Dear Mr. Buffett: What an Investor Learns 1,269 Miles From Wall Street

by Janet M. Tavakoli

  Yet, despite increasing attention in the financial press, the New York Attorney General’s office dropped its case. The SEC’s rumored civil enforcement action involving Bear Stearns’ CDO pricing practices fizzled, and the investigation was closed.3 The Slumbering Esquires Club rolled over and went back to sleep.

  The SEC’s new head struck me as the Antichrist of investor advocacy. On July 26, 2005, just a few days before Goldstein’s article and my first reply to Warren Buffett, Christopher Cox attended a Congressional coffee klatch—commonly known as his confirmation hearing—for the post of chairman of the SEC. One of Cox’s former clients pleaded guilty and got a 10-year sentence in a case involving defrauded funds.4 Cox had worked on a separate public offering that was not implicated in the case. Among other things, Cox wrote a letter for his client saying it “would unfairly and unreasonably harm the investors’ rate of return”5 to appraise pools of mortgages. Cox also wrote that suitability—a standard meant to ensure that naïve investors did not get saddled with risky product—should not apply: “Because all of the trust fund loans are secured and overcollateralized, there is relatively low risk.”6 Cox wrote his letter in 1985, and unbeknownst to Cox, his client’s fraud involving separate vehicles began in 1982 and continued until 1994. Appraisals may have stopped the fraud earlier. On July 29, 2005, Floyd Norris of the New York Times reminded his readers of Cox’s letter, yet the Senate Committee unanimously confirmed Cox later the same day. It was time to short CDOs, since the value added by the SEC’s Cox seemed likely to shrivel.

  When I read prospectuses for CDO deals and CDO-squared deals, I felt as if I had opened a box of candies and found only one or two good pieces.The rest were either missing altogether, or had a bite taken out of them with someone else’s teeth marks. These were definitely not See’s Candies, a Berkshire Hathaway company. Some CDO-squared deals were so bad it left me thinking: Where is the candy?!

  To use an extreme example, if you only use subprime-backed fraud-ridden mortgage loans as collateral for residential mortgage-backed security deals, and the RMBSs lose 60 percent of portfolio value, if you use investment grade tranches but with ratings lower than the top AAA of these RMBSs as collateral for a CDO, all of the collateral of your CDO will vaporize. If you use tranches of this defective CDO in yet another CDO called a CDO-squared, you are starting out with nearly worthless collateral, so the entire CDO-squared is nearly worthless on the day the deal is brought to market. It seems to me that some investment banks knowingly participated in predatory securitizations.

  One does not need to read hundreds of pages of prospectuses or perform complicated modeling to know that. Warren looks at every investment as if it is a business, and the only “business” these investments have are the loans backing them. If the loans do not do well, the CDOs backed by them soon follow them down the tubes.

  It will be too obvious if all of the collateral you use is this bad, so you might mix it in with some Alt-A or even some prime collateral in an RMBS. That way, if you use this collateral for a CDO, it won’t look so bad, and it will be devilishly difficult to analyze. For example, if you use BBB rated tranches of RMBS deals backed by a variety of types of loans, you can mix in 30 percent risky subprime loans. It sounds pretty safe, but losses will probably still eat through the BBB rated tranches. Now you take those doomed BBB rated tranches and combine them with A and AA rated tranches to create a CDO. All of the BBB rated tranches will disappear and probably some or all of the single A. If you buy the AAA tranche of this CDO, and it has around 25 percent subordination, your principal may or may not be in jeopardy, but most of the tranches below it are in trouble. Now if you use those lower tranches to make a CDO-squared, most of those tranches will probably lose principal. In some deals, all of the tranches below the senior-most triple A will lose the entire principal amount, and the senior-most triple A will lose substantial principal.

  Credit derivatives enable a further level of gamesmanship and opacity. The documentation of many CDOs is dense with all sorts of cash flow tricks, and the contracts for the credit derivatives embedded in the CDOs are not included with the prospectuses. The ratings are completely meaningless.

  In January 2007, I noticed that U.S. institutional investors curtailed their buying of CDOs. But investment banks had created new types of structured investment vehicles called SIV-lites, or structured investment vehicles with less protection (or lite protection). These vehicles invested in the overrated AAA tranches of CDOs backed by subprime debt, and the rating agencies rated the vehicles AAA. These vehicles, in turn, issued faux AAA asset-backed commercial paper.

  These new entities seemed like corporations, but the only “business” they have is investing in assets and those assets have to provide “earnings.” Benjamin Graham’s disciples look for better quality of earnings and for earnings growth.

  As the collateral in the structured investment vehicles inevitably took massive downgrades, the vehicles had to liquidate their wasting collateral, and investors lost a significant amount of their principal. Mutual funds, bank portfolios, insurance companies, local government funds, private investment groups, and more lost billions. Canadians heavily invested, and our North American neighbors lost billions. Since these assets carried high ratings, European and Asian investors also took losses.

  Despite their “efforts,” investment banks were still stuck with tens of billions of unsold CDOs. They reduced exposures by buying bond insurance, buying credit protection from hedge funds, and doing a variety of leveraged sales. Some of that risk boomeranged back onto bank balance sheets.

  The madness did not stop with mortgage loans. Collateralized debt obligations can be backed by any combination of debt: credit derivatives, asset-backed securities, mortgage-backed securities, other collateralized debt obligations, hedge fund loans, credit card loans, auto loans, bonds, leveraged corporate loans, sovereign debt, or any kind of combination of actual or notional debt man can imagine and create.

  Stephen Partridge-Hicks, co-head of Gordian Knot, probably the best run structured investment vehicles in the world, felt the effects of a nervous market reluctant to invest in the debt of any investment vehicle. Risky overrated AAA commercial paper issued by risky structured investment vehicles caused investors to shun sound investments. He told me he bought zero subprime-backed investments and rejected a lot of other misrated AAA deals.Yet shortly after Lehman’s bankruptcy, Sigma, one of his two funds, collapsed.

  If I had a large bonus in my sights and mischief on my mind, how would I unload toxic CDO tranches? This is all hypothetical, mind you, but here’s just one of a number of different gimmicks.

  If you work at an investment bank and you stuff the toxic tranches of only your own CDOs into another CDO, it will be too obvious.You need help from your friends who work for other investment banks, hedge funds, and CDO managers. Since you all have toxic CDOs and still want to earn high fees, you can all play investment banking hawala similar to the complex, but highly effective, money brokering system used in the Middle East. Hawala makes it virtually impossible to trace cross-border money flows. It will be hard for anyone, except the SEC or someone with subpoena power to examine your trade tickets, to figure out what you are doing. Since the SEC seems to have lost its will to exist, you are good to go.

  There is just one more thing. As Warren told me at lunch, many people seem to have a perverse desire to make things overly complicated. Yet, the fundamentals of finance do not change. Most value investors will not be fooled, and they actually read your documents. If you really think you can confuse unwary investors about the basics by hiding behind a label such as “synthetic CDO-squared,” you are good to go.

  Mix your toxic junk with your friends’ toxic junk into a CDO-SQUARED. Now you have deniability. After all, why would you buy someone else’s CDOs if they were toxic? Now get the compliant rating agencies to rate a huge chunk of this risky hairball triple A. If you are lucky, you may find an investor to buy it. Failing that, you may find a bon
d insurer to insure it. Failing that, you may find an investment vehicle or hedge fund willing to do a credit derivative or other leveraged transaction.These diversions should get you through bonus season. After all else fails, your investment banks can beg the Federal Reserve Bank to take overrated AAA paper in exchange for treasuries.

  There is one small problem with this. If you know or should know that you are not correctly pricing your balance sheet or if you knowingly sell overrated securities, you must disclose that, and you must be specific about it. If you know something is rated super-safe AAA; but it deserves a near-default rating of CCC, you cannot keep silent about it when you sell it.

  When I pointed out to an investment banker that this is a classic situation for fraud, he told me: “Our internal OGC [Office of General Counsel] disclaims virtually all liability for [our investment bank] and its bankers in small print fully disclosing the risks in the prospectuses.” I knew what he meant, but he sounded like a smart 10-year-old parroting an adult.

  “I did not attend law school,” I responded, “but I am pretty damn sure that just because you disclosed serious conflicts of interest, it does not protect you if you fail in your duty of care to investors.Your lawyers can’t give you a license to kill.”

  The moral hazard swamped any risk the rating agencies’ models could capture. One investment banker crowed to me that the rating agencies are eager for fees and the investment bank’s structurers seeking ratings for their CDOs are “shrewd bullies.”

  One synthetic CDO deal with a notional amount of more than $2 billion went into liquidation, and less than 3 percent of investors’ money was recovered. Even the investor in the top-most “AAA,” the super senior tranche, lost principal. Perhaps everyone involved with this deal, including the CDO manager, was just very unlucky. But do you want to do business with unlucky people?7

  CDO managers are supposed to be selling securities backed by actual assets—not imaginary assets.

  In November 2006, I told Asset Securitization that CDO managers are unregulated, and only a handful of managers provide good value for the fees charged. Most do not have the expertise or resources to perform CDO management or surveillance. Many cannot build a CDO model. Many managers rely on the bank arranger both for structuring expertise and to take a lead role with the rating agencies to secure the initial ratings. Rating agencies rarely ask for background checks on CDO managers.

  Chris Ricciardi, CEO of Cohen & Company, read my commentary and wrote me: “I LOVED it.” He had been thinking about how to be “the best CDO manager in the business,” had independently come to the same conclusions, and found my “insight very compelling.” Yet in April 2008, Cohen & Company’s CDO management arm, Strategos Capital Management, led managers with CDOs in default. The total original amount of the CDOs it managed that had events of default (with as yet undetermined recoveries) was $14.2 billion.8

  On December 7, 2007, I wrote Warren that many asset-backed securitization CDO prospectuses are finance comic books. For example, Adams Square Funding I closed December 15, 2006. It was an “asset-backed” deal, a collateralized deal. It was rated by Moody’s and S&P. Yet, before 2008 ended, the CDO unwound, meaning all of the underlying assets were sold in an attempt to pay investors back. Unfortunately, there was not enough cash after selling the loans to go around. According to S&P, investors in Adams Square Funding I got less than 25 percent of par value—more than a 75 percent loss—on average. Investors were wiped out, except for the investor in the seniormost AAA tranche.9 Since the prospectus shows that the seniormost tranche made up 29 percent of the deal, it appears those investors may have lost some money, too. It is reminiscent of the opening scenes of the movie Cliffhanger, in which a climber’s supports snap one-by-one ending in a spectacular steep fall. That last plastic buckle was AAA rated. Adams Square Funding I is not an isolated example, just a handy one, because it unwound. It is not even close to being the funkiest deal I have seen.

  Warren’s ability to say “no” when the risk is not priced correctly is a tremendous advantage to any investors.

  The prospectus for Adams Square Funding I disclosed the conflicts of interest between the investors, Credit Suisse Alternative Capital (CSAC), and other Credit Suisse affiliated entities, including the Leveraged Investment Group (LIG) of Credit Suisse Securities (CSS).10 I always recommend that investors eliminate this kind of moral hazard by insisting on changes to the deal. Conflicts of interest do not mean that there is anyone doing anything wrong, but when the moral hazard is enormous, things never seem to end well for investors. Rating agency models do not capture these huge risks, yet, the rating agencies never seem to refuse to rate these deals. I have written books and articles on this problem for years; the ratings on deals with this kind of risk are totally meaningless. Yet the rating agencies continue to defend their indefensible methods.

  Among many classes of bad deals, the problems of CDOs named after constellations were well publicized. Approximately $35 billion of these CDOs had been underwritten by Citigroup, UBS, Merrill Lynch, Calyon, Lehman, and others.They are mostly fallen stars. I told the Wall Street Journal that Norma, a Merrill-underwritten CDO comprised mostly of credit derivatives linked to BBB rated tranches of other securitizations, “is a tangled hairball of risk.”11 It had come to market in March 2007, and by December 2007, it was worth a fraction of its original value. The rating agencies slashed its ratings to junk. I added “[A]ny savvy investor would have thrown this…in the trash bin.”12

  Constellation deals were not the only class of dicey deals, and it seems that CDOs bought in the last half of 2006 and during 2007 were particularly awful. Investment banks found they had a huge credibility problem with investors. Merrill Lynch was not alone in having credibility problems, but I happened to review all of their 2007 CDOs that I could track. I looked at 30 CDOs and CDO-squared deals with a notional amount of $32 billion that Merrill Lynch underwrote in 2007. As of June 10, 2008, all of the deals I captured were in trouble at the AAA level. One or more of the originally AAA rated tranches had been downgraded to junk (below investment grade) by one or more rating agencies. Merrill Lynch was not alone in having a poor track record, but this sort of unprecedented performance was hard to beat. CDO managers had nothing to be proud of, either, and many saw their streams of fee income dwindle.The securitization market was in a dead calm.13 I made my concerns public.1415 As far as I was concerned, the Hall of Shame looked overcrowded. Losing trust was not the only problem. Financial institutions lost hundreds of billions of dollars.

  Bloomberg keeps daily tabulations of subprime related losses worldwide. I told Yalman Onaran that although some mark-to-market losses may be reversed as markets recover, most of the losses are permanent impairments caused by surging defaults: “[O]f course we can’t tell how much . . . may actually be good stuff that will pay back at maturity.”16

  By June 18, 2008, Bloomberg estimated that global bank balance sheet losses due to write-downs and charge-offs at $396 billion. That figure may have been tainted with denial. By October 16, 2008, it nudged past $660 billion. Citigroup had written down $55.1 billion, Merrill Lynch $58.1 billion, and UBS $44.2 billion. Wachovia topped the list with losses of $96.7 billion; Washington Mutual’s losses were $45.6 billion. The list was long and sobering.1718 Risky loans made to both risky borrowers and prime (high credit score) borrowers were only part of the problem. Predatory securitizations amplified losses. As a result, the entire landscape of global investment banking changed.

  The damage to the global markets was much worse, however. Losses reported by the banks do not include losses to hedge funds, private equity investors, mutual funds, municipalities, insurance companies, pension funds, and more.The International Monetary Fund (IMF) estimates losses related to the U.S. subprime meltdown and its fallout could reach about $1 trillion.19

  I blurted out to Warren that I was disgusted with the “douche bags who got [the nation] into this mess”; then I gasped at the realization of what I had just said. For
his part, Warren says that the documentation uses arcane language and that it is impossible to read that many prospectuses just to analyze one deal. One had to read “hundreds of thousands of pages.”20 Warren once noted: “There seems to be some perverse human characteristic that likes to make easy things difficult.”21 The simple solution boils down to the principles that Warren has espoused for decades: Don’t lend money to people who cannot pay you back. If you do not understand something, do not invest.

  Chapter 7

  Financial Astrology—AAA Falling Stars

  I can’t recall we’ve ever asked for management changes in companies we’ve invested in. If they did the wrong thing, they should go.

  —Warren Buffett,

  Wall Street Journal, May 23, 2008

  At the end of 2007, Berkshire Hathaway owned 48 million shares of Moody’s Corporation, one of the top three rating agencies (the same shares Berkshire owned when I first met Warren Buffett in 2005), representing just over 19 percent of the capital stock.The cost basis of the shares is $499 million. At the end of 2002, the value was just under $1 billion. By the end of 2006, the value was around $3.3 billion, but it dropped to $1.7 billion at the end of 2007.1 The sharp increase in revenues is due chiefly to revenues generated from rating structured financial products, and the sharp decrease was due to the disillusionment of the market with the integrity of the ratings.

  The collateralized debt obligation market grew from around $275 billion in 2000, to about $2 trillion in 2007; then the market stalled. By June 11, 2008, Total Securitization reported that CDOs in default exceeded $200 billion.2 Investors included insurance companies, bank investment portfolios, mutual funds, pension funds, hedge funds, money mangers, and more. Every sector of society is affected as misrated products cause actual principal losses combined with loss in value due to declining market prices and illiquidity. More than that, liquidity—coming up with needed cash—is now a global problem, since investors are wary of lending money (by investing) against potentially misrated assets.