Another fine mess at UBS
Tuesday September 22, 8:30pm
Apart from festive occasions, the fortress-like Federal Reserve Bank of New York has seldom hosted such an illustrious gathering of Wall Street heavy-hitters. It included Sandy Warner, chairman of JP Morgan; David Komansky, chairman of Merrill Lynch, and Herb Allison, its president; Frank Newman, chairman of Bankers Trust; Jon Corzine co-chairman of Goldman Sachs and Robert Katz, managing director; Deryck Maughan, co-CEO of Salomon Smith Barney; Allen Wheat chairman of Credit Suisse First Boston; Philip Purcell, chairman of Morgan Stanley; and Jimmy Cayne, president of Bear Stearns & Co, and Warren Spector, deputy president.
The occasion was far from festive. "I sensed a lot of fear in that room," recalls a participant.
The object of fear was Long-Term Capital Management (LTCM), a hedge fund that in four years had built up assets of close to $125 billion on a capital base of $4 billion. It says something for today's financial system that that in itself wasn't thought alarming. Banks were comforted by the fact that although LTCM's borrowings were huge, they were mostly collateralized, mainly with high-grade securities.
The scary part was LTCM's off-balance-sheet business - swaps, options, repos and other derivatives. These multiplied its declared leverage of close to 30 times capital by another 10 to 12 times: a notional principal amount of over $1 trillion.
Even that might have been bearable had the derivatives been simple interest-rate swaps. But spicing the plain-vanilla deals was a cocktail of equity derivatives, total-return swaps, index options and bets on takeover targets. "We discovered LTCM was short 30% of the total volatility of the CAC 40 French stock index," says an eye-witness.
Until the previous weekend, when teams from Merrill, Goldman and JP Morgan had been invited to examine the books at LTCM's headquarters in Greenwich, Connecticut, each counterparty had seen only the tip of the iceberg: its own off-balance-sheet positions with the firm. Now they were beginning to see the whole devastating picture.
The fear on Tuesday night was not the net position of those in the room. Each firm had an Isda (International Swaps & Derivatives Association) master agreement in place with LTCM that sanctioned close-out netting and offset of all bilateral positions, boiling them down to a single number theoretically covered by collateral. On this basis one firm even owed LTCM money.
But it was the gross position that was terrifying. LTCM had 36 swap counterparties. When it wanted to reverse a trade, instead of contacting the original counterparty to unwind the deal on penalty of a spread, it had written a mirror trade with a new counterparty to avoid paying the spread. "It was like a layer-cake of swap upon swap," says a shocked counterparty.
A single default by LTCM would trigger the cross-default of all these trades. Accelerating each deal into a net amount for close-out would be an irksome but mechanical process.
But many counterparties would then have to worry about the other side of the trade, the market position that had matched it. There would be a massive need to rehedge in the marketplace. In the case of the French CAC 40 trades, banks that had sold those volatility positions to clients would be scrambling to rehedge with a possibly severe effect on the French stock market - a worry for the Banque de France. As for LTCM's bond and equity spread trades, almost the entire speculative financial sector was in the same position - long high-yield assets and short low-yield assets. Given the already unstable state of global financial markets, with Asia, Russia and perhaps Brazil on the brink, the risk of dramatic price falls even in the US was high.
Earlier that day, New York Fed chairman Bill McDonough had told a London audience this was a "situation which I regard by some considerable margin as the most dangerous since the Second World War", before rushing to catch a plane to join the meeting.
This was not a debt rescheduling of some distant country. In this case the banks didn't need to be bullied into cooperating. They were faced with a monster they themselves had nourished and inflated. Some of them had had such faith in LTCM that they had pitched in their personal capital and the deferred benefits of their employees. If LTCM blew up, not only might world financial markets blow up too, but some of the heads around that room tonight would roll.
As those shell-shocked bankers set about a rescue plan, the full-length portraits of McDonough's predecessors looked on. One of them, Gerry Corrigan, had warned in January 1992: "...in the event of a major market disruption I assure you that it will be the gross, not the net, that will really matter in most segments of the financial marketplace nationally and internationally." At the time many dealers in the decade-old swap industry had dismissed Corrigan's warning as "naive".
But the blow-up could come as early as the next day. Bear Stearns, which acted as clearing agent for LTCM, warned that it would no longer take the clearing risk.
Gasping for cash
LTCM had been haemorrhaging cash on unwinding trades since August, and perhaps before, ever since it was known to be in search of new capital. Its counterparties knew enough about its positions to trade against them at every turn, pushing up the price on that side of the market. On September 2 LTCM sent a letter to investors admitting a 44% fall in its net asset value in August alone. JP Morgan and Goldman were asked to find a fresh capital injection from somewhere, fast.
On September 8, with some bravado, LTCM paid off $80 million that it had drawn on a $900 million standby facility syndicated by Chase - it had agreed not to draw on it while the facility was renegotiated. But within a week, when Bear Stearns demanded an extra $500 million in collateral, LTCM was scrabbling for cash.
As the weekend of September 19/20 loomed, Goldman and JP Morgan knew they had only a few days to find more equity. Peter Fisher, executive vice-president at the NY Fed, had heard enough in daily chats with bankers to know they were all concerned about Long-Term Capital Management. He also sensed there was too much rivalry between the main players for them to get round a table and sort this out by themselves.
On Sunday September 20 Fisher went to LTCM in Greenwich with assistant treasury secretary Gary Gensler and two other NY Fed colleagues. Goldman and Morgan had also sent people. Having looked through LTCM's books, they were all staggered at the sheer volume of trades. LTCM management had stress-tested those trades, but for only a 10 basis point daily interest rate move. And they had looked at only the 12 biggest deals with their 20 biggest counterparties. Their loss estimate on that gentle stress test was around $3 billion "but we reckoned it would be more like $5 billion", recalls a source close to the investigations. And that was ignoring all LTCM's other trades. The thought of closing them out in one to three days "boggled the mind", the source says.
Monday September 21 was the Jewish new year holiday. But senior partners from Goldman, Merrill and JP Morgan, in a number of telephone conversations, some involving Fed officials, reviewed the sheer enormity of the problem. They were casting about for a solution that might involve a single big buyer. "But Mr Big wasn't forthcoming, although we kept looking for him," recalls one of those present. Goldman Sachs had some big institutional relationships, and the others hoped it would strike gold. But if Goldman had a Mr Big standing in the wings, it wasn't saying. That evening UBS, another bank with heavy exposure to LTCM, had a team in Greenwich doing "due diligence" on LTCM, which perhaps it should have done months before.
Meanwhile there was a further sell-off in global equity markets, not in anticipation of the video broadcast of US president Bill Clinton testifying on the Monica Lewinsky affair, as rumours had it, but because of LTCM.
Fisher saw that time was slipping by while markets plunged and LTCM's counterparties were still sparring with each other rather than uniting to solve the problem - a classic case of the prisoner's dilemma.
The three US banks and UBS came to a breakfast at the Fed on Tuesday morning, September 22. Once again the request was made: "If there's a Mr Big out there, let him step forward." But no-one in the room could produce a Mr Big.
The bankers decided to form working groups to study possible approaches which included a purchase of LTCM's fixed-income positions, mostly spread trades between high-yield and low-yield bonds, and a "lifting" of LTCM's equity positions, which were a mixture of index-spread trades, total-return swaps, and bets on takeover targets. During the day, teams working in Greenwich on the first two options encountered legal obstacles that made lifting the positions near-impossible. But another option, a recapitalization of LTCM by its creditors, looked more promising. It needed a consortium of more than four banks, however. Based on a commitment that the core group was ready to present a joint proposal, major creditor banks were invited to a meeting at the Fed that evening.
There was no time to lose: Bear Stearns was threatening to pull the plug on LTCM. LTCM, gasping for more cash, drew $500 million from the Chase facility it had promised not to touch. Only $470 million was delivered - since two syndicate members refused to chip in.
At the Fed at 7pm, Fisher met with the four lead firms to review a draft term sheet drawn up by Merrill. At 8:30 Fisher opened the wider meeting of 13 banks, quickly giving the bankers the floor to ensure this wasn't seen as a Fed rescue. The Fed's initiative had been needed to help the bankers overcome their street rivalries, and perhaps to shield them from accusations of operating a cartel. But now they were in the Fed boardroom, around this fine walnut table, it was up to them.
After some initial Wall Street skirmishes, Chase head of global credit risk David Pflug, veteran of many country workouts, warned that nothing would be gained a) by raking over the mistakes that had got them here, and b) by arguing about who had the biggest exposure: they were all in this equally and together.
The working group had decided they would call on 16 banks to put in $250 million each to restore LTCM to its former $4 billion of capital. But there were only 13 banks around the table and some might plead poverty.
Merrill president Herb Allison quickly became the natural moderator of the meeting. Merrill after all had been in at the beginning, raising LTCM's start-up equity in 1994.
A large part of the discussion centred on LTCM's management - should they be kicked out or kept on to help dismantle the monster they had created. John Meriwether, the chairman, had been the darling of Wall Street, one of the greatest fixed-income traders Salomon Brothers had ever known, until the treasury-auction scandal of 1991 led to his quitting the firm. By 1994 he had collected around him to found LTCM not only a handful of Salomon's star traders such as Eric Rosenfeld and Lawrence Hilibrand (who earned $25 million in a year at Solly), but also among the best brains in finance - Myron Scholes and Robert Merton (Nobel prize-winners last year) - and David Mullins the former vice-chairman of the US Federal Reserve Board. It was a dream team combining Wall Street brawn, Chicago brain and Fed respectability. Although unregulated, LTCM became almost another bulge-bracket counterparty - the Rolls-Royce of hedge funds. It made 47% return on equity in 1995, 45% in 1996, and a still respectable 17% in 1997. At the end of 1997 it announced it was handing back $3 billion of capital to its investors. The message seemed to be that the fund had grown too big to do efficient bond arbitrage without distorting the markets it moved in.
But that clearly hadn't been the case. What it had wanted was more leverage to make a higher return for fewer investors. As well as bond arbitrage it was straying into areas far from its known expertise - total-return swaps, equity indexes and takeover arbitrage. Off-balance-sheet instruments allowed it to do much of this without disclosure even to its closest investors. The management were also increasing their personal stakes, borrowing money to invest in the fund. It was reckless gambling, either because they were so sure their bets would come home, or because they were desperate to recapture the high returns of 1995 and 1996.
Unfit but unsackable
By any measure they had proved themselves unfit to run their own shop. But getting rid of them would be more dangerous than keeping them. If there were any nasty surprises still hidden on or off LTCM's balance sheet they would know where to find them. Too many derivatives firms had suffered shocks after the departure of managers who had left behind positions whose full complexity only they knew how to unscramble. Better to keep Meriwether and his team locked in, better still with the incentive of an equity stake. That argument prevailed.
But matters couldn't be resolved that night. Although Komansky, Warner, Corzine and others might have the authority to pledge their institutions' money, others in the room didn't. Chase also wanted an assurance that $470 million of any new money would go straight to repaying the funds LTCM had drawn that day. Although LTCM's drawdown had been quite legal, it seemed unfair that this 11th-hour call for cash should be regarded as part of any new capital. They decided to resume the meeting the next morning.
As they were all gathered on Wednesday September 23 at 9:30am the Goldman partners took McDonough aside. Mr Big had arrived.
Warren Buffett, regarded as one of the world's canniest investors, and the man who rescued Salomon Brothers at the time Meriwether resigned over the treasury bond scandal, was offering to take over LTCM for $250 million, and to recapitalize it with $3 billion from his Berkshire Hathaway group, $700 million from AIG and $300 million from Goldman. Meriwether and his team would have no further management role.
Buffett had first been contacted by Eric Rosenfeld as far back as Sunday August 23. The following Wednesday Rosenfeld asked whether Buffet could meet Hilibrand. The fund was worried that at the end of the month it would have to reveal its losses to investors and wanted to be able to reassure them that new funds had been secured. Buffett wasn't interested. Nor was he when called again on September 18 by Goldman partner Peter Kraus, but over the next four days the two did hammer out a plan Buffett liked. That was the bid put to Meriwether on September 23.
There were conditions: the buyers would not assume any of LTCM's existing liabilities, and current financing arrangements must stay in place. The toughest condition, however, was that the offer would expire in around 90 minutes, at 12:30 that day.
The consortium meeting was adjourned while Meriwether deliberated. By 1pm the meeting was reconvened. Meriwether had been unable to accept the Buffett offer: his lawyers told him he didn't have the authority to sell the fund, especially in 90 minutes, without polling his investors.
The banks were somewhat shocked at Goldman's dual role - as consortium member and potential investor with Buffett. But Goldman had been acting for LTCM with no agreement on non-disclosure. Moreover, at all the meetings at the Fed, a continued search for any and every new investor had been unreservedly encouraged. Client confidentiality probably prevented Goldman from disclosing earlier that it was talking to Buffett, although Goldman has refused to comment publicly on this or any other aspect of the LTCM debacle.
The consortium meeting resumed and in five hours a deal was struck. Still reluctant to believe the Meriwether magic was dead they decided to go for equity not debt. A consortium of 11 banks would invest $300 million each, one other would put in $125 million, and two more $100 million each, bringing the total investment to $3.625 billion. LTCM's management and original investors would be left with 10% of the equity (which valued their asset at around $400 million, rather more than Buffett's $250 million).
The ink on Merrill's term-sheet was hardly dry before the accusations began of crony capitalism at work, Wall Street looking after its own, top bankers protecting their own investments and their jobs.
The Fed also underwent a grilling. Was it right to intervene to help save not a bank that was too big to fail, not even a small bank, but an unregulated hedge fund? McDonough and Federal Reserve chairman Alan Greenspan put their case to the house banking committee on October 1. Most financial experts judge that the Fed's initiative was timely and admirable. "Either it did the right thing, or it didn't," says Henry Hu, professor of financial law at the University of Texas. "But there are some costs to this. The Fed had to cash in a lot of chips. The risk is you lose yet more market discipline." He cites the evil of "private gains and socialized losses". Yet, in this case, not a penny of public money was chipped into the rescue.
While such views may have some popular appeal they don't take much account of the dilemma that faced the Fed that September weekend and the job it is supposed to do. Should the Fed have sucked its thumb while the financial markets descended into chaos? Fed sources are aware of the dilemma. The Fed, while wanting a refinancing, kept distant from the terms and conditions of the deal, providing only a venue, sandwiches and coffee. Investigations may one day establish whether the rescue simply picked one errant trader off the floor or prevented a global meltdown. Michael Goldman, who runs Momentum UK, which tracks and invests in many hedge funds, is unequivocal on this: "For a few hundred million the rescuers saved themselves billions."
If the Fed made one mistake it was to cast its net too narrowly, suggests one rescue participant. It implied the rescue was specific to LTCM but it should have invited in a bigger cast of banks citing "general market reasons". The losses subsequently published by BankAmerica and DE Shaw, and hinted at within Goldman itself, suggest that market chaos would have hit the biggest players when they were already highly vulnerable.
The lessons for the street are: don't be star-struck; and take account of your gross as well as your net exposure.