Requiem for a dead mall
A personal note on the past few years
This summer, I spent a few days backpacking in Olympic National Park. Driving back, I found myself stuck in traffic on I-5. As the traffic slowed, I remembered I once resolved to never change lanes in such a situation. A earnest believer in general equilibrium, I reasoned that if there’s speed to be gained by weaving, others would also do so, congesting that lane further. Nothing to be gained for your trouble.
For the past few years, I’ve worked as an analyst at a hedge fund. In truth, when I started I wasn’t sure if hedge funds made conceptual sense. Didn’t the market reflect all available information? Sure, there were pennies to be picked up by providing liquidity, but hadn’t everyone competed away all of those years ago? The very idea of excess uncorrelated returns is a bit of a paradox. What incentive is there to trade if the mere existence of an offer is evidence they know something you don't? Yet from the trading floor to the highway, there is endless jockeying for position. Was it all irrational?
I started out focusing on machine learning, but soon moved over to Commercial Real Estate (CRE for short). By now the story of the deep distress in suburban malls and vacant San Francisco offices that dominated financial headlines last March has been told a thousand times, but for me it was just beginning. Reading through the documents for each loan, I first began to see how every one had a unique fingerprint, a bespoke mess of covenants and clauses ready to ensnarl each property if things started to go awry. Then there were the endless chat messages with bankers, often talking their own books but still invaluable. Even in the relatively liquid securitized market of Commercial Mortgage Backed Securities, or CMBS, nearly all trading is done over those instant messages. In college, I had imagined finance as largely algorithmic. Calling someone on the phone to buy anything felt like a bygone era. Soon, I realized not only were phone calls normal but critical, especially when you could count on one hand the number of participants involved in struggling offices and malls. This was hardly some sleepy corner of the market that time forgot – while the CMBS market only makes up 15% of the total Commercial Real Estate market, that’s still at least $636 billion. Accordingly, as malls and offices kept moving into that “distressed” bucket, the small number of parties became increasingly apparent.
A few more basic characteristics of the CMBS market help to understand the kind of breakdown that happened over 2023. First, records of completed trades are shared through dealers to their own networks, often louder when it benefits someone to do so and only in a general “price context”. There is no requirement to publicize any trade. Second, while some share of inventory is sold through publicly advertised first-price auctions, these can contain secret reserve prices (which can be decided ex-post), are often multi-round with shifting rules and generally heavily favor the seller. Finally, a significant share of sales do not receive bids at all: instead, they are facilitated directly through a dealer connecting two parties. Dealers make money from fees, so they are simply incentivized to get any trade done, rather than get the best for their client. As a repeat buyer, you would often be favored more than the seller and could work with the dealer to give information and help “talk them down” on the price.
All of these points lend themselves to rather colorful characters taking the role of small broker-dealers. It doesn’t help that regional banks do an outsized share of CRE lending and that their trading desks seem to have an outsized involvement in the secondary market as well. I had countless questionable conversations with people that seemingly couldn’t manage to string together a coherent sentence, let alone represent any reasonable facts. (Looking back, misspellings may have been on purpose: allowing plausible deniability if the SEC ever read our messages.) Each day felt more like being on a pirate ship.
Beyond the liquidation lists, the strange and conflicting incentives of bondholders seemed to have an ever-increasing impact on the real-world fate of these empty offices and malls. Derivatives priced on top of these buildings gave incentives to buy and sell at non-economic prices, and with few real transactions happening it was impossible to tell what real prices were and how bad this all really was. [1] How long the banks and insurance companies holding these loans could hold their breath while remaining underwater was anyone’s guess.
I don’t want to linger on the financial dimension of all of this, there is far better commentary you can read to understand that. What was most striking was how in the thick of things how little large investors seemed to understand about what was really going on. A loan officer for a large Asian bank once told me: “I ordered a valuation for our Boston office, and it went down from $250 million to $125 million. I didn’t even know it could decrease that much. Now we need to fly to Seoul and explain this to them!”. That was a good scenario. In the height of the zero-rate mania of 2021 and 2022, it wasn't uncommon to see brand new retail complexes in cities like Austin and Durham get three or even four layers of securitized debt, "mezzanine loans" on "B-pieces" on first mortgages on ground leases ... no wonder why Barry Sternlicht of Starwood Capital predicted in May 2024 a regional bank failure “every day or every week”.
That isn’t to say I understood what was going on either. One of my strongest memories was hearing a bond traded at a level that seemed way off to me, only to hear news the next day that made the sale that looked cheap suddenly seem extremely rich. I learned to think of an offer alone as information. Strange sales became evidence I was missing a key piece of the puzzle rather than errors on someone else’s part. Bit by bit, the landscape of “unknown unknowns” grew larger, and I came to understand I didn’t know even more than I thought I didn’t know.
It’s probably not surprising that money managers are often incompetent; the story of AAA bonds being sold to insurance companies as safe investments before eventually taking losses sounds more like 2008 redux than novel insight. But how did we get here again? Over a decade after over-indebtedness and leverage brought down the residential mortgage market, once again the same forces come to bear on suburban offices and strip malls. First as tragedy, then as farce. When a insurance ratings officer says they see "through the cycle" when assigning ratings to commercial buildings (meaning, they underwrite recovery and ignore the decline in collateral value), it is hard not to think of the blind ratings agencies depicted in The Big Short. There is a bitter irony in the Commercial Mortgage derivative index (CMBX) being the only one after 2008 to avoid becoming centrally cleared, thus keeping the same counterparty risk and opacity concerns alive on a miniature scale.
What is most frustrating about all this is not the blind search for yield that fueled excessive leverage or the dealers that roped insurance companies and mutual funds into this whole mess but the fact that very little will come out of the so-called “CRE reckoning”. For all the hand-wringing around CRE CLO funds and the incessant short seller reports on Arbor or Blackstone, not a single one has failed or even seems close to collapse. Last year, the vast majority (over 65%) of CMBS notes coming due were simply extended to 2024 or 2025. This year’s share of extensions is even higher. Even if all publicly traded debt collapses, it simply isn't large enough to force banks to take on the pain of revaluing the vast majority of CRE lending, which continues to sit in hold-to-maturity buckets. “Extend and pretend” it hardly a strategy, but it seems to have won.
Much of this comes down not to the scale of the error but how it impacts people, which differs greatly from 2008. Back then, images of ordinary people getting foreclosed on were so salient they spurred the United States into passing some of the largest financial regulations in its history. Landlords slowly losing their office buildings doesn't elicit the same sympathy. Despite that, the consequences for American downtowns and cities are real, setting aside the impacts to pensions and retirement funds, which are far more subtle. If lenders choose not to walk away entirely (e.g. see 350 California St., which Mitsubishi UFJ sold at an 80% discount), they are faced with the difficult prospect of managing a building that is no longer generating revenue alongside a sponsor that may no longer see the “light at the end of the tunnel” and is unwilling to help contribute finances necessary for new leasing.
In either case, cities are left with buildings which no longer “work” as financial assets yet can not receive new investment without realizing losses that would be simply far too painful. For banks, that might be a regulatory or legal constraint, for other funds, it could just be wishful thinking. The result is the same: zombie buildings proliferate. Like their horror movie counterparts, these zombies are also highly contagious. A lack of sales allows firms to continue to reference past prices as the current market rate, fueling denial. All the while, these vacant buildings suck dry the property tax budgets of cities and contribute to “doom loops” that draw downtowns further into crisis. From Los Angeles to D.C., the specter of urban decay looms large once again.
It was only March, but it was already hot and sunny in Downtown Los Angeles. I was glad to step off the street and into the cool and quiet lobby of the office complex I was touring. 60% occupied, but just 40% physical occupancy once you took into consideration tenants with work from home policies not using their space. That didn't seem to bother my tour guide, who excitedly talked to me about the renovations to the elevator and new “community spaces” equipped with team-bonding materials and games. They looked nice, I agreed. You could tell it was a former storage space, but the softer lighting and glass walls did give it a more comforting feel.
“January 6th, 2023, 9:00 AM.” The building manager knew the time down to the minute. “That was when the foreclosure against (Fund A, original owner) was completed and (Fund B, mezzanine lender) took over.” I asked if anything changed. “Not much, at first. They got to keep all the new improvements (Fund A) had just put in, so they set to work advertising those.” I didn't have the heart to tell her that the second fund, too, was on the brink of foreclosure by the bank, and that that was why I was there. As we went through one of the newly renovated floors, we ran into one of the executives at the old owner’s fund. The building manager warmly greeted and introduced him, now one of the few tenants left. Having to stay and work out of the building you just lost must really sting. I was glad to not have to introduce myself.
We stopped again for a moment in the lobby, a vast open space made tranquil by the large flowing water installation that separated the elevators from the main entryway. I thanked my guide for her time and sat by the water for a while, taking notes and collecting my thoughts. It was a weekday, around lunchtime, but no one walked through the doors. Above the entrance, a large news ticker listed out the day's stock market changes for no one. A lone security guard kept watch over the gleaming renovated elevator banks. It was hard to imagine this space as the same as the small line item tucked neatly into the spreadsheets I had looked at for so long before.
Nearly three thousand miles away, slices of its mortgage were changing hands once again. Soon a new manager, a fire sale, or refreshed appraisals could arrive. Even sitting right there in the lobby, though, I wasn’t sure what any would reveal, let alone what would happen next for the building or the many others just like it. I had been looking through a small gap in the fence; having caught a glimpse over the wall, I realized just how little I had been able to see. I checked my watch. There were four more buildings to tour today; I couldn't stay too long. Collecting myself, I stepped back out into the hot Los Angeles sun.
[1] A brief footnote explaining these conflicting incentives:
• AAA holders would like prepayments but lack the legal rights to control the receiver. They can hold “hostage” excess cash flow by denying new leases in an attempt to force a sale (Gas Company Tower in Los Angeles), but that risks further driving the valuation lower such that even the AAA class may take losses (1740 Broadway in NYC).
• BB/BBB- holders control the special servicer, and can extract fees during the re-tenanting process, so would like to keep the “zombie” alive as long as possible – so long as they do not trigger new appraisals that put them out of the money. Even a few months of delay are valuable, since unpaid interest prior to new appraisals is typically senior to principal, even in a liquidation (this is a potentially unintended consequence of post-GFC “2.0” CMBS deals).
• Finally, Mezzanine holders can delay or disrupt the re-appraisal process by demanding new appraisals, negotiating modifications that allow for them to take current leasing revenues before eventual sales, or simply extending the servicing process to take yet more fees (e.g. CXP Portfolio).