Deal Breakdown - Seemed Too Good, and it Was: 1257 Bush Street, SF
What at first looked like a home run sold for 20% below the list price
Like most things in life, if a deal looks too good to be true, it probably is.
I'm going to walk through one that we recently evaluated and passed on, despite our initial screen seeming very positive.
Consider this a stream of consciousness reenactment of how I kick the tires on a new deal.
I chose this one because when it first came across my desk, it seemed too good to be true: Cheap for the location even at the asking price, nicely maintained and a clear path to upside without a lot of capex.
As it turned out, the property sold for 20% below asking and we didn’t give it a second thought.
Sound strange? Let's dive in:
Property Summary
35 residential units, 34 studios and 1 one-bedroom
Nice layouts, balconies and stellar views
Concrete construction, 1960s vintage
Asking price $7,400,000
23,666 gross square feet
$313psf or $211,000 per door
Average in place rent of $1,600/mo
Stated 5.0% going-in cap
First Impressions
Initially I didn't love the high studio count, since the neighborhood is overwhelmingly studios so you'll have a lot of competition from lower priced rental options. But after seeing them live, these are not your average janky tendernob studios - feels more like a condo building with only three units per floor.
I actually like the location - what I like to consider a workforce housing location in San Francisco where they aren’t building much new supply and where prices will remain reasonable for the huge swath of people who happen to not work in tech. There are parts of Lower Nob Hill which are really the Tenderloin, and this is not one of them.
And what really got me excited was the basis - at around $300 per foot, I applied my back of the napkin that you need to sell a building for twice what you pay for it, and $600 per foot was what buildings were selling for here in 2018-2019 - so not out of the realm.
Nice per door number too, but that’s skewed down due to the high studio count.
The going in 5-cap was a bit of a turn off, but on basis alone underwriting the deal seemed justified.
Location
As I mentioned, this is much more Nob than Tender, with the added benefit of being across the street from St. Francis Hospital and an easy walk downtown.
The location is not as attractive as it used to be when people actually worked in downtown San Francisco, but the bet here is that rents have been beaten down and as downtown eventually gets its act together, the benefit of being able to walk there should return and rents will rise accordingly.
If you believe in San Franciscos’s rebound story, this is a bet you should be willing to make.
The thing to understand about the location is that it is fine - nothing more, nothing less.
Physical Plant
The building itself felt solid and didn’t lean in any discernible direction - which in San Francisco is not an insignificant finding.
Set back from the street, the entry and lobby had more of a condo feel, albeit from decades past. Nice privacy for the tenants, since each floor has only three units.
Overall the building felt a bit tired, but taking a look at the tenants coming in and out you could tell that the tenant base is mostly professionals and tenants were proud to live here. Hard to quantify that, but it’s incredibly important - especially in a neighborhood where buildings could go either way.
At first the structure seemed fine - solid concrete, no tilts or leans, minimal evidence of deferred maintenance. Although upon further investigation, the structure would prove to be the key element of why the deal was in fact too good to be true.
Rent Roll
Here is where things started to get interesting.
In a rent control city like San Francisco, value add is most often achieved by chewing through loss to lease, which in this case was almost 30% using the broker's rent estimate. (Loss to lease is the difference between market rent and what’s actually being collected).
Now, this upside isn’t guaranteed because if all your loss to lease is trapped in units where tenants have been there forever and are not likely to leave, it’s what I call "phantom value," which you can't reasonably expect to realize. But brokers are happy to sell you.
In this case, more than 60% of the leases were signed after 2020. And since the owner was clearly managing for occupancy and not pushing rents, that was a good thing.
Right now in San Francisco, rents in most neighborhoods are below their pre-Covid peaks but off their 2021 lows, so leases signed in the past few years have upside coupled with a higher likelihood of moving.
Here, it means that we could actually build into the model turning a good chunk of the loss to lease into additional NOI.
You also have seven of the 35 units with long-term tenants paying well below market, so if any of them happened to move that would be a nice pop to value that would not be included in the model.
Operating Expenses
Another feather in the deal's cap was the above-market expense ratio at 43%, which was unnecessarily dragging down the going in cap rate.
For buildings with 16 and more units in California you’re required to have an onsite manager, and operating expenses at a building like this should run between 30-35% of rents. So at 43%, so there were almost certainly cost savings to be had which would be a straight benefit to the bottom line.
The two most obvious expense line items were on site management (overpaid by a factor of 3) and insurance (around 1.5x our initial indication). Right-sizing these two line items would add around 40bps to going in cap, already making the deal look much better before units even began to turn.
Value Add
Units were already in pretty nice shape, open boxes with views, natural light and balconies.
Kitchens and baths could use a bit of love, but we didn’t see the upside in going all out on the renovation. People just don’t pay up for studios unless you’re talking new construction - and right even now those are struggling.
We budgeted $25,000 per unit turn, which in San Francisco is plenty for cosmetic upgrades focused on low cost, high ROI items like paint, lighting and counter upgrades (for example).
After our first model run, the deal was working within a hair of the asking price and we started to wonder why, given that the deal was marketed by the biggest brokers in town and every buyer who was awake would have seen the deal.
In talking through the deal with our preferred contractor after he joined us for a second tour, he asked the $60,000 question: has the concrete been tested?
Lightbulb
One of the reasons this business is so hard is that real estate investors have to become experts on a tremendous variety of topics, from macro economics to local politics to construction.
And in this case, it was a combination of the second two.
During an earthquake, buildings shift and flex in response to the shaking depending on their material makeups.
Wood frame homes dance along with the seismic waves and maybe some pictures fall off the wall, while concrete skyscrapers sway, ideally without falling over.
The nature of that response depends on a material’s ductility, which is the measure of a material's stiffness, or more specifically its ability to bend without breaking. Wood for example has high ductility, whereas concrete’s ductility is low.
But some concrete, depending on how it was originally made and/or how its withstood the passage of time, is categorized as "non-ductile" and viewed as a higher risk of collapse during a major seismic event.
Building collapse tragedies around the world are, in part, due to the prevalence of non-ductile concrete (plus lax building codes plus lax building practices plus a bunch of other stuff which mostly doesn’t apply in the US).
In 2015, the city of Los Angeles exacted an ordinance requiring owners to retrofit such “non-ductile concrete” buildings over a 25-year period, an ambitious goal given how costly these retrofit projects can be for existing, tenant occupied buildings.
So after a list of potentially non-ductile concrete buildings in San Francisco was leaked in 2023, the apartment owner community has been bracing for a similar requirement.
But nobody knows what the potential requirement could be, when it would take effect or even whether city will act at all.
It’s a risk that’s nearly impossible to price at this juncture.
And since retrofits can run into 7-figures, tall concrete buildings built in the 1960s like 1257 Bush are also almost impossible to price.
Is a $500,000 reserve enough? $1,000,000? $2,000,000? Should we reserve at all? Its impossible to know.
So after much deliberation, the deal was a pass.
Which was disappointing because the timing, location and setup were right in our wheelhouse.
As it turned out, six offers came in and the seller went into contract just below asking.
But after an elevator issue was discovered during due diligence, along with the aforementioned structural concerns which the buyer likely came to better understand the further into the deal they got, the property traded all the way down to $5,900,000.
That’s just $250 per foot, or a mere $170,000 per door.
Either a tremendous buy, or your classic value trap.