This is the first part of a three-part series on the regulation of securitization before and after the crisis. Why were the banks so affected by the run on the shadow banking sector, which was formally off-balance sheet? How can we explain the lack of regulation in the shadow banking sector? How did governments promote the expansion of the credit supply via the shadow banking system?
Securitization has had a very bad press reputation in recent years (but see “securitization is not that evil after all”), being related to overly complex re-securitizations, which became impossible to value during the financial crisis of 2007/2008. Before that crisis, securitization was en vogue, favored by most financial economists as a way to spread credit risk from banks into the financial markets, thereby increasing the resilience of the financial system as a whole (s. Bhattacharya et al 1997). The idea was to liquefy credits, to turn them into tradable assets, generating deep secondary markets in which traders could constantly readjust the amount of risk they held in their portfolio. Banks could refinance the credits they gave and thereby expand their lending. Credit would become cheaper, as the demand for securitized assets generated from credit increased, raising the available supply of credit.
Turning credits into tradable assets requires that the transfer of these assets into money is possible instantaneously and without loss of value. Otherwise, the entire idea of readjusting one’s portfolio to changing market circumstances; which is what underlies modern portfolio theory (also the Black-Scholes formula requires continuous adjustment of the traders position, which is why trading in continuous time is a necessary assumption for the pricing to work). Credits on banking books, in contrast, are held on the book of the banook at historical cost accounting (s. Sigrids work on fair value accounting); i.e. not changing their value from the contracted valuethe moment the contract is signed. Only if banks undertake corrections in value to account for expected losses does the value of these credits in the books of the bank change.
In this blog I will look at the infrastructural preconditions of securitization (Special purpose entities, in the following SPEs) and the shifts in how they were proposed to be treated at an international level and how they were treated on a national level before the crisis and after the crisis. Before the crisis, there was pressure by the Committee of European Banking Supervisors to not force these SPEs on the balance sheet of banking groups (CEBS 2004). After the crisis, we can witness a 180° shift in the position of the financial stability board which states that it wants full prudential consolidation for sponsored SPEs. In my three blog entries I will have look at the impact these transnational recommendations had before the crisis on actions of national governments and will speculate about the fate of the current regulatory proposals in the future.
Going beyond the common presentation of the credit crisis: securitization and the banks
As several sociological authors and economists have remarked, liquidity always carries its “other” (Langley 2010, Carruthers 2009, Amato and Fantacci 2011), illiquidity. Liquidity needs constant continuous pricing of assets. As markets Tthe lower the volatility of these pricing movements, the easier to handle these assets in a portfolio. The moment the pricing offor securitized credits broke down, as markets fell dry, these assets became glued to the balance sheets where they stood at that moment, requiring write-downs.
This is part of what explains the reverberations of the rise in default rates in segments of the US market in global financial markets. Falling values of assets were hurting balance sheets, but the worst was that liquidity had come to a halt and central financial actors could not purge their balance sheets from these assets at a knowable price, but instead had to hold them with incalculable losses. In the old system, it would have wiped out the mortgage banks engaged in these credit segments and possibly some of their largest lenders, but the crisis would have had a slower pace and would have been more contained.
Focusing on the organizational infrastructure of securities markets
This is a legitimate story of the financial crisis and it points to the incapacity of the financial system to handle increasingly complex products; it points to information asymmetry and moral hazard in an originate-to-distribute model. However, this story remains at the surface level of trading and it thereby risks missing the underlying infrastructure of the markets observed, which indeed is complicated (s. first picture below from the BCBS, second picture with my additions to show the real linkage between banks and SPEs).
Bankers were not simply selling their credits to financial markets, but instead to shell companieswhich they had set up themselves and, which they controlled all but in legal form, and which they using themed to generate extra fee-income. Paying attention to these shell companies, called special purpose entities (SPEs) will also allow us to explain better why banks were and are at the center of the financial crisis.
If the credit-risk transfer mechanisms had functioned as envisioned and all credit risk had been transferred out from the balance sheets of banks, the largest losses should have materialized anywhere else but in the banking system. However, this is where they arose, of all places (s. IMF April 2009, p. xii speaking of $2.6 trillion losses concentrated in banks), necessitating large bail-outs.
Besides the increased importance of the wholesale financing market for banking in the developed world (with UK banks financing 50% of their operations via wholesale finance), which increased the vulnerability of the banking systems in these markets, the biggest reason for the large impact the crisis in securitized credits had was the close relationship banks had maintained with the assets they had securitized. These linkages led to the reappearance of large parts of securitized assets credits on the balance sheet of banks, credits which had been sold before to the shell companies. Tthe largest example in this respect is being provided by Citibank, where $49 billions of assets reappeared (s. Acharya and Schnabl 2009, 2010).
This reappearance and the holding of securitized credits es of other banks in the banking sector itself led to the concentration of losses in the banking sector. In addition, this reappearance cast wider doubt in how far the banks had hidden further potential losses from their balance sheets, as assets appeared on the balance sheets of banks; which before were deemed as sold to the market. In this sense, the crisis was also very much a crisis of confidence into the balance sheets of banks, which explains the reluctance of banks and other lenders to lend to other banks.
But: In how far did the balance sheets of banking groups represent the actual risks banks were exposed to and how much was lurking off-balance sheet?
In order to understand these legitimate concerns of investors, we need to pay attention to the infrastructural preconditions of securitization. Doing so reveals how precarious this infrastructure is and how much aid of the state was given in order for securitization to work. Securitization is inextricably linked with the shadow banking sector, and the way the crisis was resolved also points to the power balance between banks and the shadow banking sector.
Securitization, the transformation of credits in the banking books of banks into tradable assets in the trading books of banks or other financial market agents requires not only the contractual reconfiguration of the cash flow generated by these credits, which give value to the new assets but also the change of ownership. In order for banks to be able to eliminate these credits from their balance sheets and to receive cash which they could reinvest into new credits, they had to sell them to a special purpose entity, which was bankruptcy remote. This means that the creditors of the bank did not have any right to these assets in case the bank goes bankrupt.
The special purpose entity, often in the legal form of a trust used these credits as assets, refinancing them by issuing debt instruments, to which they allocated the cash-flows from the credits. The transformation of contracts thus also required a transformation of ownership. What is remarkable about Special Purpose EntitiesSPEs, in contrast to banks is that they have almost no employees, make quasi no important business decisions and hold almost no equity to speak of. The actual decision making is made beforehand by the bank initiating these trusts, which specifies in the contracts what the trust is allowed and not allowed to do and also how much it has to pay for the ongoing services of the bank to the SPE. By transferring the assets to SPEs, the banks could achieve regulatory capital relief while at the same time generating fee income from the SPE.
Dutch Special Purpose EntitiesSPEs in the business of securitization usually held 18000 Euro as equity, when the assets they held were worth 500 000 000 Euro. Its equity to capital ratio was 0.000036%. From this construction, it is clear that the special purpose entity was not capable of withstanding the slightest unexpected credit risk. This credit risk therefore was contractually distributed to the bond-holders. In case the rate of default increased over a certain trigger, the assets were to be liquidated immediately and the generated sum paid out to bond-holders as the contracts specified.
Liquidity as the achilles-heel of the shadow banking system
This is how the issue of credit-risk was contractually dealt with. But this left the issue of liquidity risk which became the more central in the business model of special purpose entities, the greater the maturity mismatch between assets and liabilities. When Special Purpose Entities were used to refinance short term credit with short term debt, there was no maturity mismatch and liquidity risk was almost zero. But when long term assets, such as mortgages were refinanced with 60 day commercial paper, this meant that the maturity mismatch was huge and accordingly liquidity risk was great.
At the same time, the greater the maturity mismatch, the greater the interest rate differential between assets and debt and therefore the greater the gain to be had from undertaking this refinancing. But how was this liquidity risk dealt with? By asking banks to provide liquidity lines which could be used if refinancing in the markets was not possible. These liquidity lines generated fees for banks, and given that banks could determine how much SPEs had to pay, banks could appropriate approximately the entire interest margin SPEs generated.
Maturity mismatches were not regulated for special purpose entitiesSPEs, and so the temptation was huge for banks to use these vehicles to engage in large scale maturity transformation. At the same time, the bigger the maturity mismatch, the greater the probability that the liquidity line would be drawn and the bank would have to take up most of the debt instruments issued by the SPE.
However, before the crisis, liquidity in the markets for debt instruments was ever-increasing, making the probability of illiquidity seem highly unlikely. In this sense, banks were betting on the non-occurrence of a tail-event, while selling insurance on it.
In this sense, banks were still maintaining some exposure to these assets, while officially transferring them from their balance sheets. Or as Gorton and Souleles (2006) put it, they were exchanging their exposure to these assets with their exposure to these shell companies. This official status was important to the banks in order to reduce the capital requirements for their balance sheets, so that they fought all accounting standards that might threaten that off-balance sheet status of these shell companies like the SIC 12 of the IASB and its application nationally (s. Larsson 2008, for a further discussion s. Thiemann 2011, forthcoming).
Regulatorypermissiveness in Europe 2004
Interestingly, domestic finance ministries largely sided with the banks arguing for light touch regulation. But also the Basel Committee for Banking Supervision as well as the Committee for European banking supervisors issued a non-binding proposal on prudential filters (s. CEBS 2004) aligned themselves with these interests, requesting a prudential (!) filter that would remove these assets of SPEs onfrom the balance sheets of banks for prudential regulation if they were captured on the balance sheet of banking groups for financial accounting purposes a true sale had occured. But why did they do this? On the one hand one can say that they were simply expressing the spirit of the time in which securitization was seen as making the financial system more stable, so the more securitization the better. Next to that, an uneven global accounting framework meant that banks in some legislation would have to withhold regulatory capital for these assets while others did not. For example in the most significant market for securitization, the United States of America, the rule for qualifying SPEs (FAS 140, in force since September 2000) allowed all securitization vehicles to be off-balance sheet, thus pursuing a prudential consolidation of SPEs without changing US accounting rules in accordance with IAS 27 and SIC 12 would have imposed costs on European banks, but not on American ones. This competitive disadvantage might have convinced the Committee of European banking supervisors to favor this common exception rather than strict rules only for their banking groups. Banking groups in Europe were pointing out this disadvantage to their advisors repeatedly.
As I show in my forthcoming paper, these recommendations left large national policy space to European financial policy makers if they should pursue prudential consolidation of these SPEs or not. How this one was used, which effects it had on the national banking systems during the crisis and which regulatory consequences were drawn on an international level after the crisis will be part of my next blog entry.
Matthias Thiemann is a PhD candidate in the Sociology Department of Columbia University and has worked as a UN consultant on the question of financial market governance for emerging countries. His dissertation deals with the regulation of the shadow banking sector before the crisis, in particular the ABCP market which involved special purpose entities.