Confidence Game

IF ONE WERE to pinpoint the day when faith in the bond insurers was irreparably undermined, it would probably be October 25, 2007. That was the day MBIA reported the second quarterly loss in its history. It had taken years, a short-seller crusade, and a two-year regulatory investigation to get MBIA to acknowledge its first quarterly loss.

This time, there was no disguising the fact. MBIA reported mark-to-market losses of $342 million to reflect the decline in the value of collateralized-debt obligations (CDOs) it had insured. Still, not everyone was convinced that the company was coming clean. “I’m trying to understand how the guarantors can take such low levels of mark-to-market losses relative to what the rest of the Street is taking on these securities,” Ken Zerbe, an analyst with Morgan Stanley, said during the company’s conference call.

The day before, Merrill Lynch had reported the largest loss in its 93-year history, taking $7.9 billion of writedowns mainly on super-senior CDOs backed by subprime mortgages. The writedowns reflected a discount of 19 percent to 57 percent of the face value of various types of securities Merrill held. “Bottom line is we got it wrong by being overexposed to subprime,” Merrill’s chief executive officer (CEO) Stan O’Neal told listeners on the firm’s October 24, 2007, conference call.

Then O’Neal did something extraordinary. He apologized for having been too slow to lay off the firm’s exposure to other firms and investors. As the subprime mortgage market collapsed during the first quarter, Merrill Lynch faced a desperate scramble to get mortgages, mortgage-backed securities, and CDOs off its balance sheet, O’Neal explained.

Unfortunately for MBIA shareholders, Merrill Lynch had succeeded in getting MBIA to assume more than $5 billion of its CDO risk. MBIA and Merrill had exposure to very similar securities. Yet if MBIA had taken writedowns of the same magnitude on securities it guaranteed as Merrill took on securities it held, then MBIA would have reported $3 billion, rather than $342 million, of losses, the equivalent of more than 45 percent of its insurance company capital.

MBIA chief financial officer Chuck Chaplin gave Zerbe an answer as to why Merrill and MBIA were taking such different markdowns on super-senior CDOs, but it was not easy to follow: “The vast majority [of CDOs] are marked using market quotes for the collateral in the deals and then the subordination of the deal as inputs into an analytical mode that then incorporates assumptions about correlation and the relationship between collateral spreads and financial guarantee premiums to calculate an implied premium for the deal,” Chaplin said. “And then the present value of the difference between the implied premium and the premium that we receive is the balance sheet value at period end,” he added.

Zerbe wasn’t the only equity analyst raising questions. Heather Hunt, who covered the bond insurers for Citigroup, asked why all of these super-senior CDOs were getting insured in the first place. Was it because the market is just at a standstill and having insurance is the only way of getting the deal done? Hunt asked.

Equity analysts had cheered MBIA’s move into the CDO business, yet it seemed they had a poor understanding of what was driving that business. This wasn’t the bond insurance business anymore. MBIA was providing a service that allowed banks to make huge amounts of securities disappear from their balance sheets. This disappearing act was known as a “negative basis trade.” It allowed financial institutions to book all their profits on vast CDO holdings up front while assuming away the risk of default. As the risk seemingly disappeared, so did the need to hold capital. All the banks had to do was buy protection on the CDOs from a bond insurer, and then hope the bond insurer didn’t go bust.

This was the trade Ackman had asked Warren Buffett about at the 2003 Berkshire Hathaway annual meeting, the transaction that Buffett had warned would bring out the brokers “with their fins showing.”

Then Chaplin touched on a topic that equity analysts watched very closely. MBIA had halted its buyback plan. “Going into a period of uncertainty, we think it’s critical for our balance-sheet strength to be unquestioned,” Chaplin explained. Within minutes, the stock swooned.

As the call drew to a close, the questions became more accusatory. “Are you saying that the mark-to-markets here are irrelevant because the markets are wrong?” one caller asked. “Because so far this year, the markets have been more than right, and the rating agencies have been totally wrong on virtually everything.”

The logic of the entire bond-insurance business was under fire. “Why would someone have wrapped these portions of the portfolios? The point of insurance is to protect yourself from losses. It seems that there is no instance that you could have a loss. That’s what MBIA seems to be telling us. So why do people buy your product?” a caller asked.

“What we are providing is protection against remote losses,” Chaplin replied. “That’s an important part of our business model, and there has been ongoing demand for protection against those remote losses at least going back over the 34-year history of our company.”

“Okay, if it’s based on remote losses and your model goes back 34 years, have you ever seen a housing market like this in the last 34 years?” the caller asked. “And what is your definition of remote? I mean, at which level is this a problem? At what point do you start to worry?”

If Chaplin knew where that level was, he didn’t appear eager to share it. “It is possible that as we see the housing market deteriorate that there will be pressure on transactions in our portfolio, and I think I said that pressure could be manifested as losses in the future,” Chaplin said. “We have not seen any to date. That doesn’t mean, given the potential for an unprecedented experience in the housing sector, that there will never be any losses, so it would be unfair to say that we expect that there never will be any loss no matter what happens.”

During the Gotham investigation four years earlier, the question of MBIA’s CDO losses had generated intense debate. Bill Ackman was asked to justify the dealer estimates he used in his report. To regulators, the gulf between Ackman’s and MBIA’s estimates for the bond insurer’s CDO losses suggested wild exaggeration on Ackman’s part.

“Why is their number $35.5 million and yours comes out to somewhere between $5 [billion] to $7 billion?” an attorney for the Securities and Exchange Commission (SEC) asked Ackman. “That is a big number.”

Someone was clearly wrong in the fall of 2007. This time, however, it wasn’t so easy to say that MBIA’s critics—or those skeptical of Merrill Lynch’s or Citigroup’s numbers—were the ones exaggerating.

By October 30, credit-default-swap (CDS) premiums on MBIA Inc. were bid at 250 basis points, meaning it cost $250,000 to buy protection against default on $10 million of MBIA Inc. debt. The next day it cost $295,000; by November 1, it jumped to $345,000. “The problem now is that investors fear the relationship between the rating agencies and the guarantors is so incestuous that the rating agencies cannot issue downgrades without implicitly conceding that their proprietary capital models are flawed,” wrote Kathy Shanley, an analyst with the research firm Gimme Credit.

By November 2, CDS rates jumped to $470,000; one day later, the cost stood at $510,000. “As the credit market continues to weaken, our confidence that guarantors will survive the credit meltdown is waning,” Zerbe, the Morgan Stanley analyst, wrote in a November 2 report ominously titled Financial Guarantors on the Knife’s Edge. Zerbe was forecasting that MBIA and Ambac each would take losses of $2.3 billion to $11.7 billion. For years, MBIA had guided analysts so that their forecasts came within a few cents of the company’s earnings-per-share results. Ideally, MBIA would beat the forecasts by just enough to show that any surprises were positive ones. That steady, predictable, pleasing world was gone.

Then on November 5, Fitch Ratings, the smallest of the three rating companies, said it planned to spend the next six weeks reviewing the capital of all the bond insurers. Fitch stopped short of putting the companies under formal review for downgrade. The credit-rating company warned, however, that recent downgrades of mortgage securities had been “broader and deeper” than Fitch had anticipated. Any bond insurer that failed the test would have a month to raise capital or have its rating cut. The companies were being given a running start.

What would it mean if the bond insurers lost their triple-A ratings? Bloomberg News collected some data and came up with an estimate that it would cost investors roughly $200 billion. It was almost certainly a conservative guess. Greg Peters, head of credit strategy at Morgan Stanley, summed it up. “We shudder to think of the ramifications,” he told me. It would be “a huge destabilizing force.”

For several weeks, the financial world waited for the credit-rating companies to rule. “We’re all rooting for the bond insurers,” a banker on the asset-backed securities desk of a bank in London told me. Without bond insurance, underwriters were going to find it next to impossible to sell asset-backed securities. Everyone’s job depended on the market picking up again. “Believe me,” he added, before leaving me with an indelible image of Wall Street’s support for MBIA and its competitors, “we come into work every morning with our pom-poms.”

ON NOVEMBER 8, the day ACA Capital, the smallest bond insurer, announced results, it became clear just how devastating a credit-rating downgrade could be. ACA Capital was the only A-rated bond insurer among a field of companies with triple-A ratings. The New York-based company was founded by Russ Fraser, a longtime ratings-company executive, who had been forced out by the board of directors in 2001 over his refusal to insure CDOs.

Fraser didn’t understand or trust CDOs that required analysts to abandon traditional credit analysis and rely instead on complex financial models. “Every PhD in the world is saying this is noncorrelated risk,” Fraser recalled of the CDOs brought before ACA’s credit committee. “But every CDO we looked at had the same 60 to 100 names. And they were never Exxon, Mobil, and Shell. The names were WorldCom and Enron and Tyco.”

By the fall of 2007, ACA Capital backed $65 billion of CDOs. Now, instead of exposure to Enron, the insurer had exposure to subprime. As a result, it announced a $1.7 billion writedown on its portfolio. Alan Roseman, ACA’s CEO, came out swinging on the conference call to discuss the quarter’s results. “We do not have a capital-adequacy issue according to S&P’s commentary and analysis,” he said. “Any public speculation or rumor to the contrary is at best misguided or at worst an intentional effort to negatively influence our stock price.”

The bond insurer’s clients were being “fed fiction” about the company, Roseman told listeners. The company faced no liquidity issues, at least not at its current rating level. And there was no reason to believe its rating would be cut.

“ACA is principally paid to absorb market price volatility,” he explained. “We assume the positive and the negative over the life of the transaction, thereby providing a market hedge for our counterparties.” The company was “absorbing” losses of $1.7 billion for its counterparties. But that was only from an accounting standpoint. ACA Capital didn’t expect to actually pay any claims on the CDOs. In fact, the company had only $425 million in capital.

If ACA Capital was downgraded to triple B, however, it would be required not only to “absorb” those losses on its income statement but also to terminate the contracts and pay its counterparties in cash for the decline in the market value of the CDOs.

An analyst on the call asked Roseman to run through the downgrade scenario just to be sure he understood the implications: “So right now, you’d have to meet about $1.7 billion, or whatever the number would be,” the analyst said. “Something is wrong here.”

But at 4:25 p.m. that afternoon, S&P placed ACA Capital’s A rating under review for a possible downgrade. There was no way the bond insurer could meet termination payments if it was downgraded, JPMorgan Chase & Company analyst Andrew Wessel told Bloomberg TV. “ACA is a likely candidate to get thrown to the wolves first,” Wessel said.

JUST WHEN IT SEEMED that no sensible person could believe in the bond-insurance business anymore, Warren Buffett tossed his hat into the ring. The Wall Street Journal reported that Berkshire Hathaway might provide capital to the bond-insurance industry. On November 13, 2007, the day the story ran, MBIA shares surged on the prospect of a capital infusion, though there was no indication Buffett was considering investing in the company. Investors hoped that Buffett—who coined the adage “Be fearful when others are greedy and be greedy when others are fearful”—believed the worst was over for the bond insurers.

The industry remained on the defensive. At a Bank of America conference in New York on November 27, 2007, executives from the bond-insurance companies sought to assure investors. “We’re going to defend the triple-A credit rating,” Ambac’s chief financial officer Sean Leonard told the assembled group.

“Job No. 1 is to show that we have the stability around our triple-A ratings that investors expect,” said Edward Hubbard, president of XL Capital Assurance, the bond-insurance unit of a company called Security Capital Assurance.

Chuck Chaplin, MBIA’s chief financial officer, reminded the group that all the bond insurers, also referred to as monolines because they insured only debt, were in this together. “If any major monoline were to have a rating change, it would have a real impact on all the business of the monolines,” Chaplin said.

That’s what made Domenic Frederico’s comments so shocking. Frederico, the CEO of Assured Guaranty, explained why his company had not followed its competitors into the business of guaranteeing CDOs backed by subprime mortgages.

As homeowners default on their mortgages and the underlying properties are sold for less than the amount of the mortgage, losses begin to build up in the pools of loans backing bonds. If a bond insurer guaranteed a security with a 30 percent subordination cushion beneath it, then the insurer would start to pay claims once losses on the pool hit 30 percent. Every percentage point increase of loss beyond 30 percent would translate into a dollar-for-dollar loss for the bond insurer. For a bond insurer to lose 100 percent on a guarantee, every mortgage would have to default and every foreclosed house would have to be sold at a complete loss. That scenario was unimaginable.

CDOs can create a more devastating outcome once bond insurers begin to pay claims. The insurer might start paying claims on a CDO guarantee once losses in underlying mortgage pools hit 30 percent, just as it would on the straight mortgage-backed bonds. But the insurer may find itself paying off 100 percent of the value of the entire CDO it has insured by the time underlying pool losses reach just 35 percent.

The slicing, dicing, and repackaging of securities can concentrate risk. Mortgage-backed bonds are supported by payments on thousands of mortgages; everyone has to default before there’s a complete loss. CDOs, however, rely on the cash flow from hundreds of mortgage-backed securities. Recall that mortgage-backed securities have different ratings depending on where they reside in the payment waterfall. Lower-rated securities are completely shut off from the cash flow as losses increase. If all the securities backing a CDO are rated triple-B, they may be completely starved of cash as losses on underlying mortgage pools move from 30 percent to 35 percent. That means the entire CDO is cut off from cash even though most people whose mortgages are in the pool continue to make their payments.

The speed with which losses can burn through a mortgage-backed bond versus a CDO is akin to the difference between a fireplace log burning over the course of an evening and a newspaper incinerating within seconds. “We told Moody’s that ABS CDOs don’t meet our credit criteria,” Sabra Purtill, head of investor relations at Assured Guaranty, once related to me. “We told them ABS CDOs are dangerous. They never asked why. It was like we were talking to a tree.”

For the first time, Ackman had hired security guards to accompany him to a presentation. Many people believed Ackman was fanning a dangerous fire with his criticism of the bond insurers. During a Barclay’s Capital conference call on the financial guarantors earlier in the month, one caller prefaced his question by stating that he believed those writing and speaking negatively about the bond insurers were “financial terrorists.”

For years Ackman had asked regulators, reporters, and just about everyone he spoke to about MBIA to see no contradiction in shorting a company and being a decent person. It is an idea that many people simply can’t accept.

When Ackman had testified at the Securities and Exchange Commission in 2003, this issue of doing good by short selling had come up. Ackman had said it was one reason he decided to short Farmer Mac. “I prefer investments where I’m not fighting against the country. You know, where there’s public policy on my side instead of against me,” Ackman had told the investigators.

Now Ackman asked the packed auditorium of investors to believe that he had devised a plan that would both hasten the collapse of a company he was betting against and do some public good in the process. Not many people would try to pull that off, but Ackman had his unrepentant idealism.

Ackman’s 146-page presentation was titled “How to Save the Bond Insurers,” and it laid out his plan for conserving capital within the insurance subsidiaries of MBIA and Ambac. Holders of insured bonds and the Wall Street firms that had sold credit-default swaps to MBIA would benefit from the additional capital, Ackman said. “The holding companies are the problem,” he told the audience. The holding companies wanted to take as much capital as possible out of the regulated insurance subsidiaries, but the policyholders were going to need that capital. Take back improper dividends, said Ackman. Void the inter-company guarantees that allow MBIA to run what is effectively a fixed-income hedge fund. Void the transactions done using credit-default swaps. Possibly even void all the deals backed by home-equity loans since bond insurers weren’t permitted to guarantee mortgages. The plan was fatal to the holding companies.

Ackman also announced that—at least as far as his personal profits were concerned—he was now shorting the bond insurers for charity. “We are going to make hundreds of millions of dollars on the failure of the holding companies,” Ackman concluded. “So I’m pledging to give my share of the profits to the Pershing Square Foundation.” Ackman had set up the foundation to provide funding for education and other charitable causes.

“He may be aggressive, he may be over the top, he may not be able to speak in short sentences,” New York Times columnist Joe Nocera concluded following the presentation. “But he’s doing the hard work, and thinking the hard thoughts that they refused to do for so long.”

On Sunday evening, just after 11 p.m., Roy Katzovicz got an e-mail from Ackman. He’d been copied on a long message that Ackman sent to the Securities and Exchange Commission. Katzovicz read the subject line—“The SEC’s Enforcement Failures with the Bond Insurers and Other Issues”—and scanned the addressees in the “To:” column: SEC Chairman Christopher Cox; SEC Commissioners Annette Nazareth, Paul Atkins, and Kathleen Casey; the SEC’s director of enforcement, Linda Thomsen; Brian Cartwright, the SEC’s general counsel; John Nester, the SEC’s director of communications; and Mark Schonfeld, director of the SEC’s New York regional office.

Without reading any further, Katzovicz told his wife, “I may have to quit my job tomorrow.” His boss’s habit of writing long, emotional, late-night missives without having him vet them was one of the aggravations of Katzovicz’s job. But this one was the worst yet. There was no point now in reading the message and spending the entire night tossing and turning over all the points that might have been made with less fractious wording or—better yet—deleted altogether before the message was sent. He let it wait till morning.

When he got up at 5 a.m., Katzovicz opened the message. “The SEC has completely fallen on its face in stepping in to prevent the upcoming mushroom cloud that will envelop the bond insurers,” the letter began. Ackman added that he had pointed out the problems with the bond insurers years ago in the Gotham report. “I addressed CDO and derivative mark-to-market issues, the illegality of bond insurers entering into derivative transactions, their inadequate reserves, their aggressive and fraudulent accounting, and a whole host of other issues,” Ackman wrote. “The SEC, as you well know, first spent time investigating me and my then firm, Gotham Partners, for writing the original report on MBIA, rather than focusing on the issues I raised in the report.”

Even though the SEC exonerated Ackman of all wrongdoing and invited him to make multiple presentations to regulators on the industry, “the SEC still has not had the courtesy to give me a letter for my files about the termination of the Gotham investigation without any finding of wrongdoing, despite repeated requests from my counsel for the same,” Ackman wrote.

The letter was also addressed to the three New York SEC staff attorneys with whom Ackman had met over the years: Alan Kahn, Steve Rawlings, and Chris Mele. But the real barbs were directed to those higher up at the SEC. “My sense is that it’s more likely that the responsibility for the lack of pursuit of these issues lies at the highest levels of the SEC including Chairman Cox,” Ackman wrote.

He concluded by offering to meet with the SEC again to further explain how losses might yet be prevented for policyholders. “It is an incredible embarrassment to the SEC and to the quality of our country’s regulatory oversight for you to have dropped the ball here,” Ackman concluded.

When he got into the office that morning, Katzovicz walked into Ackman’s office to confront him about the e-mail. “Why shouldn’t I just quit?” Katzovicz demanded. “This is the kind of thing you send off to our principal regulator without running it by me?”

Katzovicz said he would not have stopped him, but he might have suggested some changes, some omissions. Perhaps the letter wouldn’t have been such a personal attack on Cox. Sometimes Ackman needed to be protected from himself, and that had to be a part of Katzovicz’s job. If Katzovicz was going to stay with the firm, they needed a new rule. Katzovicz spelled it out: “You cannot kick our chief regulator in the nuts without consulting with me first.” Ackman agreed, and Katzovicz stayed.

Within a week, a junior staff attorney at the SEC called Pershing’s outside counsel to find out where to send a notice of discontinuance. The letter arrived on Ackman’s desk a few weeks later. It stated, “This investigation has been completed as to Gotham Partners Management Co., LLC and William A. Ackman, against whom we do not intend to recommend any enforcement action by the Commission.”