensive than building a new development.

A recent Moody’s analysis of New York offices found that just 3 percent of the buildings it tracked would be viable for apartment conversions. The median rent for apartments in New York is $55 per square foot, which just 36 percent of office properties now fall at or below — and on top of that, there’s all the cost of conversion.

$500 is nowhere near the fully-loaded cost to build new in any US city. The quoted man is comparing the cost to the delta in rent to his out-of-date understanding of what offices can yield.

It is just a simple fallacy to believe that SF/NYC offices can command $75/sqft at any reasonable occupancy level.

The economics of conversions only don't work when offices command premium rents, it's not some axiomatic fact about the world (the way it was before the pandemic).

[alephnerd]

> Certainly you must agree retrofits are cheaper than building an entire new building.

Yes I agree with you, but who pays for it? JLL and CBRE are large and diversified enough that they can eat the cost of their SF commercial properties being vacant.

It doesn't make sense for them to spend funds on retrofitting their properties into apartments when they can regeotiate their mortgage commitment to be much more amenable in the short term, especially when similar residential properties in SF like NEMA, The Gateway, and others continue to have elevated vacancy rates.

You can't force private businesses to do residential retrofits - they'll only do it if there is a viable financial case for them to do it.

Most luxury residential property in SF is owned and managed by Greystar, Avalon, and UDR. These are different companies from JLL and CBRE who own the vacant office buildings. At that point, who pays for the retrofit - JLL+CBRE or the residential landlords line Greystar+Avalon+UDR? It doesn't make sense to either because they are large, diversified international companies that have better opportunities to deploy the capital they have at hand (eg. NoMA in DC, NYC, etc). Why spend $1-2 billion renovating+retrofitting when you can spend the same amount with a better RoI in other markets.

[RC_ITR]

I really can't emphasize enough that residential in SF generates $55/sqft/year, which in almost every US market outside NYC, is more than you get for commercial anyway. If SF commercial generates meaningfully less than $55/sqft/year (which it is clearly on track to do), you don't need to force anyone to do anything.

To your capital allocation point, you're lucky to get $40/sqft/year in DC.

This is real money you're talking about. It's not some pie in the sky idea that a residential building in SF is profitable.

The line of argument that conversion can't be done is just repeating things that were true before the pandemic and that commercial lenders/developers hope will be true again.

If buildings default, all the commercial owners/lenders eat the capital loss (hence their current loud and public press tour about why conversions are not feasible), a new buyer buys at a lower cost basis and then converts. Why is that extremely practical outcome so outlandish?

SF office buildings are already trading at 75% 2019 prices. If you (as the new low cost basis buyer) can yield even 25% more by residential converting why wouldn't you?

[zamnos]

I hear you on all that, but that seems to be an argument that corporations with headquarters outside the bay area (like landlords who live outside the bay area) shouldn't be allowed to own real estate in San Francisco.

[alephnerd]

Not really. Adding a regional ownership test de facto violates the Commerce Clause in the Constitution.

[zdragnar]

The alternative, as I mentioned in my earlier post, is to demolish the building and start over. That was the cheaper route, according to the contractor who had gotten a job to retrofit an existing office space.

To run new 240v to each apartment, plus water, plus septic, plus whatever additional load bearing requirements there are means you're effectively gutting the building anyway. Then, on top of that, you have to deal with light, fire escapes, etc. which may not be easy to do with the existing building's footprint.

[RC_ITR]

We just fundamentally disagree if your argument is tearing down high-rises is cheaper than retrofitting them. In the history of mankind we've only every voluntarily demolished like 30 buildings higher than 35 stories for that exact reason (and most of them were in situations where new taller highrises replaced them). [0]

I think your contractor is just way off base.

Maybe they're correct for 1-3 story office parks, but there aren't really many of those in SF.

[0]https://en.wikipedia.org/wiki/List_of_tallest_voluntarily_de...

[bilbo0s]

The point is that you can do these conversions in places where you'll make a ton of money from them. No one's talking about converting them to 3 or 4 thousand dollar a month apartments. I think most people understand that you do a conversion in lower Manhattan and you're doing it for the "multi million per" class. So there is plenty of money to pay for these in Boston, NYC and San Fran.

Of course, that also means there is plenty of money to pay for a teardown. But being torn between making a killing tearing down and rebuilding vs making a killing doing a conversion is a good problem to have.