Blackstone paid $39B for 543 office buildings — and sold 70% of them in five months

The trade underneath it is simpler than the size suggests. Equity Office was 543 buildings and 103.1 million square feet spread across sixteen states, and nobody wanted all of it.⁷ Institutions wanted the towers in Manhattan, Boston, Seattle, Washington and San Francisco, and they would pay full price for those individually. A company holding all 543 could only be bought at a price for all 543. Blackstone paid the price for the pallet and sold the good pieces separately, and the gap between those two ways of pricing the same real estate is where the money was.

What made it survivable was that the selling started before the owning did. What made it possible at all was a fee.

The seller had already named his price

Every purchase is somebody’s sale, and the proxy statement Equity Office filed for its own shareholders is unusually candid about how this one was reached. Approaches had been arriving since late 2005 and being turned away — $37, $40, $40.75 a share. Zell told his own board he would not support a sale below $45 “at a minimum”.²

Blackstone’s Jonathan Gray made his first approach in the middle of August 2006, at $40 to $42 a share, paired with a plan — later abandoned — to bring in another office REIT to take a third of the assets off his hands. He was told the price was not sufficient and submitted nothing in writing. In the last week of October he asked what a compelling number would look like, and Merrill Lynch, advising Equity Office, told him informally that it would have to exceed $45 and would probably have to be closer to $50.² He came back on 2 November at $47.50.

Two and a half weeks of negotiation moved that to $48.50, cut the termination fee from the $275 million Blackstone had proposed to $200 million, and raised the guarantee behind the deal from $500 million to $1.5 billion.² The agreement was signed on the evening of 19 November 2006 and valued the company at about $36 billion.¹

Zell’s own account, given at Wharton ten months later, was that this was a“Godfather offer” — a price no publicly held company could responsibly refuse.¹¹ He also said, in the same lecture, that private equity firms awash with capital were benefiting from “preposterous” leverage, and that “today, you would never be able to replicate the Blackstone deal”.¹¹

The bidding war, and the fee that ended it

On 17 January 2007 a consortium of Vornado Realty Trust, Starwood Capital and Walton Street Capital, operating as Dove Parent LLC, wrote to Zell offering $52.00 a share — sixty per cent cash and forty per cent Vornado stock, a premium they put at about $1.6 billion over the Blackstone agreement. Their letter named the money behind it: $13.4 billion of equity and $25.8 billion of debt, committed by Lehman Brothers, JPMorgan, Barclays, UBS and RBS Greenwich.³

Blackstone answered on 25 January at $54.00 a share, all cash, and the termination fee went from $200 million to $500 million.⁴ Vornado raised again, to $56 a share. On 1 February the Equity Office board unanimously reaffirmed the lower Blackstone bid.⁵

That decision is the whole deal in one move, and the board explained it in arithmetic rather than loyalty. A cash-and-stock offer is not worth its headline on the day it is made, because the shares arrive later and the deal takes months to close. Running Vornado’s $56 through that discount, the board’s own analysis put its present value at between $54.81 and $55.07 a share.⁶ Blackstone then went to $55.50 in cash — above the top of that range — and the termination fee went to $720 million.⁶

Vornado withdrew the next day, saying the premium it would have to pay was not in its own shareholders’ interest.⁸ More than 92 per cent of votes cast, about 71.3 per cent of all shares, approved the sale on 7 February.⁸ Asked later what had actually decided it, Zell did not talk about price at all:“The key to the deal was structuring a $720 million break-up fee.”¹¹

The ladder is the argument in five lines. The company sold for less per share than the number sitting beside it on the table, and every party involved could explain why in a sentence about time.

What 103 million square feet looks like from the pavement

The selling started before the owning did.

The Worldwide Plaza complex, Manhattan, photographed in September 2021 from Ninth Avenue — the residential buildings in front, the office tower behind.Photo: Epicgenius, CC BY-SA 4.0, via Wikimedia Commons

A fee that size is not compensation for wasted effort. It is a toll charged to anyone who wants to keep bidding, and at three per cent of the equity it was large enough that beating $55.50 meant beating $55.50 plus $720 million. Blackstone bought certainty of closing, and certainty of closing was the thing the whole strategy depended on — because the buyers on the other side had already been called.

What it took, and what came back

The $39 billion headline is enterprise value: what the equity cost plus the debt that came with it. The financing actually raised to do the deal is a smaller and more exact number, and it appeared in an SEC filing the day before the vote. Blackstone contributed $3.75 billion of equity against $31.9 billion of debt financing.⁸ Its own cheque was under ten per cent of the price.

The buyer who took Manhattan, and lost it

The Macklowe trade is the part of this story that gets left out, and it is the part that shows what Blackstone actually sold: not just buildings, but its own timing risk.

Harry Macklowe bought seven Manhattan skyscrapers for $6.8 billion — the figure Fortune gives; Reuters reported about $7 billion at the time — and put up $50 million of his own cash against $7 billion of loans from Deutsche Bank and Fortress Investments, due in February 2008.^8,12 When the subprime crisis unfolded there was nothing to refinance into. He had personally guaranteed a $1.2 billion bridge loan controlled by Fortress and secured in part by the General Motors Building, which he had bought in 2003 for a then-record $1.4 billion and which was never part of the Equity Office portfolio at all.¹² Deutsche Bank had already sold pieces of the debt on to other lenders, which made any workout a negotiation with a crowd.

By February 2008 he was handing the keys to the seven towers back. In June the General Motors Building went to Boston Properties and its partners for about $2.8 billion.¹⁸ The building he had bought before the deal paid for the deal he could not finance.

The broker who had sold him the buildings on Blackstone’s behalf, Eastdil Secured chairman Ben Lambert, described the mechanism without any euphemism at all:“There was a game of musical chairs. The music stopped, and there was no chair for Harry.”¹²

Beacon Capital did not collapse, but it did not escape either. The $2.7 billion loan behind its share of the portfolio went to a special servicer, and in December 2010 Beacon negotiated a five-year extension to May 2017 — agreeing, as part of it, to sell some of its strongest Washington assets to pay the debt down.¹⁴ Its president told the Post that lenders had stayed with the firm because it had kept the capacity to manage and lease through the downturn.¹⁴ Eleven of the twenty properties it had bought were in and around the District.

None of that is incidental to judging the trade. The seventy per cent Blackstone recovered by July 2007 was recovered from buyers, and several of those buyers were holding the same assets a year later at prices the market no longer supported. The skill was real. So was the fact that somebody had to be on the other side of it.

What the record will not settle

Three things about this deal are repeated confidently and are not in the record.

The first is the size of the portfolio. Equity Office said 580 buildings and 108.6 million square feet when the deal was signed, using figures from 30 September 2006, and 543 buildings and 103.1 million square feet when shareholders approved it, using figures from 31 December.^1,7 Both are the company’s own counts. The second is closer to the truth of what actually changed hands, and it is the one used here.

The second is how much had been arranged by closing day. Two deals are dated to that week and they come to about $13.5 billion, not the twenty billion often quoted.^8,9 The larger number belongs to July, and it is a percentage rather than a sum.

The third is the rate of return, and it does not exist. There is a profit figure and a multiple, reported once, twelve years after the fact, on Blackstone’s own accounting.¹⁶ There is no audited number, no investor letter in the public record, and no IRR anywhere.

There is also a legal record, and it is short. Eight shareholder suits were filed in Maryland, Illinois and federal court, alleging the trustees had breached their duty by not running a full auction.² The federal case was dismissed and the dismissal affirmed by the Seventh Circuit in March 2009, Judge Richard Posner writing that requiring delay for every new competing offer would “sink the process of corporate acquisition into a sea of molasses”.¹³ The process survived review. That is not the same as the process having been open.

What transfers

The lesson is not that Blackstone timed the market. Nobody at the firm has ever claimed that, and the record does not support it: they paid a full price at the top and said so by paying it. What they did was refuse to hold what they had bought.

A portfolio trades at a discount to the sum of its parts, because the number of buyers who can absorb five hundred buildings is very small and the number who want one tower in Seattle is not. That discount is not a market inefficiency to be discovered; it is a structural fact about who can write which cheque. The trade is available to anyone willing to do the work of selling the parts — and the work is the point, because it means lining up the buyers before you own the asset, not after.

Two things made it survivable rather than merely clever. The exit was arranged in parallel with the purchase, so the sell-down began the week the deal closed rather than the quarter after. And the fee bought certainty of closing, which is what let the pre-arranged buyers treat their side as real. A deal that might close in four months with stock cannot support a pre-sold exit; a deal that closes on Friday for cash can.

And judge the outcome by the whole hold. Seventy per cent of the price came back in five months, and the last building took another twelve years to sell. Both of those are true, and only one of them is the story usually told.