This offer was rejected. In the words of one LTCM partner, “If we’re worth so much money, why would we sell?” This rejection was hubris. If the partners had sold, LTCM would have been part of JPMorgan when the 1998 crisis hit. JPMorgan would have saved LTCM to protect its own reputation. At the time, Bank of America owned D. E. Shaw, another hedge fund that lost billions in 1998. Bank of America quietly propped up D. E. Shaw in the panic. Today D. E. Shaw thrives as a $37 billion asset manager and technology firm. Shaw had a big brother. LTCM did not.
Style drift into merger arb and rejection of JPMorgan’s 1996 buyout offer were the first nails in LTCM’s coffin. The last nail came in 1997, just prior to the collapse, when the partners embarked on a plan to buy out their original investors. They would own the management company, and the fund itself. This plan was a portal to dynastic wealth.
By September 1997, LTCM’s fund capital approached $7 billion, a huge leap from the $1 billion we started with in 1994. Still, investment returns were declining. LTCM’s large size in favored trades meant diminishing marginal returns, as those trades grew even larger. Banks were copying LTCM’s strategies, making those trades less profitable for all participants. LTCM’s partners realized if they pushed out their original investors, they would own a larger piece of the pie and could capture more profits for themselves. To enrich themselves they would leave their original backers out in the cold.
The partners’ plan had two parts. The first part was simply to give back money to the outsiders through a forced redemption. This was done as of December 31, 1997, with a $3 billion all-cash distribution. The redemption reduced LTCM’s capital to about $4 billion. LTCM partners personally owned about $2.6 billion of that capital with the remainder in third-party hands.
The second part of the plan, led by Myron Scholes, involved an ingenious options strategy to control another $1 billion of the fund’s capital. LTCM persuaded UBS to sell the partners a seven-year, $1 billion at-the-money call option on their own fund. This option allowed the partners at any time from 1997 until 2004 to pay $1 billion to buy that amount of the fund plus performance on $1 billion from the day the option was sold. This effectively made the partners owners of the future performance on $1 billion of fund capital.
UBS charged the partners about $300 million for this option based on Scholes’s own options pricing formula. UBS hedged the option with a $1 billion investment in the fund. As LTCM made money, UBS would owe more to the partners under the option while profits on the LTCM investment would compensate for that obligation. UBS was hedged on the option and could pocket the $300 million premium.
The new UBS investment was used to finance part of the 1997 distribution to the outside investors. When the dust settled in early 1998, the fund had $4 billion, which was owned $2.6 billion by the partners, $1 billion by UBS, and $400 million by a few foreign banks as strategic relationships. Because the partners effectively owned the UBS investment through the options structure, their real economic ownership was $3.6 billion, or 90 percent of the fund. LTCM was morphing from a hedge fund to a multifamily office with no outside investors at all.
What was curious about the $1 billion option sold by UBS was that it hedged only future profits. UBS did not hedge future losses. It never occurred to anyone that LTCM could lose money. UBS thought it had insured an unsinkable vessel even as it set sail on the Titanic.
Vortex
The early months of 1998 were quiet in capital markets. LTCM’s profits were small but steady. The fund was on track for a good, if not spectacular year.
The year before, a financial crisis had emerged in Asia in July 1997, starting with devaluation of the Thai baht. Devaluation caused massive capital flight by hot money that chased the carry trade there. In the mid-1990s, investors borrowed cheap dollars, converted to baht, and invested in high-yield Thai development projects in resorts and other real estate. Conversion of dollars to baht was considered low risk because the Thai central bank maintained a fixed exchange rate to the dollar and the baht was freely convertible. Unexpectedly on July 2, 1997, Thailand broke the peg to the dollar. Immediately the baht dropped 20 percent. Lenders suffered colossal losses. Thailand turned to the IMF for technical assistance. Foreign investors dumped local investments and pulled their money out of Thailand. A global panic arose.
The turmoil next hit Indonesia and Korea, which had pursued policies similar to Thailand’s. On August 14, Indonesia broke the rupiah’s peg to the dollar. The rupiah went into free fall. Panic spread to the streets. Money riots erupted. Police responded with force, and some rioters were killed. The IMF imposed austerity measures that made matters worse.
Global investors no longer trusted emerging market exchange rate policies. They wanted their money back. Dread spread to developed economies. On October 27, 1997, the Dow Jones Industrial Average fell 554 points, its biggest one-day point loss ever. The word “contagion” was widely used in financial circles for the first time since the new age of globalization began in 1989.
The International Monetary Fund acted as a first responder to put out financial fires. The IMF provided cash to Korea, Indonesia, and Thailand to reinforce their reserve positions. In exchange, the IMF imposed harsh conditionality including budget cuts, tax increases, devaluations, and other draconian steps designed to save banks and bondholders at the expense of everyday citizens. Despite misery, the IMF’s castor oil approach worked. By January 1998, events seemed under control. The IMF fireman put out the Asian fire.
From the serenity of Greenwich, LTCM’s partners viewed these events not with alarm but with curiosity. If markets collapsed so suddenly, there must be cheap assets around that the computers could locate. Meriwether asked his analysts to find Indonesian corporate debt to buy on the cheap. There was blood in the streets, but for the partners and computers in Greenwich, Indonesia was just another trade.
April 1998 was a losing month for LTCM; the partners were not sure why. Markets seemed quiet. But below the surface the earth had started to shake.
On April 6, 1998, Travelers Group and Citicorp, parent of Citibank, announced a $140 billion merger, the largest in history. Travelers was controlled by legendary Wall Street denizen Sandy Weill. The merger was the capstone of Weill’s comeback after he was pushed out of American Express in 1985. The year before the Citicorp deal, on September 24, 1997, Weill and Travelers announced plans to purchase Salomon Brothers from Warren Buffett. This marked Buffett’s exit from his Salomon rescue in 1991. Unbeknown to us at the time, the admixture of Weill’s Salomon and Citicorp deals doomed LTCM.
Salomon was Meriwether’s alma mater. Unsurprisingly, a new cadre of Salomon traders trained by Meriwether were mimicking his trades. These spread trades were volatile. Spreads could widen before they converged. In that case, mark-to-market losses appeared on the books. These losses never troubled the quants because they were sure the spreads would converge again once markets settled down. Trading losses were sometimes viewed favorably because they offered a chance to buy more at a better price, like doubling down after a losing bet in roulette. The difference was the quants believed they had house odds, not gambler’s odds. The traders were betting on a sure thing; it was just a matter of time before they won big. They just had to keep doubling down.
Weill despised the double-down mentality, and the volatility that came with it. His technique was to build financial empires by buying targets using his own stock as currency. Weill wanted Travelers’ stock price as high as possible to buy Citicorp with minimal dilution of his position in Travelers. Stock markets punish stocks with volatile earnings by discounting their price. Weill ordered traders at Salomon to close out their spread positions to reduce volatility in Travelers’ earnings. Traders resented this order, but had no choice.
An unwind meant traders sold spreads instead of buying them. This pushed spreads out further, causing losses at firms like LTCM and Goldman with similar trades. At first, LTCM partners thought t
his looked like more of a good thing. They added to their positions at attractive valuations. Still, the spreads widened. Weill’s order to close out positions was a snowflake; an avalanche awaited.
Market calm returned in June. I took advantage of the quiet to join an Alaskan expedition to climb Denali, the tallest mountain in North America, more than twenty thousand feet high. The 1998 season was one of the worst ever on Denali. Bad weather led to fatalities, including a friend, guide Chris Hooyman, blown off a ridgeline by a 100-mile-per-hour gust as he unclipped his harness to rescue a struggling client. A British special forces team training at nineteen thousand feet was rescued after suffering injuries in one of the highest helicopter rescues ever attempted. Several Korean climbers were killed falling three thousand feet down a steep couloir nicknamed Orient Express. I was fortunate to have an excellent climb with legendary guide Dave Hahn, who arrived in Alaska from Nepal after summiting Mount Everest. I had no idea my dangerous season on Denali was just a warm-up for what awaited on my return.
In August, bond spreads widened again, and losses began to mount at LTCM. Still, 1998 was shaping up as a so-so year for the fund with gains in the single digits instead of the higher returns we were used to—a bad year but not a disaster. In mid-August, I went on vacation with my family on North Carolina’s Outer Banks. The other partners were on vacation too, mostly at golf resorts around the world. The Hawk was in Saratoga for the Thoroughbred racing season. Markets were choppy. Still, it was steady-as-she-goes at LTCM—golf, Thoroughbreds, and cocktails at sunset.
Then came the earthquake.
On Monday, August 17, 1998, Russia defaulted on its internal and external debt and devalued the ruble against the dollar. Defaulting on external dollar-denominated debt and devaluing the ruble was shock enough. Still, there seemed no reason to default on internal debt because it was denominated in rubles, which Russia could print. The internal debt default was senseless, yet it happened.
The global financial crisis returned with a vengeance, although it had never really gone away; the virus was merely dormant for a time. Contagion spread from Asia to Russia. Investors studying Russia’s inexplicable moves decided anything was possible. Brazil was fingered as the next domino to fall. Suddenly everyone wanted her money back. Stocks went into free fall, liquidity was king, nothing else mattered.
On the morning of Friday, August 21, the phone rang at my vacation home on the Outer Banks. The caller was Jim McEntee, an LTCM partner, and the only one with an old-school trader’s temperament. McEntee did not have a Ph.D.; he had worked his way up from the back office at Chase to found his own investment bank, which he later sold to HSBC. He had an uncanny feel for markets you could not put into an equation. He said, “Jim, we lost $500 million yesterday; the partners are meeting Sunday. You should get back for this.” I did. We packed the car and drove nine hours to Connecticut. The next six weeks were like one long day of damage control.
LTCM had 106 trading strategies involving stocks, bonds, currencies, and derivatives in twenty countries around the world. From the outside, the trades seemed diversified. French equity baskets had low correlation with Japanese government bonds. Dutch mortgages had low correlation with Boeing’s takeover of Lockheed. The partners knew they could lose money on a given trade. Yet the overall book was carefully constructed to add profit potential without adding correlation. Trades were designed to produce composite profits based on relative value spreads, even if spreads widened in one particular trade.
This diversification was a mirage. It existed only in calm markets when investors had time to uncover value and cause spreads to converge. However, there was a hidden thread running through all 106 strategies, what Scholes later called “conditional correlation.” All of the trades rested on providing liquidity to a counterparty who wanted it at the time. LTCM was a buyer of risk others wanted to sell. Suddenly everyone wanted to sell everything. Investors did not care about relative value; they wanted absolute value in the form of cash. LTCM’s solution to this was a capital cushion so it could ride out temporary liquidity demands. The $4 billion of capital was supposed to be enough. Now it appeared LTCM had constructed a ten-foot seawall to stop a fifty-foot tsunami. After losing $500 million in one day, the $4 billion would not last long.
LTCM’s first response was to raise new private capital. The estimate was that $1 billion was enough to cover losses and restore confidence. Time was short. The partners knew what was lost, but banks and regulators did not. Hedge funds, including LTCM, typically report results monthly; the daily internal update was not public. The next investor report would reflect the losses in a closing valuation on August 31. We had one week to raise $1 billion in cash before the world discovered what had happened.
August is the worst month to get anything done—let alone raise $1 billion. The rich and powerful are on vacation on yachts and in villas in exclusive locales. Still, the LTCM partners had the best financial connections in the world. Calls went out to George Soros, Prince Alwaleed bin Talal, and Warren Buffett. I call these three the “usual suspects.” They always get the urgent phone calls; they don’t always invest.
Buffett turned us down after a polite meeting in Omaha with LTCM partner Eric Rosenfeld. Buffett is notoriously wary of derivatives, what he later called “financial weapons of mass destruction.” No amount of valuation math from a Harvard professor was going to change that.
Soros and Prince Alwaleed also said no. Their reasons were subtle. A bad situation can always get worse. If LTCM was holding spread trades with embedded gains, those potential gains would only increase as spreads widened. Why throw a lifeline to a drowning man when you can wait for him to drown and collect the life insurance? Soros could afford to wait; desperate sellers only get more desperate.
By August 31, losses at LTCM were $2 billion, 50 percent of our original capital. It seemed surreal that we were still standing, still meeting margin calls, and still operating every day. The reason was our contracts did not give counterparties a way out. LTCM consistently refused to sign termination clauses with subjective criteria such as “material adverse change.” We insisted on a numeric trigger of $500 million of remaining capital for an early termination of contract. At that level, counterparties could cancel trades and take collateral. This made sense in 1994 when capital was $1 billion; the $500 million trigger equaled a 50 percent decline. Once capital reached $4 billion, the same numeric trigger represented an almost 90 percent decline. At that point, collapse is unstoppable; a 10 percent cushion won’t save you. Banks realized to their horror that they were locked into seat belts on the same flaming plane as the LTCM partners. We would all crash together.
Then a new panic hit the banks. What if losses from LTCM caused one of the banks to fail? What if your bank was also exposed to that failing bank? Who were the weak hands, and how would the panic end? Now the banks not only feared LTCM, they started to fear one another.
On September 2, we announced August results to our investors. I wrote the investor letter, and located Meriwether in our private gym locker room to ask him to sign it. He looked like a man about to sign his own death warrant. I knew the letter would leak immediately. In 1998, we still used fax machines. I had about forty letters to send. The first fax recipients leaked the letter to Bloomberg before the last fax was sent. CNBC also picked up the story. The panic was no longer about Russia or Brazil; it was about LTCM. We were the eye of the storm.
The fund continued to bleed in early September. We carried on with our capital-raising strategy; only now the target was $2 billion. Having failed with our private network, we hired Goldman Sachs as bankers. They arrived as a deal team in our Greenwich offices. I approached Goldman’s lawyer and asked him to sign a customary nondisclosure agreement. He burst out laughing and said, “We’re not signing anything.” I had no leverage, but I knew where things stood. I had been on Wall Street long enough to know predatory behavior was the rule, not the exception.
By September 17, the death watch had begun. LTCM still had cash and capital, but no hope of recovery despite wishful thinking by certain partners. A discreet phone call was placed to the Federal Reserve Bank of New York. There was no request for a bailout, no expectation of one. It was inconceivable to us that the Fed would bail out a hedge fund. We simply wanted the Fed to know the situation. It seemed bizarre that Goldman had our information and the Fed did not. So we invited the Fed in.
On Sunday, September 20, a Fed and Treasury delegation arrived in our Greenwich offices led by Peter Fisher, head of open market operations at the New York Fed. Fisher was accompanied by his close associate Dino Kos and Gary Gensler, then deputy assistant secretary of the treasury and protégé of Treasury Secretary Bob Rubin. Fisher, Kos, and Gensler sat down with Meriwether and me in the partners’ conference room. For the next five hours we went through LTCM’s positions line by line, trade by trade, counterparty by counterparty. When we were done, Fisher’s face was white. He remarked, “We knew you guys could shut down the bond markets, but we had no idea you would shut down stock markets too.” He was reacting to the $15 billion in takeover deal stocks on our books. If LTCM defaulted, Bear Stearns’s hedged stock positions would instantaneously become net long, as the short to LTCM would disappear. Bear would then dump $15 billion in stocks into a falling market to balance its own books. Market contagion and panic from such selling was inevitable.
The Road to Ruin Page 15