Tools vs. Interests? How the GSE’s brought risk management technology into mortgage finance…
August 19, 2009
Daniel’s last post draws a clear distinction between a) an institutionalist approach to markets as an expression of political agendas; and b) the science studies approach to technologies as generators of unpredictable yet inherently political forms of order.
I’d like to contribute some observations on how the study of market making devices can take the involvement of both states and firms seriously – but without following the institutionalist argument that markets are sheer expressions of either political power or private interests.
Securitization did, indeed, begin life in the US as a means of attracting private capital for the public purpose of increasing homeownership. It was government sponsored because extra guarantees were necessary to entice investors into buying these types of products.
The puzzle of the current crisis, however, is this: The liquidity freeze was triggered by rising defaults in the subprime sector of housing loans, loans that by definition fail to meet Freddie and Fannie’s definition of a sound or prime investment grade loan.
From an SSF perspective, the question is: Starting in the late 1990’s how do loans initially defined as non-investment quality by the GSEs, become investment grade paper that was freely underwritten by private label players and actively traded? What technical change is put in place to permit this transformation to occur?
Here is one way to unknot this puzzle. Although securitization is decades old, the ability to securitize subprime loans exists based on the ability to assess mortgages according to a quantified and gradated measure of risk. This ability comes from a modification in underwriting practices that is relatively new.
As I have traced in a recent paper [also here], in 1995, Freddie and Fannie adopted a statistical risk management standard called a FICO score, marking 660 as the threshold to redefine the prime market. These commercial credit scores gradually became a standard in mortgage finance.
Working from the information endorsed by Freddie and Fannie, private industry built up a flourishing circuit of lending by acting on FICO<660 and jumbo loan segments. In other words, once expressed quantitatively, low scoring loans became manageable risks.
How did the government enterprises participate in the recent real estate bubble? Not only through ‘politics’ as generally understood (i.e. through directive policy statements), but also through their role in the adoption, interpretation and dissemination of specific risk management choices.
What is striking is that in the shift from rule-based to statistical underwriting the GSEs put in place the groundwork for subprime lending, but they did not anticipate or promote its rise. The proof is that the GSEs, unlike private players, were only very weakly involved in underwriting subprime securities.
How does private industry come to express its voracious appetite for growth and risk in mortgage finance? Not through the spontaneous expression of naked self interest, but rather, by being parasitic upon a change in the information infrastructure of the industry, initiated and validated by the GSEs.
Subprime lending, the trigger of the contemporary credit crisis, has involved the delicate intertwining of both government and private actors through risk management technology. It is, I think, a strong example of the kind of innovation mediated politics – the creation of new financial order through achievement of common technical platforms – that SSF explores.