Deal: Block trade in BP stock
Amount: £1.2 billion
Launched: May 14
Firm: Goldman Sachs
At 4.45pm on May 14, Gary Williams, head of equity trading at Goldman Sachs in London, was in a routine meeting at UK regulator the Securities & Investments Board. A note was passed to him from the secretary of Wiet Pot, Goldman’s London head of equity sales. It was terse: “There’s a big trade on, get back here.” Williams sensed something out of the ordinary and immediately announced he was leaving. He grabbed a mobile phone – luckily one belonging to a Goldman colleague – and was dialling on the way out of the door.
“Big trade” was an understatement. What was beginning to unfold was the largest-ever block trade in the UK equity market, the sale by the Kuwait Investment Office (KIO) of 170 million shares in British Petroleum (BP) – 3% of the company’s equity worth £1.22 billion (nearly $2 billion). In the next 18 hours, Goldman won and executed this trade – a complicated cross-market block deal including ADRs in New York and ordinary shares in London, each trading on different settlement cycles.
The process that led to Williams’s breathless rush back to Goldman’s offices in Peterborough Court had begun months earlier when the KIO appointed Schroders to advise on and structure a placement of part of its 9.3% holding in BP. This holding was a legacy of the KIO’s purchase at the time of BP’s difficult flotation just after the 1987 crash and its later sale back of part of its holding to the company, at the behest of the UK government. In the succeeding decade the value of KIO’s holding had grown to some $6.5 billion. The KIO concluded that this was too much exposure to a single stock. Its motive for the sale was simple: portfolio rebalancing.
Over the years, many banks had approached the KIO with suggestions for selling part of its BP stock and once or twice a year rumours would circulate of a deal in the offing, the latest rumour being this February. But when the sale finally came, it caught everyone by surprise. With most UK block trades the sale is flagged in advance, bidders pre-market to key investing accounts, the market begins to price in the new supply and the block is sold at a narrow discount but to a depressed stock price. This deal dropped out of the sky.
Other options had suggested themselves – a fully roadshowed offer for the block, a truncated 48-hour book build, a tender auction with some kind of underwritten back-stop price. Finally, though, a block trade in which secrecy was paramount was chosen. This allowed the KIO to select the absolute best time for the disposal and to proceed with lightning speed, rather than commit itself to an open timetable and risk either a general equity market downturn or a collapse in BP’s share price because of professional short-selling.
At 4.40 that Wednesday afternoon, Philip Mallinckrodt, a director in equity capital markets at Schroders, contacted three potential bidders: Goldman Sachs, UBS and a consortium of Salomon Brothers and Kleinwort Benson. The bidders were self-selecting on the basis of their previous proposals to the KIO for disposal of the block, their capital strength, their good oil analysts and their positions as leading traders in the UK and US stock markets.
Schroders faxed a confidentiality agreement to each bidder, requiring that it commit itself not to deal in the stock for 48 hours after the auction, if it lost the bid, and not to alert its investing clients that the block was coming, if it lost. Bidders were told that they had a small number of competitors, but not exactly how many or who they were. The losers would not be told who had won, though the winner would be told which other firms had also bid and lost. Schroders asked each bidder to sign the confidentiality agreement, send it back and dispatch their documentation teams to Schroders’ offices to study the underwriting agreement. Bids were to be submitted in pence within one hour. By 6pm or shortly after the winning bidder would be told that it now owned the stock.
Schroders was being paid a fee for its services and was not taking principal risk. But still it was sticking its neck out. No-one had ever asked a firm to put up $2 billion of its capital in such a way, in a blind competitive bid with no time for pre-marketing. What if no-one played ball? But they did. “What we achieved is a testament to the depth and development of the international equity markets,” says Mallinckrodt.
In well under 20 minutes Goldman assembled on a single conference call the top seven executives in its equities business on both sides of the Atlantic – as well as Williams, Pot, there was Patrick Ward, and Robert Steel, co-chief operating heads of the equities division; Eric Dobkin, global head of equity capital markets; Roy Zuckerberg, global head of equities and a vice chairman of Goldman Sachs; David Rogers, global head of equities risk management. Also on the line was John Thain chief financial officer of Goldman Sachs. A later call was made to the chairman, Jon Corzine.
The seven equity specialists discussed the price at which they could sell on the BP shares to investors. BP stock had spiked up earlier in the week, following strong quarterly earnings announcements, and stood at 744p. Goldman knew that many UK investing institutions were underweight BP stock against their UK market benchmarks. These investors, offered the chance to buy a big block of the largest UK stock at a discount to the market price, would need to come up with a good reason not to bite. American investors were tougher to read because their holdings in BP would be assessed against different benchmarks, of leading European stocks or of oil stocks. But American holdings of BP stock were clearly below their highs.
“Our starting point was that enough investors would care at 715p and we could probably get it done at 716p and maybe even 717p, but that any higher would be pushing too close to 720p and we could not get it done at that level without sloppy placement,” says Williams. “We had an obligation to all our clients, both potential buyers and the seller, who would retain a $4.5 billion investment in BP, to achieve the best quality placement at a price that would be sustained.” Goldman settled on 716p, and then factored in a 75bp turn for itself on the deal, leading to a bid of 710.5p.
As the conference call continued, documentation teams had been arriving at Schroders’ offices at 120 Cheapside. These teams had to be kept apart, in case they recognized each other and so identified the other bidding firms. Knowing that a rival firm is long $2 billion of a particular stock would be a dangerous piece of information for a trading firm to have. Schroders closeted the teams in separate meeting rooms, each with an open line back to their own firms’ sales and trading desks from where bids were called in.
By 5.45pm the bids were in. Goldman’s came back first with its bid of 710.5p. It beat Salomon Brothers and Kleinwort Benson by the narrowest conceivable margin. Market sources say they bid 710. The glory could so easily have been theirs. UBS bid around 705p, implying a spread between highest and lowest bid of less than 1% of BP’s prevailing share price.
By 6.10pm, Goldman knew it was the proud owner of 170 million BP shares – a $2 billion exposure it had not planned on taking, just 90 minutes earlier. It was now on a tight timetable. Because the KIO owns more than 5% of BP, it has to file major trades with the Securities & Exchange Commission in Washington. It was now 1pm there. By 9.00am Washington time the next day (2pm London time), the trade and the price would become public knowledge. Goldman would be exposed.
What Goldman did next was perhaps the most impressive thing of all. It did nothing. For nearly two hours the firm collectively paused and drew breath. Even the losing bidders must have wondered what was going on. By 8pm nothing had been seen in the market. Had the deal been placed? In fact, Goldman had not yet called a single account.
Between 7pm and 8pm, through a series of internal conference calls, Goldman made sure its economists, strategists and salesforces were prepared and brought up to speed on BP. The stock had just gone ex-dividend in New York, leading to heavy trading as the stock washed through various hands. That volume might offer Goldman some cover for the huge trade it was about to undertake. The firm briefly considered, and rejected, hedging strategies. It searched for correlated instruments to sell. Shell shares suggested themselves. But selling $500 million of Shell shares as a hedge might be counterproductive because many investors would value BP stock relative to Shell, and driving the Shell price down would lower the price at which these investors would buy BP. Taking out UK equity market risk by selling futures was also a non-starter. “The danger is that the placement fails and the markets rally and you increase your risk exposure,” says Pot. “We decided our best hedge was to put all our distribution power behind it.”
As at most firms, Goldman distributes shares through various salesforces, one for private clients, others selling different products to institutions, including teams covering UK, US and continental equities, convertibles and derivatives. BP could fit into most of these areas and was of potential interest to many accounts. Up to 500 people were involved in selling the block.
At 8pm London time, Goldman unleashed this army. It had waited until then to limit its potential price risk in New York, where there was just one hour left to the close of trading in ADRs. It was now 9pm in Europe. Clients were contacted at their homes. Goldman was concerned that some investors would worry about the Kuwaitis selling off an oil stock. What might the Kuwaitis know? But it was confident that BP was a very well regarded company, with a record of strong market performance. The KIO had said publicly that its remaining 6.3% holding in BP was a core position and the underwriting document spelt out a lock-up period during which this stake would not be sold. Crucially, BP had strong stock placement and reasonable liquidity in New York and London. Although 170 million shares is a huge number, perhaps equivalent to a month’s trading in BP (volume in BP stock in London on Tuesday May 13 was 8.1 million shares), turnover in the six oil majors in New York can be 200 million shares in a day. Oil stocks are among the most actively traded in the world. The high level of institutional liquidity was another crucial consideration.
By the close of trading in New York, just after 9pm London time, Goldman knew the risk it had taken would pay off. “It’s like early poll results. From 9.30 we could project the size of the final book,” says Williams. Goldman was now able to indulge in some refinement. Instead of selling the stock on a first-come, first-served basis, as is usual when a winning bidder is desperate to unload a big block, it took orders and promised to decide final allocations on Thursday morning, rather as in a book-building exercise.
Goldman eventually generated more demand than it had stock. It could easily have laid off its risk by placing far more than the eventual 40% it sold in the US. The KIO owned both ordinary shares and ADRs and Goldman had the right through the underwriting agreement to take up to 100 million of the 170 million shares as ADRs. But Goldman held back to ensure it had stock to sell in the UK and Europe, which eventually took 50% of the deal, and the rest of the world.
Partly, it wanted distribution to reflect the importance of the stock in each market, partly it was a matter of tactics. “If a client responds by indicating demand for two million shares and you immediately confirm stock sold, that suggests desperation, which is not the im pression you want to give,” says Pot. At the same time, institutions were asked to respond quickly, within just an hour or two, to be fair to early respondents and to create a sense of urgency and even competition between buyers. In such a speedy process, distribution was not entirely smooth. There was always the risk of an institution’s oil analysts or key decision maker being on vacation, off sick, at dinner at the wrong moment. “There were some surprises both ways, investors we thought would participate who didn’t and ones we thought would not buy who did,” says Pot.
By 11.00am on Thursday May 15, investors were informed of their allocations. The biggest single order accounted for 5% of the block. Goldman had no position. The stock dipped to close at 724.5p that day and later rallied up to 737p. But trading was stable. On May 16 volume in London was a normal eight million shares. As Euromoney went to press it once again traded at 724.5p, implying a discount for buyers at 716p of 1.2% to that price. Many people in the equity markets would have loved to have been in the middle of Goldman’s trading floor that night as the BP deal gripped the whole firm. But by the morning after, the atmosphere was one of anti-climax. Did the Goldman team celebrate? “We were too tired,” says Williams.
