Office is dead. Long live office?

A rant first. I detest this long-established trend where the branding consultants march across the economy removing anything unique and interesting from the names of businesses. From “Kentucky Fried Chicken” becoming KFC to my local grocery changing from “Glen’s” to “Family Fare” (ugh!!), this trend has now destroyed the uniqueness of “Boston Properties” and made it “BXP REIT” (BXP).

Thanks for listening to that rant. A pox on all branding consultants.

Four years ago, as the US was emerging from the pandemic, predictions of the demise of office were a dime a dozen. All work would be from home and central business districts would soon be wastelands.

And indeed, there have been large impacts. Nationally averaged property prices for office remain down 37% from their peak in 2022. Nationally, office construction has nearly shut down.

So an average office building financed with a mortgage at 70% Loan to Value (LtV) is still under water. There are real losses, whether yet realized or not.

Even so, the widespread predictions of the complete collapse of the office market, and sometimes also of the US banking system, have not come to pass. Yes, there is a lot of “extend and pretend” going on, but with each month the odds of a cataclysm go down.

From this national perspective one would expect that nothing good is happening or will happen soon in the office markets. And one would be wrong.

Variegated Markets

The old saying, that real estate is about “location, location, and location,” is certainly true. It is also about property attributes.

What we’ve been seeing now for years is a bifurcation in the office market. Here is how BXP described it in their Q4 2024 earnings call:

BXP competes primarily in the premier workplace segment of the office sector, which continues to materially outperform the broader office market. Premier workplaces are defined in CBRE's research as the highest quality 7% of buildings, representing 13% of total space in our 5 CBD markets. Direct vacancy for Premier workplaces is currently 13.2% versus 18.8% for the broader market.

Likewise, net absorption for premier workplaces has been a positive 8.8 million square feet over the last 3 years versus a negative 15.6 million square feet for the broader market. Asking rents for premier workplaces are more than 50% higher than the broader market, up from approximately a 40% premium 3 years ago.

Premier workplaces are located near transit, are architecturally appealing, and typically offer at least two “amenities” such as health clubs and restaurants.

So with the slow return to office and ongoing economic growth, vacancy in premier workplaces will continue to drop. Rents have already surged as vacancy has dropped in Manhattan. And even San Francisco, the weak sister for several years, showed positive net absorption in Q4.

Here is a quote from one random example. I see similar stories frequently.

Blackstone and DivcoWest bought the 416,000-square-foot office tower [at 199 Fremont] for $111.3 million, or about $265 per square foot, earlier this month, the San Francisco Business Times first reported. That’s a more than 70 percent discount from the building’s price just over five years ago.

The private equity company and DivcoWest plan to add a full floor of tenant amenities as part of the renovations. And they’ll rebrand the property as 300 Howard, according to an announcement about the project last Tuesday.

The Lower-Class Crunch

Premier workplaces are a fraction of Class A buildings. For several years, a friend of mine has opined that upgrades or conversions of lower-quality buildings will suppress rents in premier spaces and assure that the woes of office will continue. And certainly we have seen some such conversions.

But for Class B and Class C it often would take a rebuild to generate a premier space, and those rarely “pencil” at the moment. The photo above shows one that did, about which more below.

On top of that, there has recently been a huge surprise. We are now seeing supply shortages for Class B properties in both Manhattan and Boston. Quoting CRE Daily:

A Market Flip No One Predicted

According to the Commercial Observer, only a few years ago, Class B buildings were written off as the inevitable casualties of a remote-first world. Now, they’re the unlikely stars of a recovering office market.

Mid-tier properties — typically older and less flashy than their Class A peers — are experiencing a dramatic turnaround. Executives at Adams & Company, which oversees 36 such buildings, report several properties are fully leased, with more demand than space available.

What is going on is that residential conversions have turned out to be viable for many of those buildings. A lot of them are underway (for example 2.5% of the current office space in downtown Boston), which helps address the longstanding shortfalls of housing while also improving the office market.

Demolish and Rebuild

We are even seeing circumstances where it makes economic sense to demolish and rebuild. This includes that (future) building shown in the image above, scheduled to deliver in early 2029.

We learned on the Q4 2024 earnings call that:

we acquired 725 12th Street, a vacant office building from its lender for $34 million or $112 a square foot. The site is very well located, immediately adjacent to Metro Center, Washington, D.C.'s busiest transit stop where 4 train lines converge.

And more recently bisnow tells us this:

Law firm Cooley LLP has signed a big deal at a planned downtown D.C. office development as new trophy space continues to dominate the leasing market.

The Palo Alto, California-based firm signed a 20-year lease for 126K SF at 725 12th St. NW, where BXP plans to demolish a 33-year-old building and develop a new 320K SF office, the developer announced Thursday.

The latest lease brings the project to 87% preleased before it has even started construction, a sign of the high demand for increasingly scarce large blocks of trophy office space in D.C.

The trophy segment of D.C.'s office market had a vacancy rate of 12.2% at the end of the first quarter, according to CBRE, much lower than 22.6% for the overall market. Only one new trophy building is under construction today: Stonebridge and Rockefeller Group's redevelopment of the former WMATA headquarters that also landed a law firm prelease.

The tightening of the trophy market has pushed rents up: CBRE reported that trophy asking rents last quarter averaged $91.21 per SF, up from $86.74 per SF a year earlier.

All the above suggest good times ahead for BXP. And this in turn raises the question of whether it is a buy, and for whom. Let’s look.

[My conclusion, which I did not know when I wrote that sentence, is “nobody.” But instead of reworking this article as a hit piece, I decided to share my exploration.

If you just want the bottom line, skip to the end or stop now. If you would like a chance to learn something about how such a REIT works, keep reading.]

Overview of BXP

This slide summarizes a lot of key elements about this REIT:

BXP is big. They have long leases and so have modest annual lease expirations.

What drives BXP is development. They are active developers, with current developments (and redevelopments) at the level of about 5% of their current square feet.

While they do operate a collection of properties, they look to regularly sell properties, or partial interests, to fund further development. This ends up making some trends non-intuitive. Consider occupancy:

Occupancy has plunged nearly 600 bps since 2019. For many REITs that would lead to distressed property sales and dividend cuts. But as you can see, total revenues increased throughout, save for a 10% drop in 2020 (that’s a GAAP number, perhaps hiding a bigger drop, but no matter what revenues are up more than 10% since 2019).

The story is that BXP delivered enough occupied properties to offset the decline in overall occupancy. This has kept NOI/sh climbing, but increasing interest costs (on which more below) pulled down CfO/sh in 2024:

BXP reports a FAD/sh, but it does not represent actual cash available because they do not subtract capitalized interest. This graphic shows actual cash available as FAD/sh.

In addition, BXP finds FAD by adjusting Funds From Operations (FFO). Below, the section based cash flows shows results for FAD starting with Adjusted CfO/sh. The two numbers differ over time by about ± 10%. I did not try to reconcile the differences.

FAD is running about 43% of NOI. It is that small because BXP has a high level of sustaining capex, to support tenant replacement. Still, FAD/sh has been climbing nicely, increasing at a CAGR above 4% in recent years.

Whenever the office market recovers and occupancy moves up, and after covering associated capex, we should see FAD/sh move up nicely. Even so, at minimum that is a few years away.

It is also informative to look at the previous plot with the dividend overlaid:

What happened for a decade prior to 2015 was that BXP decided to sell a lot of properties. This reduced earnings but shareholders benefited by receiving some very large special dividends (for tax reasons).

After 2010, save for dispositions in 2014 and 2015, BXP focused on growing per share earnings. Holding the dividend flat since 2020 let FAD/sh pull above the dividend, so that the payout ratio is now a healthy 70%. But BXP needs that headroom to allow for the capex that will come with occupancy increases and to be able to pay increased interest costs (see below).

Cash Flows

The cash flows provide more insight. Let’s start with the cash flows associated with ongoing operations:

This and the next cash-flow plot are in my standard format, with the left stacked bar for each year showing sources of cash and the right stacked bar showing uses of cash. Here the source is adjusted CfO.

You can see that about half of CfO goes to dividends (purple). Most of the rest goes to sustaining capex of one kind or another, with a bit more soaked up as capitalized interest.

The difference between the height of the two stacked bars for each year is the retained cash — the FAD less distributions. While small, this is not zero.

The remaining cash flows are involved in the main focus of BXP — building, buying, and selling office properties. Here they are:

Two of the large and persistent categories here are construction in progress (purple) and full or partial dispositions (blue). Central to BXP is that they raise money, mainly from those dispositions, and use it to support construction.

Once the new buildings have stabilized, substantial value has typically been created, because the build yields are well above the disposition yields. BXP then levers up that new value by adding debt (gray), which adds cash for further construction.

It has been many years since BXP has issued stock (red), and it is clear from their comments that this is not part of their thinking at the moment. Occasionally BXP does acquire property (yellow), very often for redevelopment.

They also accumulate or spend large amounts of cash (pink) to enable their construction. Some REITs use their credit revolver for this, but BXP seems to be more careful.

This is a viable business model, but not without some risk for shareholders. As usual, that is associated with debt.

Debt

With credit ratings of BBB from S&P and Fitch, BXP is not in the top group of REITs but also is not far from them. BXP has very steadily run LtV (Debt to Gross Assets) in the mid 50s. This is risky; here is why.

G&A costs push 10% of NOI. Sustaining capex runs near 30% and can spike higher. For 2024, interest expenses are above 30% of NOI and they are increasing (see below). And dividends are up near the remaining 30% of NOI.

So when NOI drops, what gives? Sustaining capex can be pulled down but perhaps not immediately. Otherwise those costs can’t move much.

One is left either cutting the dividend or making damaging choices (increasing debt, making forced asset sales, or diluting shareholders). Having a lower LtV would buy more headroom.

That said, the NOI growth engine of the construction in progress can help offset declines from loss of tenants, as we saw above. My view is that the BXP dividend is not fundamentally safe but also is unlikely to be cut outside of a quite severe recession.

The dividend was cut 25% across the Great Recession and there was a bit of dilution totalling just over 10%. Things today don’t seem that different financially for this company.

Debt Maturity Ladder

The Debt Maturity Ladder does have a fair bit going on through 2028. But categories matter. Here is a plot:

The term loan (red) is intended to be rolled. Plus BXP issued new debt last August that they have already used to pay off the 2025 Senior Note. I always appreciate such pro-active behavior.

Plus, mortgages (green) are never of very long maturity. That big green bar in 2027 is the $2.3 B mortgage on the GM Building on Fifth Ave. in NYC.

There are good disclosures about that building in the Supplemental. Marking the interest rate to market on a new mortgage will still leave the building with solid coverage of the interest costs. So my view is that the mortgage should roll easily.

That leaves Sr. Notes to replace: $1B in early 2026, $1B in late 2026, and $750M in late 2027. This is unlikely to be difficult, and if they need some time they have an empty $2B credit revolver.

The question is what it will cost.

Interest Costs

To address this, I took the debt details from the 8-K supplemental and let the interest expense increase to a current rate of 5.75% when each tranche was paid off at maturity. Here is what one gets:

The red bars show the GAAP interest expenses from the 10-Ks. These are distorted but I did not try to work out differences from reality.

The purple bars show the direct calculation of interest costs from the debt stack. We find that interest costs will increase by about $300M/yr over the interval from 2025 to 2035. This is an average increase of $42k/year.

The total increase will fully absorb the difference seen above between FAD and the dividend. Worse, the average is 4.7% of the current FAD.

Most REITs are seeing this cost come in near 1% of their FAD. BXP looks much worse. Why?

• Their 50% Debt Ratio

• They got really low rates on most of their existing debt

• It does not help that their credit rating has dropped from BBB+ to BBB.

BXP may well not grow FAD at a 4.7% rate and certainly will not do much more than that. So FAD/sh will be flat or declining for the next decade.

Forward Dividend Coverage

We saw in the earnings numbers that the current dividend is only 70% of FAD. There is ballpark $300M of headroom right now. But that headroom will disappear over the coming years as FAD likely drops and interest costs increase.

Whenever office recovers, BXP’s signed-not-occupied pipeline will go up. Then they will face substantial tenant improvement expenses in order to get those properties occupied and paying rent. This could easily pull down FAD/sh further for a time.

Sustaining capex has averaged $580M/yr since 2020, to re-lease an average of 6.6% of their space. This suggests that the incremental cost to grow occupancy, at 100 bps per year, will be in the $100M/yr ballpark. Combined with the increase in interest cost, coverage of the dividend by the current FAD would go to zero in 2031.

Eventually, as leased occupancy and interest costs level off, FAD/sh will grow. But that could easily be a decade from now.

Whether that happens sooner depends on how much FAD/sh actually drops. It would appear that the dividend is not safe for more than a few years. That said, any cut likely will not be extreme and will be recovered within a few years.

For Whom?

This is clearly a great time to be an opportunistic office investor. The next generation of Sam Zells is getting properties for dimes on the dollar, or less. But it seems that the REITs have too much history to overcome.

Considering the imminent recovery of premier office, I thought maybe BXP would have a chance to grow rapidly enough to be an upside play. After this analysis I don’t see that, although of course you never know about Mr. Market.

Failing that, the dividend yield got up near 7% in early April. I thought there might be a dividend where BXP would be worth buying for income.

But thanks to their likely increase in interest costs, combined with a need for incremental capex to lease up their vacancies, a dividend cut within a few years appears likely to me. So one might buy in a market crash if the yield got up near 10%, anticipating that dividend cut. I can’t quite see doing that at 7%, though.

That view of the future might change, if interest rates come back down strongly . No counting on that, though.

Absent that market crash, it seems to me that one might wait 5 if not 10 years before revisiting this one.

Please click that ♡ button. And please subscribe, restack, and share. Thanks!