[Engineering the Financial Crisis: Systemic Risk and the Failure of Regulation ♦ by Jeffrey Friedman and Wladimir Kraus ♦ University of Pennsylvania Press, 2011]
There have been a good deal of books attempting to find and explain the causes of the ongoing financial crisis. Authors have approached the issue from all sorts of ideological perspectives and with different sets of evidence. Most of these works are lacking, incomplete, or even flat-out wrong. Many of them do not even care for the facts, instead using vague generalizations to justify the application of broad economic theories. There has not, until recently, really been a meticulous analysis of the mechanics of the causes of the Great Recession, despite the enormous interest displayed by the economics profession in the subject.
This lacuna has been filled by Jeffrey Friedman, editor of Critical Review, and Wladimir Kraus, in their recently published book: Engineering the Financial Crisis. The authors make the purpose of their study evident from the very beginning. They shed themselves of any ideological priors which may have otherwise impaired their analysis, even going as far as to disprove a number of general theories from either side of the spectrum (insufficient regulation versus insufficient economic freedom), and task themselves simply with accumulating, analyzing, and interpreting the evidence. The data they look at has to do with the regulations which governed the financial institutions which presided over the network of financial instruments which suddenly lost the bulk of their value. The question they ask is a simple one: based on the facts, was the recession caused by under-regulation or was it something in the regulation itself which may have influenced the ways banks invested?
Friedman and Kraus give reason to believe that it was the latter — perverse regulations — which gave way to the great contraction which took place between 2007 and 2009. Looking through the relevant legislature which dictates the laws governing the banking industry, the authors find that it was this regulatory web which led banks to invest into the specific financial assets that would soon after be deemed nearly worthless.
Friedman and Kraus emphasize the importance of the fact that the banking industry had no idea — what they call “radical ignorance” — about just what kind of quagmire they were investing themselves into. They use evidence to illustrate the fact that, predominately speaking, the bankers, regulators, politicians, and other major actors in this crisis had absolutely no idea of the relevant potential for a recession to occur, let alone that the highly rated assets they invested into would soon become relatively valueless.
We can see now the broad thesis of Engineering the Financial Crisis. Bankers did not buy large amounts of soon-to-be “toxics assets” because the risk had been externalized to the taxpayer.1 Neither is there any evidence suggesting that bankers purposefully ignored high risk in favor of pursuing high profits. The majority of assets purchased were actually AAA rated, and because of this the risk-load they carried, as perceived at the time of purchase, was relatively low. What manipulated the relevant price signals which funneled investment into the housing market were regulations which awarded these type of investments. To a lesser extent, the authors also point at government programs which pushed for house ownership and the relatively low rates of interest on new loans which made borrowing seemingly more affordable.
Leading up to the crisis, bankers were generally very risk sensitive, preferring assets with lower revenue returns. Roughly 93 percent of mortgage bonds held by U.S. Commercial banks were AAA-rated mortgage backed securities (both private label [PLMBS] and agency rated [MBS]), and almost another 7 percent were AAA-rated collectivized debt obligations (CDO).2 Missing from this collection of assets were any mortgage bonds rated less than AAA, which were usually the bonds which the highest rate of return.
Most commercial banks in the United States were also above their legal capital reserve minimum on the eve of the financial crisis — the twenty largest banks held capital levels averaging 11.7 percent, where the legal minimum was 10 percent of a bank’s total assets (as dictated by the Federal law, whereas the Basel I accords had set it at 8 percent). More specifically, if one only accounts for “Tier 1” capital3 banks retained a capital cushion of 50 percent, even while federal law required a minimum of 5%. In other words, most banks opted to retain a substantial capital cushion, where one would expect a bank interested in maximizing profit (while ignoring risk) to push the boundaries of its legal requirements.4
The issue, then, was not about bankers with low risk aversion, seeking high profits by investing mostly in high-risk assets. The evidence suggests quite the contrary, actually. U.S. commercial banks invested in what were perceived as low-risk, low-return assets, and on top of this held higher than required capital cushions. Also going out the window is the “too big to fail” theory, or any other case that argues that it was risk externalization which created an incentive for an over-concentration of investment into the mortgage market. Simply put, there was a high aversion to risk during the years leading up to the crisis.
Friedman’s and Kraus’ explanation of what caused the crisis can be divided into two parts: what caused the over-concentration of investment into mortgage-backed securities and why these securities, which by 2008 had lost the majority of their value, had been rated so highly by the major rating agencies. The strength of their book is found in its accounting of the first — what led to the pattern of investment that characterized U.S. commercial bank assets prior to the recession.
Because different types of investments generally have different degrees of risk, a capital requirement minimum that encompasses all assets is illogical. Rather, it makes more sense to create different capital reserve requirements for different sorts of investments, based on the general perceived riskiness of the different types. The Basil I accords was an attempt to correct the issues of a homogenous treatment of assets by creating different categories and attaching a capital reserve minimum to each category. Higher risk assets, therefore, were categorized into higher risk “risk buckets” and required a greater capital cushion. In other words, the greater the risk the asset carried the more it cost the banks to protect against, by reducing the amount of capital available to invest.
The issue, as explained by Friedman and Kraus, is that these regulations led to “regulatory arbitrage”. For example, a bank could invest into mortgage bonds with a risk-weight of 50 percent, then re-sell these bonds to a government sponsored enterprise (such as Freddie Mac and Fannie Mae), and buy them back as an agency bond. These agency bonds were risk-weighted at 20 percent, effectively reducing the capital cushion necessary to back the asset, even though the composition of the asset remained exactly the same. The Basel regulations made it more expensive to issue business loans than home loans, creating a financial incentive to issue more home loans. Basel I created incentives for banks to make certain types of loans and then securitize them.
A further boon to securitization came with the adoption of the Recourse Rule, which borrowed the Basel II accord’s method of rating privately issued securitized assets by risk. This system caused banks to increase investment in AAA-rated securitized mortgage bonds, especially since the risk-weigh of unsecuritized mortgages remained at 50 percent (as dictated by Basel I). It remained cheaper to invest in mortgages, rather than other types of loans to consumers and businessmen, and then banks could further increase profitability by securitizing these mortgages and releasing part of their capital reserves for further investments. This explains the concentration of investment in mortgage backed securities.
We see a pattern between 2001 and 2007 of an increase in housing loans and investment into mortgage backed securities. We know that at the time that these types of investments being made were being pooled into buckets which were considered generally less risky than other forms of investment. It was not an issue, therefore, of carrying on more risk. In fact, this pattern of investment was created out of the fact that the regulations incentivized purchase of less risky assets. The problem which led to the financial collapse, therefore, deals exclusively with the fact that these assets carried more risk than was originally perceived by the regulators (and banks). In fact, the collapse of the subprime mortgage market came as a total surprise, both for the bankers and the regulators.
It was not just the architecture of the impending financial collapse that the regulatory web was responsible for, but also the magnification of the disaster. Thanks to the capital reserve minima, many banks faced insolvency even though the circumstances did not really call for it. In order to remain legally solvent, U.S. banks are forced to maintain a certain capital reserve minimum. As the crisis unfolded, bonds which had been previously rated at AAA were suddenly downgraded, raising the necessary capital reserve minima for each risk-bucket. In other words, as ratings fell for different types of bonds, banks were suddenly forced to raise new capital to cover their loss.
Furthermore, regulations forced banks to mark-to-market their assets to reveal their “true value”. This process was not done on an individual basis; rather, different bonds were lumped together and them marked-to-market as a group. So, even individual bonds which may have not actually lost value were readjusted on a bank’s balance sheet as an asset that was suddenly worth less than it had been perceived prior to the crisis. It was on the basis of these new market values that a bank’s solvency was judged at. The issue is that by late 2009 many of these same bonds had recuperated much of their value, and so a bank that had been legally insolvent in early 2008 may not have been two years later. In other words, many banks were forced into insolvency that could have survived the crisis, and other banks had to radically contract outstanding liabilities in order to remain solvent.
The consequence of these regulatory restrictions was a giant credit contraction — much larger than was actually necessary. And, of course, the monetary contraction only worsened the situation, as it reduced the financial viability of the various investments that depended on this credit.
Friedman and Kraus blame the inadequate rating of mortgage bonds on the cartelization of the major rating agencies — Moody’s, Standard & Poor’s (S&P), and Fitch. Basel II and the Recourse Rule had effectively tied their capital reserve requirements to the ratings provided by these three agencies. It was these three rating agencies that had been classified as Nationally Recognized Statistical Rating Organizations by the Securities and Exchange Commission (SEC) in 1975, and the various regulations that relied on risk ratings depended exclusively on this cartel. None of the three “nationally recognized” rating agencies, furthermore, had accounted for the possibility of a nationwide hosing crisis leading up to 2008.
There were private rating agencies, though, that had recognized the potential for crisis. According to Friedman and Kraus one such company was First Pacific Advisors, which sold its $1.85 billion investment in mortgage-backed bonds in late 2005.
Friedman’s and Kraus’ analysis hinges on the notion that what was ultimately at fault were these rating agencies. Had they been more accurate in their risk assessments then banks would not have malinvested in so many mortgage-backed bonds. But, why had the banks relied exclusively on the risk-assessment provided by the three “nationally recognized” agencies? Were these banks not aware of the risk assessments being made by private investment companies?
Perhaps the authors put too much weight on the notion that it was this cartelization of the rating agencies which made possible the crisis, and that had there been more competition in this industry the recession may have been less destructive than it turned out to be. But, Friedman and Kraus do not make clear exactly how many private agencies had foreseen the crisis, or at least were aware of the likely possibility of an impending crash. It seems as if the number of investors who had sold off their mortgage-backed bonds in the face of a threat of a market-crash were in the few. Indeed, the majority of investors were still fairly confident in the strength of the housing market.
Here is where the explanation of the crisis becomes more detached from the data. Friedman and Kraus leave room for further interpretation, since their explanation for why the different bonds were assessed as they were comes off as inadequate (or, at least, incomplete). Understandably, they are looking to separate themselves from ideology — even though the book’s conclusions are extremely pro-market — and thus avoid applying far-reaching theories to the evidence they were able to collect.
However, that there is still room for further interpretation is not necessarily a bad thing. As far as their analysis on the impact of regulations on the housing boom and the consequent financial crisis goes, it is difficult to refute. That there is still room for the application of theory means that their empirical findings can easily be assimilated into grander explanations of the financial crisis.
For the task it sets out to accomplish, Engineering the Financial Crisis is undoubtedly one of the best books written yet on the causes of the Great Recession. Jeffrey Friedman and Wladimir Kraus painstakingly dig through the data to provide a solid picture of why there was such an over-concentration of investment in the mortgage market. The evidence clearly shows that it was the web of regulation on the banking industry that shaped the structure of banking investment by favoring certain investments over others. The entire system collapsed when it turned out that these regulations had depended on agency assessments that had totally miscalculated the risk these favored assets carried. Thus, banks had loaded themselves up with mortgage backed bonds, completely unaware of the fact that these bonds would soon be relatively valueless. It was not the market which caused the crisis, rather the distortions to the market that were created by government intervention.
1. That banks did not buy mortgage-backed securities and other similar assets because the bankers knew that there was no risk for them is a very specific claim, and it does not include many banking practices which are undertaken because of risk externalization (such as the extent of fiduciary expansion, which in our present banking system is a product of its cartelization under the Federal Reserve System).
2. Friedman and Kraus 2011, p. 42 (table 1.3).
3. Basel I divides bank capital by the type of asset, Type I being the safest pool. Type I is composed of “funds received frm sales of common equity shares and from retained earnings.” Ibid., p. 40
4. A capital reserve basically allows banks to take losses, since it gives it a cushion of assets it can capitalize on to make up for net losses (before liabilities exceed assets).