The New Robber Barons Page 21

  The Fed uses tax dollars to keep some of our largest banks—weakened by reverse-Glass-Steagall mergers with troubled entities—from collapsing under heavy loan losses. U.S. taxpayers became unwilling unsophisticated investors funding Wall Street’s bailout.

  In the wake of the global financial crisis, Bloomberg estimated that by March 21, 2009, pledges by the Federal Reserve, Treasury Department, Federal Deposit Insurance Corporation topped $12.8 trillion. The U.S. GDP was $14.2 trillion. At the time the pledges represented more than $42,000 for every man, woman and child in the U.S.

  Massive hiding-in-plain-sight fraud damaged the U.S. economy. Housing prices didn’t just fall; they plummeted as the fraud unraveled. By the end of 2010, four million home loans were more than 50% underwater with an average of $107,000 negative equity; this alone is around $428 billion in negative equity. 7.8 million loans were 25% or more underwater. Most of these damaged borrowers were paying higher interest rates than the national average, couldn’t refinance, and were ineligible for HAMP, a government initiated refinancing program.

  Securitized assets were dumped on the balance sheets of the Fed (and the Fed supplied near zero cost funding to plug holes in bank balance sheets and special government guarantees on bank debt), special purpose entities set up by the Fed, Fannie Mae, Freddie Mac, and the FHA (through government guarantees). Government guarantees allowed for new issuance of securitized mortgage loans that otherwise would have no buyers, since the rating agencies are widely mistrusted.

  The effect of existing government entitlements and obligations combined with expanding bank debt, flawed loans, flawed securitizations, and leverage significantly weakened the U.S. economy stunting growth as money was diverted to the banks and away from capital spending, the biggest driver of any economy.

  The damage of financial malfeasance was so extensive and the protocol of the bailout so profligate that the “AAA” rating of the United States has been rendered an incorrect but politically expedient label. The U.S. issues its debt in dollars, and if necessary can engage in a silent default by devaluing the U.S. dollar. To some extent, that has already happened.

  The U.S. dollar is still the world’s reserve currency. It is also an “intervention” currency, and an “invoice” currency for much international trade. Central banks hold dollars as a reserve, and private agents hold dollar denominated investments. So far, the U.S. has the most well developed short-term financial markets, so when there is a “flight to quality,” which these days means the lessor of other evils, foreigners rush to liquid financial instruments like T-Bills.

  Unlike the Yuan, the U.S. dollar is freely exchangeable in most of the world. The U.S. buys its major imports such as oil in dollars. But change is already in process. Iran has said it will accept Yuan for oil. There is talk of oil prices being expressed in a basket of currencies that will include gold. Gold is already being accepted as collateral on derivatives exchanges, and in this sense, gold is now recognized again as money for trading purposes.

  The problem for the U.S. is that despite the temporary advantages, the world is diversifying into other currencies, including gold. The Yen and Deutsche Mark (before the Euro) were the former candidates for diversification. Future candidates may be the Yuan and perhaps a new Euro (if Germany prevails in getting its way). That will push up the value of those currencies and push down the value of the dollar.

  Even if this doesn’t happen right away, the U.S. is in a weakened state and it is not immune to shocks. Recall the 1973 oil embargo, for example. Oil prices quadrupled overnight and the dollar fell from 350 yen to 250 and from 3.25 Swiss francs to 1.80. Two hour moves eclipsed the previous two decades. Then there was the 1995 U.S. dollar crisis brought about by the “tequila effect” when one of our debtors, Mexico, neared default and had to be bailed out.

  When foreigners lose confidence in the United States, they don’t suddenly sell all of their assets or pull out all of their money (although some of that does happen), the immediate effect is a fast fall in the dollar. The Fed has compensated for the explosion in new debt by purchasing government bonds, but this strategy may not be sustainable, and if not, there may be a swift rise in interest rates.

  The current “AAA” rating of the United States is not on merit, but it is a convenient fiction for the global financial markets, because no one yet has an alternative. It would be more accurate to say that the United States remains investment grade, because of our current role in the infrastructure of the global economy. Whether or not the rating agencies put an explicit downgrade on the United States is largely irrelevant. The United States has seen huge U.S. dollar fluctuations and higher interest rates in the past, and on its current course will see them again. The fundamentals of the U.S.’s increasing debt load, massive future entitlements, high unemployment rate, and weak economy with no meaningful growth plan speak for themselves.

  VIII.RESTORING CREDIBILITY: WHAT REGULATORS MISSED AND WHAT THEY GOT RIGHT

  The U.S. based rating agencies, Moody’s and S&P are subject to Congressional legislation and ostensibly regulated by the SEC. The Dodd-Frank Act attempted to address issues with the rating agencies, and it made several good suggestions, but missed key problems. Japan has a set of rules for the rating agencies, and Europe is developing a separate set of rules. Ratings are used globally, and it seems rating companies should be regulated by an international body.

  A. Create Objective Benchmarks and Rating Scale: Revoke the Raters’ Ability to Legislate

  Neither U.S. nor European regulators address the problem of rating agencies’ ability to make up their benchmarks and rating scales and change them at will. The discussion below has been further developed from the work of Dr. Arturo Cifuentes, a former Moody’s rating methodology developer, and Jose Miguel Cruz in their “White Paper on Rating Agency Reform,” issued from the department of industrial engineering at the University of Chile in May, 2010.

  Moody’s and S&P each have a nine category scale with 21 notches. (See III. A. And APPENDIX VII. Note that “D” is not counted in the nine category scale, since it denotes default.) There is no reason to consider the ratings as having equivalent meanings. Moody’s and S&P base their scales on different concepts, expected loss and probability of default, respectively. The rating agencies employ different computational methods and different models. They each use their own data sets.

  It is more than a little interesting to note, that the ratings are often similar, even when they are egregiously wrong, and even though they supposedly employ different theoretical frameworks. For example, both Moody’s and S&P awarded “AAA” ratings to CPDOs, a product developed in late 2006 that merited a non-investment grade rating, i.e., a junk rating and quickly fell apart. (See Section V.) It begs the question of whether the rating agencies are actually independent or whether they are being consistent with each other to maximize revenues.

  Solution: A third party (the SEC, for example, unless it is replaced as the regulator, and it should be) should define the benchmarks and the rating scale. The rating agencies would no longer be free to choose benchmarks and effectively redefine the meaning of ratings. Rating agencies would only be entitled to issue ratings, i.e., determine where the creditworthiness fits in relative to the pre-defined benchmarks that correspond to the rating scale.

  If a combination of levels of probability of default and loss given default (from which recovery value is determined) is used to create a rating scale—an appealing prospect—then two supplementary rating scales are in order. One scale would reflect the probability of default and one scale would show the loss given default.

  Said another way, the ratings scale and the cutoffs would be objective and the same cutoffs and scales would be used by all rating agencies. The rating agencies would have no ability to change the cut-off values.

  Implementation would require some global buy-in, and a comment period of say, three months, from stakeholders. Legacy ratings would be mapped to the new scale with the proviso that the m
apping might be flawed due both to the flaws inherent in the original ratings process, and the fact that the theoretical approaches were originally different.

  B. Solution: Eliminate “Mezzanine “AAA” Tranches and “Super Senior” Tranches

  Neither U.S. nor European regulators address the problem of “first loss AAA” tranches and “super senior” tranches. Many “AAA” tranches had insufficient subordination to ever merit that rating based on the quality of the assets in reference portfolios. But even solving the problem of quality of assets in the portfolio does not solve the conceptual problem of “super seniors” being somehow superior to what was formerly a genuine “AAA.” The definition of “super senior” and the benchmark requirements to merit the designation were left undefined by the entire industry.

  Super seniors were meant to be super safe, and even BIS enabled them by awarding favorable regulatory capital treatment to them to accommodate banks. In fact, “super seniors” are a gimmick created to achieve favorable regulatory capital treatment.

  Not only did “super seniors” make a mockery of the “AAA” label, they did not live up to their name. Many investors in “super seniors” lost almost all of their initial investment when they invested in late vintage CDOs. Investment banks and banks that retained “super seniors” on their balance sheets and “insured” them to try to obtain income via a flawed regulatory capital and economic “arbitrage” ended up with heavy losses. (See also Section III. A. and APPENDIX VIII.)

  C. Focus Is On Wrong Metrics with Uninformed Guesses as Model Inputs

  Neither U.S. nor European regulators address the rating agencies’ apparent incompetence. For example, the heavy reliance on correlation models is the wrong way to go about determining whether any benchmark meets the requirement of a rating when the most important metrics are probability of default and loss given default (or said another way, recovery value in the event of default). It would have been more productive to spend all of the time and effort determining probability of default and recovery value and none of it on correlation.

  In fact, the exercise of understanding the granular risk well enough to determine probability of default and recovery value gives one a much better understanding of correlation and how correlation among assets changes and converges under different scenarios. Once one understands that, one is keen to abandon the unstable correlation metric in favor of the granular approach. This may be part of the reason my analyses prove superior to the rating agencies time and again. Another reason may be conflicts of interest within the rating agencies themselves. Revenues were more important than ratings accuracy. (See Section VIII. E.)

  Rating agencies adapted the Gaussian Copula model and black box correlation models for structured financial products. Correlation models try to determine if a portfolio’s assets will behave similarly when a reference asset strengthens or weakens. The models are highly unstable. They are like a chair that collapses beneath you as soon as you sit on it. Small changes to model inputs result in huge changes to the results. (See article following the letter in Appendix I.)

  If you play with coins or dice, you know exactly what your inputs are and you can model all potential outcomes. You can examine the coins (heads or tails per coin), and you can model all of the possible outcomes. You can examine dice (one to six dots on each face of each cube), and again, a mathematical model can describe all potential outcomes. We may use a Monte Carlo model (or other model) to randomize the inputs (the flips and tosses), but we do not have to guess at the inputs for dice and cards; they are known in advance and the relationship between the inputs does not change.

  Unlike cards and dice, the inputs to credit models are not exact representations of reality, they are estimates of variables. The inputs to credit models are a guess, but they should be educated guesses—not the complete stabs in the dark used by the rating agencies.

  We rely on data approximations to come up with the inputs in the first place, and the relationships between the inputs can change. For example, most of the data describing how one corporation behaves in relationship to another is based on market prices such as stock prices or the prices of credit default swaps based on corporate debt. Moreover, there is very little of this already suspect data to work with. The results are guesses about relative price or yield spread movements, which result in a guess about the correlations. When a credit upset occurs in a financial sector, correlations that were previously fractional numbers tend to converge to one. Everything seems to fall apart at once. A model will calculate the wrong answer to nine decimal places, but it cannot tell you it is the wrong answer.

  The biggest problem with correlation models is that even if they temporarily get the correct answer, they do not tell you what you need to know. The models estimate asset correlations instead of the necessary default correlations. Furthermore, the overwhelming flaw in the methodology is that if you want to make up a default correlation between two companies, you must make the false assumption that default probability does not vary, but of course it does.

  Even if the models measured the default probability of individual companies—and they do not—if a company defaults, you still have to guess the recovery rate, the amount left over, if any, after all obligations are paid.

  You cannot solve for two independent items of information from a single piece of information such as a letter grade or a price. You cannot get both the probability that a company will default and the amount of money you will have left if it does default.

  Solution: Modelers must understand the granular credit risks in a portfolio. There is nothing wrong with employing a shortcut like a correlation model (even though there are better techniques), but a correlation model is not a substitute for examining due diligence to understand the character of the granular risk in a portfolio. For example, new risky mortgage loan products with corrupt underwriting standards failed in the same way no matter where they were located, and recovery values were much lower than the assumptions used by the rating agencies. This was knowable in advance. Mortgage loan originators had a known history of fraud. (Examples include FAMCO, Ameriquest, manufactured housing loan originators.) Yet in the recent debacle, rating agencies’ correlation assumptions were grossly incorrect at the outset.

  D. Rating Agencies Must Review Evidence of Underwriters’ Due Diligence

  Both the Dodd-Frank Act and proposals by European regulators call for more transparency about underlying collateral, but don’t hold rating agencies explicitly accountable for demanding thorough due diligence from underwriters, especially in the face of red flags that suggest third party fraud audits may be in order. A fraud audit doesn’t mean anyone is being accused of fraud, only that the audit will be thorough enough to uncover it if it exists. In the case of mortgage lenders and the underwriters themselves, one would have found the reports of Clayton Holdings that suggested that the securitizations created by underwriters were invalid as many loans representations and warranties had been breached, and the documentation of loan transfers seemed missing. Beyond that, the credit quality of the loans was so poor that the ratings were invalid.

  The rating agencies are swift to point out that they do not perform due diligence on the data they use and take no responsibility for unearthing fraud; they merely provide an opinion. In past legal battles, rating agencies successfully claimed journalist-like privileges, refused to turn over notes of their analyses, and continued to issue opinions. Independent organizations exist, however, that perform rigorous due diligence for a fee. Underwriters can hire them, and rating agencies can demand to see the results. These reports include loan reviews, and audits ranging in thoroughness from a cursory review to a fraud audit depending on the circumstances.

  The consequences of the unwholesome mix of lax underwriting standards and new risky loan products combined with varying degrees of overreaching, predatory lending, real estate speculation and outright fraud, were discoverable with reasonable due diligence. The egregious overrating of securitized products
was completely preventable in the course of competent analysis.

  The rating agencies could have—and should have—asked issuers and investment bank underwriters to demonstrate that they had performed a statistical sampling and verification of underlying loans in the proposed portfolios. This includes independent checks of appraisals, checks that mortgages have not been mis-sold, background checks of mortgage brokers, background checks on mortgage lenders, background checks of CDO managers, verification that the homeowner can cope with the reset when a "teaser" period comes to an end, and income verification on "stated" income loans.