One loan wearing seven addresses: the Macklowe towers were due all at once

The usual telling — including the reel this brief grew from — is that Harry Macklowe bought seven towers with almost none of his own money, the debt was one big cross-collateralized bet, and when credit froze the default took everything at once. The spine of that is right, and it is worth being precise about, because every part the retelling compresses is more instructive than the compression.

The record disagrees with the retelling about the number of buildings. It disagrees about what happened to the GM Building. It contains no bankruptcy, no auction, and a sixteen-month unwind in which the lender ended up financing the final buyer itself. And it has a hole where the bottom line should be: no one has ever published what the lending group got back.

Ten days in February

By the start of 2007 Harry Macklowe was seventy years old and had already been to the edge twice. In 1985 he ordered four buildings on West 44th Street demolished in the middle of the night, without permits, two days before a moratorium on demolishing single-room-occupancy housing took effect; his company and one of its officers pleaded guilty, $2 million went to the city, and Macklowe himself escaped indictment because the state could not establish criminal intent.^24,25 In the early-1990s crash his bankers called in about $1 billion of loans and took back five buildings, including the hotel that carried his name.¹ What rebuilt his standing was the General Motors Building, bought from the bankrupt insurer Conseco in September 2003 for $1.4 billion — a price his peers thought excessive until the sunken Apple cube he had sketched before he even owned the plaza made the corner famous.^1,17

Late in 2006 Sam Zell decided the market had peaked and sold Equity Office Properties — the largest office landlord in the country — to Blackstone, which raised its bid to $55.50 a share against a rival, a transaction valued at about $39 billion, closing on 9 February 2007.^1,4 Blackstone’s Jonathan Gray had let it be known before the deal even closed that the Manhattan buildings would be sold on.¹ Macklowe took them the same day the merger closed — the purchase was done concurrently with the EOP closing, not after it — for $6.8 billion, having negotiated the whole thing in about ten days.^1,2,7 “We thought capturing a dominant position in the Plaza district was a real plus,” he said, still confident, in January 2008.¹

The financing is the study. Macklowe put in about $50 million of his own cash. Deutsche Bank lent $5.8 billion against the towers. The remaining $1.2 billion had to be raised in about two weeks, and the only lender prepared to move at that speed was Fortress, at a rate reported north of 15 per cent, with personal guarantees from Macklowe himself.^1,23 All of it was due in about a year — “$7 billion in loans, due in February,” as Fortune put it while the February in question was arriving.¹ The reel’s cross-collateralization claim holds in the record at the level that mattered: the towers secured pooled debt rather than seven separate mortgages. A mezzanine lender’s own SEC-filed materials describe four of them as collateral for a single financing pool, and two more were held — and later sold — through one owning entity.^11,21 One deal, wearing seven addresses.

Notice what is missing from that structure: time. A 99 per cent loan-to-cost position is survivable if the debt is long. Seven Midtown towers were never going to stop being worth something close to $7 billion by 2027. The bet Macklowe actually made was not on Manhattan offices; it was that the refinancing window would still be open in twelve months. His son Billy said exactly that, jaw clenched, in February 2008: “We always maintained it was interim acquisition financing. We had a plan to provide for a long-term capital solution. The events in August have slowed the process down.”¹

The wire said eight buildings. Every retelling says seven.

On the day the deal closed, the Associated Press reported that Macklowe had bought eight buildings for $7 billion, about 6.5 million square feet, and the list that circulated named 717 Fifth Avenue, Worldwide Plaza, 1301 Avenue of the Americas, Park Avenue Tower, 527 Madison Avenue, 1540 Broadway, 850 Third Avenue and Tower 56.³ Within eighteen months the portfolio was universally “the seven towers” — in Fortune’s deep report, in the Observer’s accounting of the unsold, in Reuters’s description of Worldwide Plaza as the last of seven.^1,6,7

No published source explains the difference. What the record does show is that the disposition history — the sale, tower by tower, of everything Deutsche Bank took back — accounts for exactly seven addresses, and 717 Fifth Avenue is not one of them: it appears in no account of the unwind, and Blackstone itself was still selling an interest at that address years later. The likeliest reading is that the eighth building never actually travelled to Macklowe. But that is a reading; the count in the day-one record and the count in every retrospective simply disagree, and a reader who has seen this deal cited a hundred times has almost certainly been given the number seven with no mention that the original wire said otherwise.

The dollar figures swim the same way. Fortune says he paid $6.8 billion and borrowed $7 billion; the wire said the price was $7 billion; the New York Times, retelling it in 2013, put the price at $7 billion and the debt at $6.8 billion — the same two numbers, assigned to opposite roles.^1,2,3 The $50 million of equity, at least, is consistent in every source that states a figure at all. One cut of the reel this study grew from says $15 million; no source we could find says that anywhere, and the $50 million cut is the one linked below.

825 Eighth Avenue, May 2008

The default was not an event. It was sixteen months of selling.

Worldwide Plaza, photographed in May 2008 — three months after the default, fourteen months before it became the last of the seven towers to sell. Macklowe had paid $1.74 billion for it; Deutsche Bank let it go at $600 million, and financed the buyer itself.Photo: Jim.henderson, public domain, via Wikimedia Commons

Six months of music, then no chair

In August 2007 the securitization market that every one-year bridge loan in America was counting on stopped. “There was a game of musical chairs,” said Ben Lambert, the Eastdil Secured chairman who had sold Macklowe the buildings on Blackstone’s behalf. “The music stopped, and there was no chair for Harry.”¹ By January 2008 Deutsche Bank was talking to the Macklowes about taking the buildings back — complicated by the fact that it had sold pieces of its $5.8 billion to as many as twenty other lenders, some of whom stood first in line and some of whom would get, in Fortune’s phrase, leftovers.¹

What happened in February 2008 is usually told as a seizure, and the word the contemporaneous record uses again and again is quieter: consensual. Macklowe defaulted and handed the towers back to his senior lender rather than filing for bankruptcy or forcing a foreclosure auction.^9,22 A mezzanine lender in the stack, Capital Trust, told its investors that May that its Macklowe position had “come due in February but was subsequently extended by the lending group to February 2009,” and that the properties were being offered for sale “pursuant to a consensual arrangement with the borrower, with the proceeds going to the lenders to repay the debt.”¹⁰ The default was real and total. It just did not look like the movie version. It looked like a workout — sixteen months of brokers, extensions, re-trades and closings, run by the same people who had signed the loans.

The sales themselves tell the story of 2008 in seven receipts. In June, before Lehman, Shorenstein paid $930 million for the pair that travelled together — Park Avenue Tower and 850 Third Avenue, held and sold through a single owning entity.^5,21 In August, Paramount and Allianz took 1301 Avenue of the Americas at about $1.5 billion, slightly above Macklowe’s $1.46 billion — the only line in the ledger that went up.^5,19 Then the market turned. A $2 billion bundle of Worldwide Plaza and 1540 Broadway fell apart; Transwestern re-traded and then lost Tower 56; and the prices on everything that closed after September carry the crash in them: 527 Madison at $225 million, Tower 56 to Pearlmark at $158 million, 1540 Broadway at about $355 million — a third of what Macklowe had paid — and finally, in July 2009, Worldwide Plaza at $600 million against his $1.74 billion.^5,6,8,9,20

That last sale is the detail the retellings never carry. Deutsche Bank did not just accept $600 million for a tower that had stood at $1.74 billion on the 2007 closing statement — it provided the buyer’s $470 million mortgage itself, put up roughly $135 million of the equity beside them, and wrote down the $1 billion mortgage it had held.⁸ A lender financing the purchase of its own collateral, at a third of the price it lent against, is what the end of a bubble actually looks like from inside the bank.

The building he sold to keep the promise

The seven towers went back to the bank. The GM Building did not — and the distinction is the part of this story the one-line versions flatten. Fortress’s $1.2 billion was personally guaranteed, and the building at 767 Fifth Avenue was pledged behind that guarantee; by February 2008 the fund was, in Fortune’s words, salivating at the possibility of foreclosing, and had begun pricing Macklowe’s art collection, his homes and his yacht in case the building was not enough.¹ Fortress extended its deadline to 15 February 2008 precisely so that Macklowe could sell the building himself.¹

Nor was the GM Building the debt-free trophy of the retelling. Macklowe’s own investment summary shows $1.2 billion of debt on the 2003 purchase, a German partner’s $300 million of preferred equity until December 2006, and a $1.9 billion recapitalization — two months before the towers deal — that made him sole owner of a building mortgaged for more than it had cost him.¹⁸ “Owned outright since 2003” is not what the owner’s own document says.

The sale, when it came, was enormous and fast. Agreements signed on 23 May 2008 sold the GM Building and three other buildings from Macklowe’s own pre-EOP portfolio — 540 Madison Avenue, Two Grand Central Tower and 125 West 55th Street — for an aggregate $3.949 billion.¹³ The GM Building closed on 9 June at approximately $2.8 billion by Boston Properties’ SEC filing — the seller’s own summary says approximately $2.9 billion, and this study reports both — through a venture in which Boston Properties holds 60 per cent, beside a Goldman Sachs-managed fund whose principal investors were the governments of Kuwait and Qatar, and Dubai’s Meraas Capital at 20 per cent.^14,15,16,18 Mortimer Zuckerman called it a commitment to the best real estate performing “even better in more challenging times”.¹⁵ The proceeds repaid Fortress; Macklowe kept a non-controlling interest in the building he had spent forty years wanting, and later testified that the forced sale left him with roughly $900 million of taxable gains and, eventually, hundreds of millions underwater on the tax.^18,22,26

Three weeks after the GM Building closed, Billy Macklowe ousted his father as head of the firm and spent the next year shepherding the tower sales he had opposed financing in the first place.⁹

The lender’s own filings never say his name

Here is a silence worth reporting. Fortress was a public company from February 2007, and the $1.2 billion Macklowe bridge loan was among the most discussed credit positions in America in the winter of 2008 — yet a full-text search of Fortress Investment Group’s SEC filings returns not one mention of Macklowe, or of the General Motors Building, anywhere. The loan lived in managed fund vehicles that never had to name their borrower. The most consequential lending relationship of the cycle, from the lender’s side, is publicly documented only in journalism.

Which is why the one lender that did have to write things down matters more than its $50 million suggests. Capital Trust’s earnings materials are SEC exhibits, and they show the unwind from inside the lending group in real time: the February maturity, the extension to February 2009, the consensual sale process — and then, in July 2008, the decision to reserve its entire economic exposure while bids came in low because “buyers sensed that the lending group was a forced seller.”^10,11,12 A junior lender announcing to its own shareholders that it expects nothing back is the closest thing the record holds to a statement of what this deal cost the people who financed it.

Macklowe after

The Drake Hotel site on Park Avenue — bought in 2006, demolished, then caught in the same credit collapse — drew a Deutsche Bank foreclosure suit over $482 million in August 2008.²⁷ It became the one loss that circled back: CIM Group bought the site’s discounted debt in the workout years and partnered with Macklowe to build 432 Park Avenue on it, the 1,396-foot pencil tower that opened as the tallest residential building in the hemisphere.² He never regained the fortune — Forbes had him at $2 billion in 2007 and he left the Forbes 400 in 2008 for good²⁸ — and his own accounting of the arc, delivered to a room of Cornell students in 2014, is the cleanest summary anyone has produced: “I have been extremely fortunate, 1) to be successful, 2) to make money, 3) to lose money, 4) to make money, 5) to lose it again, 6) to make it back.”²⁹

The retrospective verdict that matters most came from the man who sold him the towers. Blackstone’s Jonathan Gray, as quoted in Vicky Ward’s The Liar’s Ball: “His mistake was not the price he paid … it was the way he financed it.”²² Seven Midtown towers in 2007 were a defensible thing to want. Seven Midtown towers on twelve months of money were a clock.

What transfers

The first thing is the reel’s own lesson, which the record strengthens rather than softens: leverage does not decide how much you make; it decides who can force you to sell, and when. Macklowe’s buildings did not fail — their tenancies, their addresses and their long-run values were fine, and one of the seven even resold above his price seventeen months after he bought it. What failed was a maturity date. The instrument to study in any levered deal is not the price or even the rate; it is the calendar, and who holds the pen when it runs out.

The second is that a default has a mechanism, and the mechanism is most of the outcome. Macklowe filed nothing. He handed the towers back consensually, sold the GM Building himself before his lender could seize it, and kept enough standing — and enough residual interests — to be building the hemisphere’s tallest apartment tower on a lost site within five years. The borrowers who fought their lenders through bankruptcy that cycle mostly kept neither the assets nor the relationships. “Lost everything” is the retelling; the record shows a man who chose which everything to lose, in which order, and it mattered.

The third is about what the record cannot show you. This is one of the most cited deals of the century, and its bottom line does not exist in public: no total of what the lending group recovered, no accounting of Macklowe’s full loss, not one mention of the borrower’s name in the public lender’s own filings. A workout that never touches a courtroom never has to add itself up. When a deal this famous is quoted at you with confident totals, it is worth knowing that the confident totals are reconstructions — including the $3.77 billion this study summed, which is offered as arithmetic on seven sourced receipts, not as a number anyone published.

And the count is a parable in miniature. Eight buildings in the wire that day; seven in every account since; no document reconciling them. A story told often enough acquires a shape, and the shape sheds whatever does not fit. The retelling is how the deal is remembered. The record is what it was.