The Great Financial Crisis
Part I: The great financial innovation and corruption (Free)
Issues discussed:
The route to the greatest financial crash since the Great Depression, the Great Financial Crisis of 2007-08 (-2012) explained.
Actions of authorities feeding the mania in the U.S. housing market leading up to the crisis.
The main lesson of the run-up to the GFC: We should not trust authorities to keep our financial system safe.
Over the Holidays, I will publish a two-part series on the Great Financial Crisis (GFC) of 2007-2008. Some argue that the crisis continued till the peak of the European debt crisis in July 2012, but it’s also generally held view that the actual financial panic concentrated on the years 2007 and 2008. This series is based on two lectures on the GFC I gave at the University of Helsinki in 2011 and 2012 as a part of a lecture series on economic crises held by Professor Vesa Kanniainen, and my book draft. I can honestly say that my two lectures were the most comprehensive presentations of the crisis I had seen anywhere outside the U.S. during that time. The GFC and these two lectures also redirected my post-Doctoral research to economic crises from income inequality and economic growth I studied in my Doctoral Thesis.
The first part of this mini-series will show that we cannot trust the authorities to provide a stable financial system. Their actions were essential leading the world to the deepest financial crash since the Great Depression.
The GFC
Like all economic crises in history, the GFC, or Panic of 2008, was not born out of a void. In the heart of the crisis was an innocent financial innovation and a genius effort to manage risks in the banking system more efficiently. Like so many times before, regulators were seriously behind "the curve" and did not understand what had been created, how it affected the financial markets and what was brewing inside the global financial system, until it was too late. To make matters worse, regulators took an active role feeding the speculation during the latter parts of the mania. The GFC truly was a failure of the whole system, not just that of the banking system.
The long road to the crisis started when the regulated commercial banking sector found a way around the financial regulation and nurtured a new "shadowy" banking system. Shadow banking is a term used to describe the part of the financial sector that operates outside of, or is only loosely linked to, the traditional system of deposit taking institutions. It provides credit intermediation through different institutions, instruments and markets. Effectively, it's a deposit-like banking system for firms and institutional investors, but because it does not accept deposits in their traditional form from consumers and corporations, it mostly falls out of the banking supervision and regulative framework. This means that shadow banks can take bigger risks and operate with higher leverage than their regulated counterparts.
In the shadow banking system, non-financial firms and institutional investors earn interest on short-term (sometimes just over-night) deposited money. These "deposits" are backed by collateral in the form of bonds which, before the crisis of 2007-2008, where usually in the form of securitized loans, obtained by commercial banks from repo and commercial paper markets. Crisis erupted, because there was a run on money markets that withdrew the collateral of the short-term deposits, and because of a cataclysmic failure in risk assessments of the new opaque (securitized) financial products. When the U.S. housing market cratered, for the first time since the 1930s, losses started to mount in the securitized financial asset universe eventually breaking the shadow banks that had took part in the speculation. Their losses fell upon commercial banks that had participated on the securitization and brokerage of the shadow banks, leading to a near-collapse of the financial system. Drastic measures of authorities averted the financial collapse, but they came at a heavy price, which will a topic of the second installment. But, how the crisis came to be?
Prerequisites
Basically since the birth of banking, banks have been in the forefront of risk distribution through diversification and hedging. In the US, banks started to sell mortgage-backed debt obligations to investors already in the 1960s. The idea was to distribute risk outside the balance sheet of banks, which would make more funds available for bank lending. In the late 1970s Salomon Brothers and Bank of America Corporation started to securitize the mortgages by pooling together their interest payments and selling investors tranches based on their riskiness. The tranches became known as: junior, mezzanine and senior. Junior was the most-risky one, but it also received the highest payments. Senior tranche was thought to be very safe. If there would be defaults on mortgages, all the other tranches would be wiped out before senior, effectively implying a collapse of the housing market. Lewis Ranieri, working at Salomon, persistently pushed through legislation opening the secondary market for the mortgage-backed securitized bonds, which before could be sold only to the government-sponsored entities: Fannie Mae, Freddie Mac or Ginnie Mae.
In the 1990's, diversification and hedging took a big leap forward when the credit default swap, CDS, was developed. In it, the risk of a loan is insured by a third party to which the bank or, generally, the issuer of a loan pays a fee for the insurance. The first known CDS contract was made between Exxon and the European Bank for Reconstruction and Redevelopment (EBRD) after the Exxon Valdez oil tanker accident in Prince William Sound, Alaska, in March 1989. Exxon was threatened with a fine of $5 billion, and it asked for a credit line from J.P. Morgan and Barclays bank. Barclays declined. Blythe Master from J.P. Morgan credit derivatives team came up with an idea of insuring the loan. She contacted the EBRD to ask whether they would insure the Exxon credit line in exchange for an annual fee paid for the insurance. If Exxon would default, EBRD would compensate J.P. Morgan for the loss. If not, EBRD would make good (practically free) profit from the fees. Because the likelihood of an Exxon default was considered very small, the EBRD agreed, and the CDS market was born.
This rather well functioning new system of risk distribution through diversification and hedging was elevated to a new level, when the derivatives team run by Bill Demchak in J.P. Morgan decided to create a shell-company, a special purpose vehicle, SPV, to insure the loans of the bank.1 The role of the SPV was effectively the same as EBRD had had in the Exxon credit line. However, the SPV was used to bundle the risk (loans) and sell them to investors according to the calculated tranches (junior, mezzanine and senior), which then received interest income based on the riskiness of the tranche in their possession. To cover against losses, SPV bought US treasuries and other AAA -rated securities. Mathematically, it was shown that the overall risk of the loans greatly diminished, when they were bundled together. This implied that the SPV needed to obtain only a small amount of collateral to cover the possible (mathematical) losses. The risk that remained after bundling was thought to be so low it was denoted as “super-senior” (implicating that its risk was even lower than that of the senior -tranche). The construct was called Bistro, a short from broad index secured trust offering, revealed to the financial community in December 1997.
Authorities did not appreciate the concept at first. When the J.P. Morgan team sought for a regulatory approval, banking authorities required that the risk unfunded by the Bistro scheme should be insured by third party. A solution was found, when the insurer American International Group, AIG, agreed to insure the super-senior risk. Official at the Federal Reserve and OCC (the Office of the Comptroller of the Currency) first demanded this insurance, but then made an U-turn deciding that banks did not need to remove the super-senior risk from their books.2 This marked the beginning of the spectacular growth of the shadow banking sector, and speculation.
After the initial stage, further innovations followed. These included the synthetic collateralized debt obligation, or CDO, and the structured investment vehicle, or SIV. The CDO was a standardized, a more general version of the Bistro, and it could be constructed not just from CDS and other derivatives, but also from different debt securities, like mortgages. The financial innovation picked up breeding ever more complex products, like the CDO squared, where the CDO was build not from bundle of loans, but from pieces of debt from other CDO's. So, essentially, it was a CDO of CDO's, and it was usually constructed of the riskiest notes from other CDOs, because they provided the highest returns. SIV:s were (are) “quasi-shell companies”, which were found by the banks to buy, bundle and sell the loan risk. Banks funded them with loans with maturity shorter than one year. This meant that banks needed not to reserve collateral for those loans, according to the Basel I regulation. SIV was thus a perfect construct for the banks. It did not require as heavy administrative control as an SPV, but it also sat outside their balance sheets, with no need to provide collateral for loans issued to it. Bankers thought they had stumbled into a 'Holy Grail' of loan risk hedging and revenue, and the result was as expected.
Supported by the low interest rates of the Fed, the housing market quickly proceeded to a total hustle. Loan originators started to issue loans without rigorous background checks. At the peak of the speculation, mortgages were issued to people with No Income, No Jobs and No Assets, as “NINJA” loans. Ordinary people speculated with the market by buying several houses using cheap credit to obtain high leverage to be sold later with an (expected) hefty revenue. The risk of these loans were bundled together and removed from the balance sheets of banks by SIV’s, which then sold this risk to investors in CDO’s, which included other banks. This also meant that a systematic risk started to build within the financial system, but this was completely overlooked by the U.S. authorities.
The securitization of mortgages rapidly developed into a full Ponzi. Many mortgage holders were unable to finance neither the principal nor the interest of loan from their income flows. They were betting on the increase in the value of their houses. Only very few people were able to see through this Ponzi, including (most notably) by Michael Burry, the team of Mark Baum and Cornwall Capital (if you have not, you need to watch the Big Short).
The Achilles heel
When the risk management team at J.P. Morgan devised the Bistro, the original idea was to remove (most of) the risk of corporate loans. This worked for two reasons:
There were extensive knowledge (a long time-series of data) on the credit-worthiness of companies.
Defaults of companies were relatively rare and they were somewhat predictable, e.g., companies with certain types of loan risk tend to default in a recession.
The ‘Achilles heel’ of the structured mortgage products was their opaque nature. While there were usually decades worth of detailed data from companies, there was often no information at all on the mortgage holders. In many cases, such information was not even known or enclosed by the bank. The housing market had also not had a serious downturn for more than 70 years (since the Great Depression), so it was impossible to evaluate how the mortgages would behave if the housing market would crash. Thus, while the risk accounting through the Value at Risk, or VaR used to assess the risk of a loss (a loan default), could be effective only in a fairly predictable environment of corporate defaults, mortgages were a completely different animal.3
Regardless of the obvious shortcomings on assessing the risk of these products, the rating agencies gave high credit ratings, especially for the super-senior risk of CDO:s (which, naturally, was also sold or “insured”). The rating agencies received higher compensation for higher ratings, which created a strong incentive to give high ratings even to dodgy financial products. In the peak of their moral decadence, agencies even gave advice to the issuers on how to bundle the loans in order to achieve the best possible rating.
Because the CDO:s offered an exceptionally high yield with high credit ratings, their demand exploded across the financial world. They were gobbled up by banks, their SPV:s and SIV:s, hedge funds and even by pension funds. Thus, supported by the high demand for CDO:s, banks were able to transfer most of the loan risk outside their balance sheets in a systematic manner, but only to other banks and other financial institutions. That is, by applying the CDO-structure to remove risk from their balance sheets, large-scale, only meant that it was transferred into the financial system in general. While this greatly increased the leverage and profits of banks, it accomplished this at the expense of stability of the financial system. And after approving the measures leading into the mania, authorities were sleep at the wheel or even took part in the speculation. The whole financial system corrupted.
Conclusions to part I
Some of the crucial errors that led to the crises were made decades before the crisis. Fannie Mae was founded in 1938 to provide funding for the housing market as a part of the New Deal by President Roosevelt. Ginnie Mae was created in 1968 to provide affordable housing and guaranteeing loans associated to it. Freddie Mac was founded in 1970 to provide “competition” to the secondary market of the housing loans, dominated by Fannie Mae at that time. It’s rather odd that a Government Sponsored Enterprise, GSE, is founded to ease the monopoly created by another GSE. It’s like fixing a fire with fire. Together Fannie Mae and Freddie Mac are said to control about half of the secondary mortgage market, where housing loans are bundled and sold to investors. All these institutions created artificial demand for sub-prime loans (issued to low-income households).
Three laws were also passed to redirect mortgage funding to poorer households. The Community Reinvestment Act (1977) mandate regulatory authors to check that banks do not discriminate their mortgage activity based on the income-status of the community. The Housing and Community Development Act (1992) forced Fannie Mae and Freddie Mac to use 30% of their capacity to acquire loans of poor and middle-income households. The American Dream Downpayment Act (2003) provided federal assistance to down-payments and transaction costs of mortgages of poorer households. While these laws and institutions did not cause the crisis, they forced funding into the sub-prime loans (with higher risks) and were thus directly causing the ‘securitization mania’ in banks, who tried to remove the risk of these loans from their books.
The Fed and the OCC green-lighted the practice of not insuring the super-senior risk creating a massive incentives to securitization and hedging of mortgages in banks. The Fed also kept interest rates too low for too long after the collapse of the Tech-bubble, feeding the housing market mania. However, the biggest factor contributing to the crisis was the (utter) corruption of rating agencies. It’s unfathomable that during the last stages of the mania, they instructed banks, SIV’s, etc., on how to bundle the loans for maximum credit rating.
The point is that, yes, banks created the precipice for speculation in the U.S. housing markets, but actions of authorities fueled it into an outright mania. Law-makers created artificial demand for the sub-prime loans and thus established conditions for large-scale hedging of these high-risk loans into opaque financial products. Moreover, authorities were completely unaware for quite some time on the monster they had unleashed. Most damagingly, both the mathematical hedging of the risk within banks and regulatory oversight failed, massively.
I think that the run-up to the GFC shows very clearly why we cannot trust on authorities to keep our financial system safe. This is for two reasons.
Authorities operate with a massive lag on what comes to the over-sight of financial innovation and risks.
Authorities have tendency to become corrupt, either through actual profits they receive from financial innovations or the prospects they create (before the crisis, some U.S. politicians hailed the financial innovations for allowing the “American dream” for many).
During the past few years we have seens central banks covering their own ass with the expence of long-term financial stability. The repeating financial market bailouts by central bankers, especially after the GFC, has created a highly levered and fragile financial system (on which more in the second installment). Like I explained, the panic of 2023 was mostly due to the actions of authorities turning the U.S. banking system fragile.
Our financial system is not currently threated by “greedy banks”, investors nor speculators, but authorities and especially central banks. They have failed us again, and much worse is likely to come.
Tuomas
Large parts of the history of derivatives development (CDS, OCD, etc.) are from Gillian Tett’s: Fool’s Gold.
1/11/2025: Small updates. 1/15/2025: Updates on footnote 3
Disclaimer:
The information contained herein is current as at the date of this entry. The information presented here is considered reliable, but its accuracy is not guaranteed. Changes may occur in the circumstances after the date of this entry and the information contained in this post may not hold true in the future.
No information contained in this entry should be construed as investment advice. Readers should always consult their own personal financial or investment advisor before making any investment decision, and readers using this post do so solely at their own risk.
Readers must make an independent assessment of the risks involved and of the legal, tax, business, financial or other consequences of their actions. GnS Economics nor Tuomas Malinen cannot be held i) responsible for any decision taken, act or omission; or ii) liable for damages caused by such measures.
These were usually off-shore companies setup in the Cayman Island’s or Bermuda.
They did not let the banks completely off-the-hook, though. Banks needed to show that the risk of default on the super-senior portion of Bistro was negligible. Moreover, the notes issued by Bistro structure needed to carry an AAA credit rating from from a nationally recognized credit-agency. While the conditions seemed strict, they were earth-shattering for bank profits. For example, a $10 billion corporate credit line would usually require $800 million of reserves, but with the Bistro this dropped to $160 million, a drop of 80%!
In VaR, historical data is used to calculate, how much money a bank could lose from its loan portfolio assuming certain levels of probability, which were (are) based on an assumption of normally distributed losses. This effectively meant that the data on gains and losses from mortgages of past few years formed a distribution, which seemed to follow a hump-shape form, or so called 'normal distribution'. By using, for example, the 5% measures of statistical significance, any observations that were beyond the 2.5 percent threshold in each side of the distribution were deemed "highly unlikely". This practice made the CDO’s look safe, because the models assumed that their loss-distribution would follow a normal distribution, while in reality it was something very different altogether. This was a crucial mistake.
The main requirement in basically any macroeconomic risk-assessment is that one needs a time series of all kinds of states of the world, including depressions and asset market crashes, to make a reliable assumption of the underlying statistical distribution (mirroring the risks of the underlying assets and their income streams). This kind of data was utterly missing for the U.S. housing market at the time. Moreover, copula models, used to model cross-correlation of assets, require data on all states of the world (including crises). Because the last housing bust had occurred during the Great Depression of the 1930s, such data was not available either.
re: "They were betting on the increase in the value of their houses."
Means-of-payment money was held constant for 4 years, i.e., a contraction not seen since the GD. Legal reserves were drained for 29 contiguous months, turning safe assets into impaired assets.