Collateralized Loan Obligations (CLOs) are securitized structured credit derivatives backed by risky corporate loans. They share features with the Collateralized Debt Obligations (CDOs) that blew up the financial system in 2007/8. By the end of 2018, the outstanding value of CLOs reached $616bn (SIFMA 2018) – an amount roughly equivalent to the CDO market in 2006. This has raised concern about the capacity for CLOs – and the corporate loan markets they shape – to cause another financial crisis (Bank of England 2018; Federal Reserve 2019)

This project marks the first data gathering phase in a larger programme which examines risk in the CLO and corporate loan market. Our programme hypothesis, drawing on social embeddedness theory (Granovetter 1985), is that financial crises can emerge from the quotidian sociology of exchange in security structuration networks: that repeat relations nurture systems of reciprocity and concession-giving which reduce costs for incumbents but erode standards. Those cost savings, in turn, force out competition from suppliers of better, more expensive collateral, leading to network concentration, normative homogenisation and a growing tolerance for ever-riskier collateral, which feeds back into primary asset markets like corporate loans. In such a context ‘the bad drives out the good’ making this a ‘network for lemons’ - a modification of Akerlof’s (1970) formulation. This thesis is informed by the 2011 US Senate report which identified concession-giving within the CDO structuration network as a cause of the crisis.

This project will use data-mining techniques to build a unique database of structuration participants (banks, law firms, trustees, collateral managers, credit rating agencies etc) in pre-2008 CDOs and post-2008 CLOs from a hitherto untapped resource: ‘offering circulars’. We will later analyse that database using social network analysis (SNA) to assess similarities between the two networks, analysing the effect of social network change on security quality.