Genre: Business & Money
Author: Bethany McLean and Joe Nocera
Title: All the Devils Are Here: The Hidden History of the Financial Crisis (Buy the Book)
Sometimes the devil is very patient. He establishes his plan, sets things in motion far in advance, allows something that is initially good to occur, and over many years’ lures one person or one group in at a time, feeding on their ignorance, fear, and greed. Then, in the end, he seeks to destroy them all. Thankfully because there is a higher power, he never fully succeeds; but in 2007 and 2008, he came close.
The seeds for the financial crisis and great recession of 2007 and 2008 were first planted in the late 1970s and 1980s by two men. Lewis Ranieri and Larry Fink created two financial innovations: securitization and tranching. Both innovations were initially applied to the mortgage market.
Securitization allowed individual mortgages to be bundled and sold from lenders to investors as a single security. Tranching allowed for that single mortgage security (which was comprised of hundreds or thousands of individual mortgages) to be broken up into various new securities based initially on their maturity dates and later on their credit rating. These two innovations launched the era of Wall Street’s involvement in mortgages.
The government was also heavily involved in seeding the crisis, initially through the publicly traded but semi-government enterprises, Fannie Mae and Freddie Mac. Both were established to support the long- term growth of the housing market as a route toward higher living standards and the movement of lower income citizens into the middle class.
However, because they were also publicly held and traded, they had shareholders to consider. Over time government regulators at the HUD essentially forced both Fannie and Freddie to lower their lending standards significantly and thereby begin lending to citizens without adequate income, collateral, down payments or credit history.
Lending to this class of mortgage applicants became known as “subprime lending” and was essentially made the law of the land with the passage of the “Community Reinvestment Act.” After initially resisting the government’s “mandate,” both Fannie and Freddie capitulated, and their Boards concurred. In the beginning, the increased homeownership rates pleased the government, and the high fees earned in the newly expanded subprime market appeased shareholders.
What about risk? The argument for the new practices was simple. Housing prices in the U.S. had supposedly never declined at the national level. This “fact” was irrefutable, and all arguments made to dispute it seemed either theoretical or an attempt by the upper class to deny access to homes for the lower class.
New and highly scientific risk models had also recently been created, specifically Variance at Risk, or VaR. This again supposedly allowed CEO’s to accurately measure the amount of aggregate risk that their firm was taking and the incremental risk any new offering might create. These models were highly mathematical and offered a seemingly “sophisticated answer” to the increasingly complex, less intuitive problem of monitoring and managing corporate risk.
Essentially VaR boiled all the complexity down to a single number, at a particular pre-determined probability of occurrence. Generally, unbeknownst to the senior managers, complex financial solutions like VaR always require a preset group of objective functions (e.g., the risk is normally distributed, which it isn’t; and the correlation matrix represents the event that will actually occur, which it didn’t.) In the end, for most, VaR proved to be a mathematical pied piper leading them to their financial destruction.
Finally, it seems that the individual or group making the original loan and underwriting it, simply no longer cared about its actual quality. In this new world of Freddie, Fannie and Wall Street, they had no plans to keep it on their own books anyway.
As these markets grew, there was also a rising need for a “stamp of approval.” Most investors had neither the means nor the inclination to deeply analyze these sometimes complex pools of mortgages and other types of collateral.
The various government-approved rating agencies were more than willing to fill that gap assuming that it would be profitable and perhaps the desired rating would be particularly so. Throughout the final few years, it seems that the rating agencies were willing to give their approval to virtually every offering.
Global agencies also jumped in and inadvertently further exaggerated the growing housing bubble. In an attempt to assure that global banks had adequate capital, they established new guidelines for capital requirements.
This limited a financial institution’s use of its capital for profit-seeking…unless there were creative and legal ways to get around the new guidelines. As it turns out, there were many. They were, however, less transparent, more complex and riskier.
Ultimately, the final act began. In important ways, everything described above also changed the culture of Wall Street. Goldman went public and then lost its way. Its culture of “protect the company” at all costs resulted in its seeming to have done so too often by using its own clients as fodder for itself.
Merrill Lynch decided to leave more than 100 years of sound business practices behind and reprioritize its efforts around aggressive risk-taking for its own account. AIG failed to recognize that one of the units literally, in an “outpost” position, had obligated the entire firm to massive Collateral Default Obligations (CDO’s), with collateral triggers that none of them were aware of.
Bear Stearn’s senior management was seemingly playing bridge while “Rome burned.” Lehman Brothers had amassed one of the largest, and worst, portfolios of them all. And on it went, with virtually every Wall Street firm having followed the financial pied piper to the cliff of ultimate disaster.
In the end, here is a sample of the result– U.S. GDP declined by around 10%, or $1.5 trillion (at least) – U.S. unemployment rose from 4.5% to 10% – The S&P 500 declined by more than 50% – Credit markets ceased to function properly – A global recession in developed countries arose – A monetary and fiscal stimulus program was launched at a size that was about ten times anything previously initiated in history – The U.S.
The dollar weakened – Sweeping new laws and regulations were drafted and are in the process of becoming law – Government debt to GDP in the U.S. has risen to unprecedented levels
In addition, most of those who cooperated most readily with the devil have been taken away, and the number of major U.S. financial institutions has been halved. Gone are Merrill Lynch, Bear Stearns, AIG, Lehman, Countrywide and Ameriquest. Remaining, but far different are J.P. Morgan Chase, Morgan Stanley, Bank of America, Wells Fargo, Citi Group, and Goldman Sachs.
Cooperating with the devil is a loser’s game.
The 2008 financial crisis can trace its roots to the late 1970s when the baby boomers began to transition into adulthood. Many boomers were starting families and buying homes, perpetuating the American dream. It was inevitable that a large population of adult boomers would lead to a boom in the housing market. Economists expected this, and on its own, this was certainly no crisis. But, the rest of the story came as a surprise to everyone.
At the beginning of the boom, the mortgage market was highly inefficient and geographically segregated in terms of lending.
The reason for this inefficiency was that mortgage lending was primarily a local practice facilitated by the local S&Ls and banks. Then the S&Ls began to struggle in the early 70s due to high inflation and rising interest rates. This meant that there could be shortages of funds for lending in one area of the country while, simultaneously, a surplus of funds existed in another region. The S&Ls troubles and eventual withdrawal from the mortgage business created market conditions ripe for Wall Street to enter an industry that it had largely avoided.
Lewis Ranieri of Salomon Brothers and Larry Fink of First Boston and Blackrock are widely credited with developing the first marketable mortgage-backed security (MBS) in 1977. Lewis Ranieri coined the process “securitization,” which at first was used to bundle mortgages to create a security, a tradable product.
Securitization would later be used to bundle everything from subprime mortgages to credit card debt. Larry Fink developed the idea of tranching, which enabled the securitizer to carve the bundle of assets into tranches with corresponding levels of risk; this made the security marketable to a much wider range of investors.
David Maxwell, CEO of Fannie Mae, was the lynchpin that kept the fledgling mortgage-backed security market going. Ranieri and Fink, at least in the beginning, needed Maxwell and Fannie Mae to be able to complete the sale of their new class of securities.
Ranieri’s and Fink’s innovations, securitization and tranching made MBSs widely appealing and caused the mortgage-backed security market to expand in 1983. In a market environment that was hungry for yield, MBSs gave investors the product that they had been waiting for and banks the perfect product to sell.
MBSs could be bundled in numerous ways, tranched and diced, and then sold for more than the value of their components. The bank would then earn an underwriting fee and eventually profits from the trading of these new securities. Even at this early stage during the first half of the 1980s, Wall Street was hooked.
The Hybrid Beast
Fannie Mae was an out of the ordinary corporate entity. It was a publicly traded company that had shareholders and a board of directors, but it also had a government-sponsored mandate and the implied full faith and credit of the United States government. Fannie Mae was half government enterprise and half private enterprise, termed a GSE, a government-sponsored entity.
A key piece of legislation was passed in 1968, which provided Fannie Mae powers that would, ten years later, allow it to participate in and take advantage of the asset-backed security craze. The powers included the ability to buy mortgages that were not issued by the government and the ability to issue securities that were backed by the mortgages they were guaranteeing.
David Maxwell took over Fannie Mae in 1981 and began running the company like a true private enterprise.
Profits and shareholders were suddenly the priority. Before Maxwell came onboard, Fannie had been very close to bankruptcy due to rising interest rates during the Federal Reserve’s fight against inflation in the 1970s. Bankruptcy never happened because banks continued to lend to Fannie.
The only reason they were willing to continue to lend was because of Fannie’s implied government backing, an association that would help Fannie continue to do business for years to come. Maxwell did not tolerate mediocrity and searched for ways to increase the company’s bottom line, firing many managers in the process.
His search for higher profits led him to the mortgage-backed security market that was still in its infancy. Guaranteeing and selling mortgages to Wall Street was an excellent way for Fannie to earn fees from banks and quickly became Fannie Mae’s most profitable business. However, since Fannie Mae guaranteed the mortgages it sold, the risk of default still existed on its books.
Fannie, and Freddie, to a lesser extent, became the middlemen linking Main Street (borrowers) and Wall Street (securitizers). Fannie Mae and Freddie Mac were essentially identical. They had the same business model and all of the same government guarantees. The only difference was that Fannie primarily bought mortgages issued by banks and Freddie bought mortgages issued by thrifts.
The GSEs and Wall Street had a love-hate relationship.
Wall Street needed, in the beginning at least, the GSEs mortgage guarantee because it was the only type of mortgage that investors wanted to buy. The GSEs needed Wall Street as a place to unload debt, earn fees, and increase profits. Ultimately, Wall Street wanted Fannie and Freddie out of the picture.
After David Maxwell retired from Fannie Mae in 1991, Jim Johnson took the reins. By the mid-1990s, the GSEs capital requirements had dropped to only 2.75% of total assets. This was much lower than what Wall Street firms had to carry on their books and was a byproduct of Johnson’s tough handling of Fannie’s regulators.
Maxwell had managed Fannie’s regulators with a light, but forceful touch. Johnson, on the other hand, made it no secret of his disdain for any regulation. The GSEs employed an army of lobbyists in Washington to get their way. The Department of Housing and Urban Development (HUD) and the Office of Federal Housing Enterprise Oversight (OFHEO), which oversaw the GSEs, had no real regulatory powers.
In the late 90s, OFHEO was without a director for two years. The regulation was lax. What Fannie wanted, it usually received; and if politicians caused trouble, Fannie would pull the “he/she is against homeownership” card. This trick always caused politicians to back off. If you did not believe in homeownership, then you did not believe in America.
During the 90s, the GSEs wanted nothing to do with subprime mortgages.
The Community Reinvestment Act forced banks and the GSEs to make more loans to borrowers in poorer neighborhoods as a boost to homeownership rates, which had dropped after the S&L crisis. This government mandate forced Fannie to enter the subprime market.
The GSEs up until this point had been very cautious about the subprime mortgage risk that they took; there was just too much money to be made in safer investment vehicles. Many government officials actually complained about the GSEs not playing an active enough role in the subprime market. After the Community Reinvestment Act, Fannie bought just enough subprime mortgages to keep Washington content.
During Bill Clinton’s term in the late 90s, Andrew Cuomo was appointed as HUD secretary. Cuomo was a true believer in the idea of affordable housing for all. In 1999, he proposed a significant increase to the GSEs affordable housing goals.
Fannie fought hard, but Cuomo was not swayed by the GSEs legion of lobbyists. Fannie eventually had to agree to increase loans to low-income borrowers. Politicians used “low income” as a euphemism for subprime. In two short years, subprime lending would make up 50% of Fannie’s overall lending.
The only way for the GSEs to increase lending to low-income borrowers was to lower their lending standards materially. The GSEs would later cite the subprime lending push by Cuomo and Clinton as the tipping point for rushing headfirst into the subprime market; however, by the turn of the century, Fannie executives had become enamored with the high returns subprime lending was producing.
Cuomo and other politicians are not the only one’s deserving blame. The truth is that the GSEs had been on the sideline for the majority of the rise of the subprime market, and they desperately wanted to take part in the record profits that Wall Street firms were reporting each quarter. The problem with the GSEs focus on profits was that Fannie and Freddie were not created for the sole purpose of making money, but rather to supply liquidity to the housing market when needed, a sort of hybrid lender of last resort.
After the S&L crisis in the 70s, coupled with Ranieri’s securitization innovation, Wall Street took over as the primary source of mortgage funding. Specialized mortgage companies began to open. Two of the most prominent examples include Ameriquest and Countrywide.
The founders of these new companies were not typical cautious bank managers. They were intense entrepreneurs with an insatiable appetite for growth, for risk, and for profit.
Angelo Mozilo was the head of Countrywide. Countrywide believed in sound underwriting principles for the majority of the subprime market expansion during the 80s and 90s. Mozilo once described Citicorp’s mortgage problems by saying, “they tried to take a shortcut and went the way of every institution that has ever tried to defy the basics of sound underwriting principles.” Unfortunately, Countrywide inevitably entered the subprime market at the behest of shareholders needs for increased market share and profits.
In 1983, Larry Fink created the first collateralized mortgage obligation (CMO).
This would later turn into a collateralized debt obligation (CDO). A CMO was a mortgage-backed security that had three different tranches: a short-term five-year debt, a medium-term 12-year debt, and a long-term 30-year debt. The actual issuer of this debt was the government sponsored entities because investors only wanted to buy debt that was insured by Fannie and Freddie.
The GSE guarantee meant that investors no longer had to worry about the risk of default. At this stage in the history of MBSs, the GSEs were not guaranteeing subprime loans, and the subprime market had yet to gain significant traction.
The process of securitization and the introduction of CMOs severed the vital link that entwined the fates of both lender and borrower. Before securitization existed, the lender had a vested interest in the borrower being able to pay back his loan. Now, because of securitization, the lender could make the loan and then sell it to the GSEs and Wall Street.
Many lenders already had Wall Street buyers lined up before the loan was ever made. This removed the liability from the lender’s book and moved it to another firm. It also let the lender relinquish all responsibility for the future performance of the loan. The debt was now someone else’s problem. The crucial link had been broken between the borrower and the lender. This decoupling would later become traumatic for the industry.
In a now infamous 2003 phone conversation between Bear Stearns and Ameriquest, Bear Stearns asked, “how can you increase your [mortgage] volume?” Ameriquest replied, “we said, tongue in cheek, well we can do a 100 percent loan-to-value stated income loan for 580 FICO scores!” Bear Stearns responded, “okay!” Ameriquest replied with, “no problem! Let’s do this all day!” “And they did it, in massive quantities.” This exchange meant that the borrower would not have had to make a down payment, their income would not be verified, their credit score could be very low, and a loan would still be provided.
Wall Street wanted mortgage lenders like Ameriquest to churn out higher and higher volumes. The only way to do this was to continually lower lending standards. Lending shops, like Ameriquest, had branches all over the country that sold mortgages at an incredible pace. Many times, fraud and other deceptive practices were used to make the sale.
Moody’s Stamp of Approval
Lewis Ranieri and Larry Fink needed three crucial components to make mortgage-backed securities a successful market product: securitization and tranching, Fannie Mae’s ability to issue and guarantee securities, and the blessing of the rating agencies. By 1985 the only component that they were missing was the approval of the rating agencies.
Credit enhancements were invented as a byproduct when the Resolution Trust Thrift, which was tasked with cleaning up the failed thrifts from the S&L crisis, wound up with a lot of bad debt on their books.
This debt could not be backed by the GSEs because of its toxic nature. Without the GSEs backing, the Resolution Trust would not be able to unload and sell the debt to investors.
For most investors, it was not specifically the GSE guarantee that was essential to investing in the securities, but rather the triple-A rating that accompanied the guarantee. Pension funds and some sovereign wealth funds required assets to have triple-A ratings to purchase them.
Wall Street’s answer to this problem was “credit enhancement.” Examples of typical credit enhancements included: purchase of insurance from a third party such as AIG, over-collateralization, and senior/subordinated debt structures. Overcollateralization was the process of stuffing more mortgages than required into the product to make up for potential defaults.
Senior/subordinated structuring was a process where cash flows from the riskier underlying mortgages were redirected to the senior bonds to make them appear more secure, as this theoretically meant that the super-senior holders would always get paid.
Credit enhancements convinced Moody’s to begin rating certain tranches of non-GSE guaranteed debt as triple-A. The invention of credit enhancement was very important for two reasons. The first was that Wall Street had finally figured out a way to get around using the GSEs as a middle man.
Wall Street no longer relied upon the GSE guarantee, they could get the securities rated as triple-A without getting them guaranteed by Fannie. It was then an easy sell to investors. The second reason was that the rating agencies had now received a taste of the big money that came from being directly linked to the success of the mortgage-backed security market.
The banks needed the rating agencies now, and Moody’s, Fitch and Standard & Poor’s were more than happy to play along and receive the increased business and fees. Wall Street now had a huge securitization business that they could do on their own, without, in their opinion, the meddlesome involvement of the GSEs, and consequently, the MBS market took off in the second half of the 80s.
Risk On, Risk Off
During the rise of mortgage-backed securities, J.P. Morgan was focused on another market aspect. They were obsessively tinkering with risk management practices. Company management wanted to be able to measure risk, model risk, predict risk and embrace new risks that other institutions had yet even fully to understand. JP Morgan’s answer to risk management was value-at-risk or VaR.
VaR would become the industry’s standard risk measurement tool. JP Morgan, though, would be one of the few firms that truly understood the tools capabilities and limitations. VaR at its core was just a bell curve. It could tell the firm what they might lose 95% of the time, but it did not measure the other 5%, the “fat tails” or the “black swan” events.
Many companies that implemented JP Morgan’s VaR model took it as the firm’s worst case scenario.
This view was incorrect. Worst case scenarios were not even measured using VaR. VaR was a more useful tool for measuring risk over time. At JP Morgan, VaR was always used in conjunction with personal knowledge about the traders themselves, their current position and trading tendency. VaR was a powerful extra risk measurement tool. It was not created to be an excuse for bad risk managers.
To help prevent undue risk something known as the Basel Committee was created. The Basel Committee on Banking Supervision was an international effort that took place in 1988 to get a global set of capital requirement rules implemented. Banks understood the importance of holding capital, but at the same time, each dollar that was held in reserve was a dollar that the bank could not loan and profit from.
The Basel Committee’s solution to the global capital requirements project was a set of risk-based capital requirements. Commercial loans were determined by Basel to be the riskiest type of debt, therefore, requiring the full capital ratio. Mortgages not backed by the GSEs were viewed as a moderate risk and given a 50% capital requirement.
Finally, mortgages backed by Fannie and Freddie, and therefore the implied full faith and credit of the United States, were viewed as the least risky and given a 20% capital requirement rate. What this amounted to is that banks now had an even greater incentive to purchase mortgages because they required much less capital when compared to corporate debt.
When the Basel global capital requirements deemed that Fannie and Freddie backed mortgages needed only 20% capital in reserves, rather than the 50% reserve for private label mortgages, bank lobbyists started campaigning to get the private banks and the GSEs on an equal playing field. It took over ten years, but in 2001 the bank lobbyists succeeded. Banks now only needed to carry 20% capital in reserve on their triple-A rated mortgages.
JP Morgan created the credit default swap (CDS), during their years of obsessive risk management and innovation.
The idea for CDSs arose when the Basel Committee deemed that all commercial paper, regardless of the company issuing the debt, required the full capital ratio. From JP Morgan’s point-of-view it did not make sense that two different corporations could have identical risk profiles. Lending to Wal-Mart has a different set of risks than when lending to Target.
JP Morgan needed a way to price the risk associated with holding different corporations’ debt. Credit default swaps were the answer. By trading CDSs, it would allow the market to judge for itself the riskiness of assets by putting a price on them. If the cost of the CDS increased, then the chance of the underlying security defaulting must have increased in the eyes of the market. JP Morgan could now price their loan portfolios.
Besides the pricing benefit that CDSs created, they were also an opportunity to lower capital requirements from the Basel standards.
A credit default swap moved the risk of default from one party that did not want it on their books to a counterparty that did. At its most basic level, it was an insurance policy against the risk of default. By purchasing credit insurance, banks were removing risks from their balance sheets, therefore freeing up capital. For example, Bank “UNO” bought credit default swaps from Bank “DOS” to protect its portfolio against the risk of default.
Bank DOS sold the CDSs to Bank UNO in return for a periodic fee. If a default never occurred, then Bank DOS would continue to receive payments and not have to pay out. In the event of a default, Bank DOS would have to cover the full loss amount for Bank UNO. The loan itself stayed on the books of Bank UNO. The fact that the actual loan did not change hands is very important to note. If an industry-wide meltdown occurred and Bank DOS was not able to cover all of Bank UNO‟s loan losses then Bank UNO would still feel the difference.
The trouble with CDSs was that by transferring risk to another entity the risk did not truly go away, it was just hidden from view. Adding synthetics and leverage to the equation would put the entire financial system at risk. The institutions that sold the CDSs never expected to have to pay out losses. The underlying debt was supposed to be safe; after all, much of it was double or triple-A rated.
In the late 90s, JP Morgan took the CDS product one step further. They combined credit derivatives with securitization. This allowed JP Morgan to bundle several credit derivatives that referenced hundreds of corporations’ debt. Securitized CDSs, later called synthetic CDOs, differed from MBSs in that MBSs were backed by hard assets, like homes, and synthetics were not.
Synthetics owned CDSs and only referenced the underlying assets as a gauge for performance. Synthetics were normally tranched into three levels. The equity portion was the riskiest and produced the most return if it didn’t default. The mezzanine tranche carried less risk but offered less return. The super-senior tranche produced the least amount of return but was considered almost as safe as US treasuries.
JP Morgan’s risk models showed that it would take a monstrous catastrophe to have defaults cut through the lower two tranches and affect the super-seniors.
The credit rating agencies agreed and rated the synthetic CDO super-senior tranches as triple-A. Synthetics at first referenced only corporate debt, but over time changed to reference mortgages. Mortgages were the much more dangerous product and an asset class that the Basel Committee and the rating agencies drastically underestimated.
Rating Agency Follies
In 2007, a Moody’s employee complained about the ratings that they were giving, saying “they make us look either incompetent at credit analysis or like we sold our soul to the devil for revenue or a little bit of both.” The truth is that it was a little bit of both.
The prominence and power of the rating agencies increased alongside the rise of the mortgage-backed security market. Firms like Moody’s became ever more essential especially once the GSE guarantee waned in importance because of the introduction of credit enhancement.
The rating agencies were paid by the firms for which they rated investment products. This relationship is similar to a giant “do loop.” The rise of the structured products division within Moody’s coincided with Moody’s rapid deterioration in rating quality.
Wall Street wanted Moody’s to rate their products as triple-A, so they incentivized higher ratings by paying more for them. By this time there was a huge volume of MBSs and CDOs that needed to be rated, and the business was a major cash cow for Moody’s. Profits were too good to say no; even if they did, the bank would just pass the offer to another rating agency who would say yes.
At the same time, Moody’s executives wanted more market share. The only way to get more market share was for Moody’s to drop its rating standards further. A Moody’s employee aptly described the situation as “a race to the bottom.”
It is interesting to note that over the long history of the rating agencies, they have been consistently bad at making timely rating calls. They seemed to get things wrong when even a slight unexpected event happened. They missed the bankruptcy of Orange County California, the Asian and Russian meltdowns, and the collapse of the hedge fund Long Term Capital Management.
The most obvious fault is that Moody’s would often downgrade companies just days before they went bankrupt, therefore skipping many levels of ratings; often going from a rating of A to D in one fell swoop. So, with the agency’s consistent failures, it is strange that they “carried the force of law” in the securities business.
Ratings had a surprising effect on the contents of CDOs. The contents ended up just not mattering. Investors, when buying CDOs did not know the exact CDSs and other assets in the CDO, but since it was rated triple-A, they would still buy it. Investors began to effectively “buy the rating,” which is why the rating agencies were so integral to the system. The CDO market went from $69 billion in the year 2000 to over $500 billion in 2006. The rating agencies‟ profits followed in lockstep.
The culture at Goldman Sachs began to shift from a client focus to a trading focus long before Goldman went public on May 4, 1999, but their IPO was the point-of-no-return. Now there were shareholders to answer to and more profits to be made. Goldman Sachs stood head and shoulders above competitors in terms of ability to price their assets.
This “price is right” attitude would greatly benefit them in the depths of the crisis.
Goldman used the mark to market accounting. They marked their books at the end of each day at the market prices for all of their securities. Goldman kept enough cash in short-term securities to deal with what they believed to be worst-case scenarios.
This was very different from what most firms practiced, except for perhaps JPMorgan. Goldman, like JP Morgan, viewed VaR as a tool, a gauge, and not gospel. Goldman understood risk and practiced active risk management more intensely than nearly all other investment banks.
At many firms, including Merrill Lynch, risk officers were mainly a back office function and were looked down upon as not as smart as the front office. People at risk were often paid less. At Goldman, being in the risk department was something to be proud of; it was an accomplishment.
Goldman’s risk team was called the Federation, and they were in constant contact with personnel on the floor and upper management.
One of Goldman’s biggest and simplest assets was that they communicated constantly; information was passed freely within the firm. Traders did not have to be afraid to disclose a loss to their boss; they were encouraged to do so as soon as possible; this helped to keep everyone at the firm on the same page and allowed each of Goldman’s departments to have a good idea of how the other departments operated. The executive suite at Goldman knew what was happening in their company each day and understood “trader talk,” a feat many CEOs could not claim.
Founded by Charlie Merrill in 1914, his lifelong crusade was to sell investments to the American middle class. His mantra was to “bring Wall Street to Main Street.” Merrill Lynch was known as the “Thundering Herd” and had brokerage offices all over the US.
Merrill had a sound business model. Their analysts were top rated, their brokerage services made money, and their investment bank division was well respected, but their trading operations were not as well known.
In 2002, Stanley O’Neal became CEO of Merrill Lynch and began to expand the trading desk.
Merrill compared itself constantly to Goldman Sachs who was making more money with fewer employees and doing so in more extravagant ways such as using hedge funds and complex derivative desks. O’Neal was going to try to make Merrill Lynch more like Goldman Sachs. This would be Merrill Lynch’s and Stan O’Neal’s downfall.
He wanted to get rid of the “Mother Merrill” culture and replace it with an emphasis on trading. O’Neal was focused on risk, but not in the same constructive way that Goldman and JP Morgan focused on it. O’Neal wanted more risk, and he became upset when larger, riskier positions were not being taken.
If the daily VaR number was not big enough, he would actually become enraged. With Stanley O’Neal and his hefty appetite for risk at the helm and deep envy of Goldman, Merrill Lynch dove head first into the mortgage market, and within a few years was the number one underwriter of CDOs.
Chris Ricciardi was head of the CDO desk at Merrill and was the aggressive risk taker that O’Neal wanted.
Ricciardi had been hired to take advantage of the exploding mortgage market. Merrill soon took a 20% stake in the mortgage originator Ownit, and this ensured that Ricciardi and his fellow traders would have an ample supply of mortgages to securitize and trade.
The normal cycle progressed as follows: Ricciardi would take the mortgages it bought from Ownit, warehouse the securities on Merrill’s books until they could be bundled into a CDO, contend with Moody’s to receive high ratings, modify the CDO with credit enhancements per the rating agencies instructions, and then sell it to investors in different tranches.
Investors around the world were lined up to buy these products. Jeff Kronthal, head of the mortgage desk, was tasked with trying to control Ricciardi and his team and placate Stan O’Neal. Kronthal was the only voice of reason during these times of excessive risk-taking. In 2005, O’Neal fired Kronthal who was probably Merrill’s best trader and had worked there for 17 years in the MBS market.
The “Allstate” of Wall Street
American Insurance Group-Financial Products (FP) under the supervision of Joe Cassano was a very well-run business. Cassano and his team did their due diligence on all deals in which the firm was involved. He would tell his traders, “Be careful out there. Don’t take big positions.” They had one blind spot, but it was a big one: credit default swaps.
FP had created a profitable niche business insuring the super-senior tranches of other firms’ CDOs.
Super-seniors were the top-tier, triple-A rated tranches of CDOs, which only got hit with losses if the first and second tranches went under. Many firms would approach FP looking for CDSs for capital requirement relief. The firms wanted to get some of the default risk off of their books to lower their capital requirements and use the money for other trades.
By the time of the financial crisis, AIG-FP had insured close to $200 billion of top-tier, super-senior debt. This was supposedly risk-free debt on par with US Treasuries that they were insuring, free money in the minds of the FP traders. FP never hedged their exposure to the CDSs they were selling. In 2004, FP began insuring multi-sector CDOs. The same as a mortgage filled CDO, but this time the CDOs had “everything but the kitchen sink” in them including student loans, credit card debt, and mortgages.
In the last few years leading up to the crisis, AIG-FP gained the lion’s share of the super-senior insurance business for two reasons. The first was because people trusted AIG, its storied insurance practices, and its great credit rating. The second, much more sinister, reason was because of a seemingly inconsequential paragraph in each of FP’s CDS agreements.
FP had written collateral triggers into their contracts with counterparties. These collateral triggers allowed FP’s counterparties to demand capital from AIG if any of the following events occurred. Collateral trigger number one was a drop in AIG’s overall credit rating to single A.
The second was any downgrade of the super-senior tranches that they were insuring. The third trigger of collateral payment was a decline in the market value of the securities FP was insuring, even if the securities were not downgraded by the rating agencies.
Frost who was in charge of CDSs for AIG and Cassano felt that having to put up collateral on these triple-A rated super-senior tranches was beyond the realm of reasonable possibility and therefore never discussed it within the company. Most employees did not know that the triggers existed, so it was that much more startling when collateral calls started coming in during the crisis.
In 2006, Lewis Ranieri gave a talk at an all-day housing symposium. He was appalled by what he saw in the mortgage market, such as the deficient underwriting standards and the documentation frauds at Ameriquest. He never thought that what he, Fink and others helped create would be used and abused in this fashion. Ranieri was shocked by what was going on in the industry.
House Of Cards
In January of 2006, the asset-backed securities index (ABX) opened. It was a joint venture between Deutsche Bank, Goldman Sachs, and Bear Stearns. It was an index based on subprime mortgages that represented certain tranches at different ratings and different issue dates and maturities. With the opening of the ABX, investors could now short MBSs.
This was a new concept, up until this point almost no one had really thought about shorting the MBS market. The way the holder of the short got paid was as the mortgage lost value. The street called it PAUG, pay-as-you-go. Conversely, the long position received interest payments periodically from the short holder. Since all of this trading was based on asset-backed securities, there had to be a finite amount of assets.
There could only be a certain number of mortgages in the country to securitize, but synthetic CDOs solved the problem of finite assets.
If mortgage volume completely dried up, firms could still make and trade CDOs. These synthetic CDOs referenced mortgage derivatives that already existed. They were traded on the ABX and could be levered several times over. The opening of the ABX caused CDO underwriting to increase dramatically. Synthetic CDOs would amplify gains and losses because now an infinite number of side bets could be made quickly by just purchasing/shorting synthetics on the exchange.
CDOs were toxic to the financial system because they had the stamp of approval from the rating agencies, they exponentially increased leverage in the financial system, and the debt upon which they were built was mainly low-quality subprime mortgages.
Build Up To The Crisis
To the great frustration of many in the financial business today, Goldman Sachs escaped relatively unscathed from the crisis. They achieved this by doing what Goldman had always preached, “protect the firm at all costs.” In 2007 and 2008, Goldman was often profiting on the other side of their client’s trades, and the client was never informed. When constructing deals for clients who wanted to short the subprime market, Goldman would sometimes get stuck with the long side of the trade.
To get a short position that they needed to hedge, Goldman would create products to sell and use their renowned franchise name to make the deal happen. The Hudson deal is an example. Goldman picked all of the assets in Hudson and then pushed hard to sell it to clients. Once sold, Goldman took the short position on the other side.
Quotes from internal Goldman emails regarding the Hudson deal and others stated, “we aggressively capitalized on the franchise to enter into efficient shorts,” prompting the head of fixed-income sales in Europe for Goldman to write: “the damage this has done to our franchise is very significant.”
In November 2006, Goldman had $7.8 billion of subprime mortgages on its balance sheet, and $7.2 billion of mortgage-backed securities.
They also had a substantial long position in the ABX and warehouse lines of credit to New Century and other subprime lenders, meaning Goldman was substantially exposed to subprime. In December of 2006, Goldman’s subprime desk lost money ten days in a row. The party had ended, and Goldman Sachs did a reality check.
To protect themselves and “get closer to home,” Goldman traders headed by Dan Sparks, head of the mortgage desk, began aggressively shorting the ABX. Traders entered into all of the short positions that they could get their hands on. They then tried to push out and sell to clients all of the subprime assets they had warehoused and had sitting on their books.
A June 24, 2007 presentation given inside Goldman described the subprime market as “game over” and that exposure on the street was especially high for Merrill and Lehman. Goldman covered much of its long position in 2007 and made billions of dollars when many firms were losing money. What Goldman did in 2007 was protect itself at the expense of its clients. “Protect the firm at all costs” was a go.
The Abacus 2007-ACI deal was the infamous short trade that John Paulson put together with the help of Goldman Sachs. Paulson wanted to make a huge bet against the subprime market, so Goldman rented the respected Abacus security name to him for 15 million dollars.
The astonishing feature of this trade is that Paulson was able to handpick the mortgages that he wanted to be referenced in the Abacus CDO deal. Rather than select the best debt (lowest risk), he picked mortgages that he thought had the greatest chance of failure. The counterparty did not know that this was happening. At one point, Paulson got rid of all the Wells Fargo mortgages in the portfolio because he thought they might actually perform well. The short netted $4 billion in profits for Paulson.
In the spring of 2007, Ralph Cioffi and Matthew Tannin of Bear Stearns were getting nervous.
They started a fund in 2003 called the High-Grade Structured Credit Fund. A few years later, they started a second fund called the High-Grade Structured Credit Strategies Enhanced Leverage Limited Partnership, a leveraged short fund. As the triple-A tranches of debt began to falter and even the super-seniors became pressured, all of Wall Street started a massive push to get as many CDOs out their doors and off their books as possible.
Goldman had already done this. This was the last push of MBSs and CDOs and the tipping point for the crisis. From January 2007 to June 2007, CDO issuance peaked at $180 billion. Cioffi and Tannin talked to Bear’s chief economist who said not to worry, the subprime market was going to come back. There was a major disconnect from reality. The Bear Stearns hedge funds headed by Cioffi and Tannin collapsed in June 2007. The whole firm would soon follow.
Moody’s and the other rating agencies had noticed the increase in defaults. Their initial response was to require more credit enhancements for the CDOs to receive a triple-A rating. The problem with that response was that the newly minted CDOs referenced many older CDOs that had fewer credit enhancements, and these older CDOs were going bad.
The house built with paper walls was collapsing in on itself. Moody’s made the whole dilemma look rosy to the outside world, but internally they knew that the CDO market had turned into a ticking time bomb.
The Final Days
On July 10, 2007, S&P and Moody’s put a total of 1,011 tranches of securities on downgrade review. While the problem was evident much earlier, the reviews were not announced until July mainly because the agencies were afraid of what would happen come the day they had to downgrade the debt.
The ratings and all of the securities were so intertwined and prolific in the market that the effect of a downgrade was unknowable, although they knew it would be all bad. Actual downgrades started in October. The downgrade reviews and eventual downgrades caused a massive loss in investor confidence. This, in turn, caused the securities market to freeze up. All of a sudden no one wanted any assets that depended on the rating agencies or that referenced asset-backed securities.
When longtime trader, Jeff Kronthal, was fired from Merrill Lynch in July 2006, Merrill had approximately $6.5 billion of subprime on its books.
One year later, Merrill Lynch had $55 billion worth of subprime. A single trader, Osman Semerci, had added close to 50 billion dollars’ worth of subprime exposure! Most of the debt was triple-A rated tranches of subprime CDOs, not that ratings mattered anymore. At this stage of the crisis AIG was no longer insuring tranches or writing security, so Semerci and his team had just kept the triple-A rated debt on Merrill’s own books.
Each time the desk created a new deal, a large portion of it would go straight to Merrill’s balance sheet. Semerci truly believed that the triple-A tranches were riskless and continued to tell the board of directors his views. Merrill was completely exposed to any downturn in the mortgage market.
The difference between Goldman and Merrill was that Goldman treated all subprime the same, as high- risk bets, even the triple-A rated tranches. Merrill, on the other hand, viewed the triple-A rated tranche as virtually riskless. Goldman realized that the slightly greater return on highly rated subprime was not worth the risk compared to what they could get by purchasing US Treasuries.
Stanley O’Neal did eventually recognize that Merrill was in a dire predicament. He approached a board member about merging with BofA. The board member shot it down stating his strong disapproval of Ken Lewis, the CEO of Bank of America.
Soon after, Merrill fired Semerci and his team and announced Merrill’s largest trading loss ever, $5 billion. O’Neal then wanted to sell to Wachovia as the losses piled up. He announced his idea to the board at a dinner, and the directors contested it vehemently. Stanley O’Neal was fired shortly after the dinner on October 29, 2007. Of the approximate $45 billion of subprime that Semerci added to Merrill’s position, close to $42 billion ended up being written off.
Eleven months after O’Neal was fired, Merrill Lynch was sold to BofA for $29 a share during one of the worst weekends of the crisis. Ken Lewis had agreed upon $100/share when O’Neal approached him about a sale a year earlier.
JP Morgan and Goldman, in the years leading up to the meltdown in 2007, needed a way to move risk off of their books.
The answer was AIG-FP. JP Morgan bought credit default protection from AIG with the added bonus of collateral triggers. Joe Cassano, of AIG-FP, championed the cause for insuring super- senior tranches of debt because it was free money; there was no real risk of default, the models said so.
Downgrades started pouring out of the rating agencies in the second half of 2007. Martin Sullivan, CEO, and Bob Lewis, CRO, of AIG did not know about the collateral triggers in FP. Only Cassano, Frost and Forster knew the details. FP’s counterparties had been paying FP millions of dollars to insure their triple- A debts and could demand compensation, because of the collateral triggers in the contracts.
On July 26, 2007, the first margin call came across the wire. It was a demand for 1.8 billion dollars from Goldman Sachs. Goldman claimed that $20 billion of AIG’s insured securities had declined by a total value of $1.8 billion.
Goldman had been aggressively marking their CDOs down, much more so than any other firm. For example, Goldman had a CDO valued at 85 cents on the dollar, Merrill had the same CDO priced at 98 cents, and FP had it at par. AIG-FP fought the collateral call saying that the CDOs had not dropped in value and had definitely not dropped to 85 cents on the dollar.
When counterparts objected to Goldman’s marks on securities, Goldman would offer to sell to the counterparty at the low price. No counterparty ever took the offer to buy, therefore proving to Goldman that the prices were still not low enough. Goldman did not like that AIG-FP was dodging collateral calls, so they started buying protection against AIG. Goldman at this time, late August 2007, bought $575 million worth of CDSs on AIG; these contracts would payout if AIG went bankrupt.
Fannie Mae, during the messy summer of 2007, was actually plotting a comeback internally.
They wanted to regain their former glory in the mortgage market. In 2007, the GSEs added a combined $600 billion of new mortgage debt to their books. This would turn out to be some of the most toxic debt to date. By 2008, the GSEs had $5.3 trillion in mortgages and guarantees and only $84 billion in capital.
Lehman Brothers, led by Dick Fuld, walked a very tight line with the Federal Reserve. The Fed told Fuld that he needed to raise capital. The Fed performed several stress tests on Lehman, and Lehman failed all of them. Instead of raising capital, Fuld actually put on more exposure in the months leading up to the crisis. The only explanation for this is that Fuld wanted Lehman to be too big to fail and receive a government bailout.
- September 11: Goldman gets back from vacation and starts demanding collateral again from AIG
- September 13: Goldman buys an additional $700 million of protection against AIG
- September 20: Goldman reports a profit of $2.9 billion for the quarter
- November 2: Goldman ups the collateral call to $2.8 billion. AIG-FP still disputes their prices
- November 7: AIG posts $3 billion profit but mentions losses at its FP division. On the conference call, analysts only want to know about AIG-FP‟s super-senior exposure levels
- November 18: Goldman now has $1.9 billion of CDS protection against AIG defaulting
- November 23: AIG posts a total of $2 billion in collateral. Goldman now demands a total of $3 billion. Goldman’s CDO prices are much lower than the rest of Wall Street
- February 5: Joe Cassano is forced to retire from AIG-FP
- February 28: AIG announces an $11.47 billion write-down of their super-seniors. Goldman increases its collateral call to $4.2 billion
- March 1: Bear Stearns stock trades at $70. (The firm has $18 billion in cash!)
- March 17: Bear Stearns is sold to JPMorgan for $2 a share
- June 15: CEO Martin Sullivan leaves AIG. He spent his entire career there
- June 18: AIG-FP posts $5.4 billion in collateral for Goldman. Many of the claims originate from the Abacus deal
- September 5: Fannie Mae & Freddie Mac are taken over by the US government
- September 15: Lehman cannot find a buyer and files for bankruptcy
- September 16: The Government, in an $85 billion deal, bails out AIG
The dominoes continued to fall. Crafting a deal with Mitsubishi UFJ, a Japanese bank, saved Morgan Stanley. Washington Mutual was sold to JPMorgan in a fire sale hours before bankruptcy. Wachovia was collapsing when Wells Fargo bought it in December 2008. Citigroup needed multiple outlays of government funding.
By the end of 2008, the subprime edifice had collapsed, and all of the institutions that had been supporting it were gone or beginning to pick up the broken pieces.
Ultimately, 91% of triple-A rated subprime mortgage-backed securities issued in 2007 went bad and so did 93% from 2006. The game was officially over. The losers far outnumbered the winners, who mostly resided at Goldman Sachs.
Treasury Secretary, Hank Paulson, helped broker many of the deals that would preserve portions of Wall Street’s storied firms. After Bear Stearns collapsed, market liquidity dried up, and Wall Street was having major issues functioning. The problem was that because of the rapid crash of Bear Stearns, confidence in the financial system had cratered.
No one knew where the next shoe was going to drop. No one knew their true exposure to subprime. Lending was frozen. Paulson knew that it would take a massive government intervention to calm the markets, inject liquidity and return to normal operations. TARP was announced.
TARP was a $700 billion Troubled Asset Relief Program designed to restore confidence in the marketplace and unfreeze the financial system. Over three years later that financial system is still struggling.
Most of the investment products that were created in the 30 years leading up to the financial crisis were designed to solve a genuine market problem. The products, over time, gained a life of their own wreaking havoc on financial markets driven by the constant need for greater profits and constant distancing from any real collateralized assets.
Before securitization, lenders had to care about the creditworthiness of borrowers. If the borrower defaulted on the loan, then the lender was left holding the bag. Securitization allowed the lenders to sell the mortgage immediately to Wall Street. Once sold, the mortgage was no longer the lender’s concern.
The decoupling of lender and borrower was at the heart of the mortgage-backed securities meltdown and the ensuing 2008 financial crisis. There was a complete loss of credibility and a lack of responsibility between all parties, borrowers, lenders, securitizers, and regulators. Perhaps lessons have been learned, but history always seems to have a way of repeating itself.
Britt always taught us Titans that Wisdom is Cheap, and
This post has been slightly edited to promote search engine accessibility.