Calculating at the Frontier: Coordinating the Reinsurance Market

This paper illustrates market coordination through a study of the calculative practices that provide the basis for reinsurers to allocate capital to reinsurance deals. For some types of reinsurance risk standardized vendor models help coordinate the evaluation process amongst dispersed risk-traders in the absence of any electronic or physical connection (Knorr Cetina and Bruegger, 2002). Yet, for many other types reinsurance risks vendor models are not available to inform risk calculation. For example, reinsurance deals from many territories have insufficient data to be modeled while other types of deals have underlying causes too complex to be standardized. Yet, the calculative purpose of the market remains the same: to turn these risks into tradable comparable objects to which reinsurers can allocate capital (Carruthers and Stinchombe, 1999; Çalışkan and Callon, 2010).

Through our global ethnography of the reinsurance industry – which includes 382 interviews and 935 observation fieldnotes across fifteen countries - we explore the practices of marketization (Çalışkan and Callon, 2010) and calculation (Beunza et al., 2006; Callon and Muniesa, 2005) of various types of reinsurance deals. The paper first introduces variation in deals stemming from the many different risk-types that characterize reinsurance. Risk-types can be differentiated according to the modelability of deals. Second, we then explain how multiple epistemic cultures (Knorr Cetina, 1999; Knorr Cetina and Preda, 2001), each pertaining to a particular risk-type, enable the calculation amidst this variation and in the absence of models. We show that underwriters specialize in different risk-types, which makes them part of specific global epistemic cultures (Brown and Duguid, 2001), each of which has deep understanding and shared practice in calculating deals within a particular risk-type. Third, we identify the practices of technicalizing and contextualizing, to show how members of different epistemic cultures enact the specific calculative practices that make deals within each risk-type tradable. Different epistemic cultures each generate, apply and blend those calculative practices in ways appropriate to evaluating each of these less-modelable risk-types. We will explain how, despite their eclectic calculative practices, underwriters in these different cultures generate the same type of output, namely, a pricing quotation expressed as a ‘Rate on Line’. This enables these vastly different deals to be compared and traded in a broadly consistent way.

We conclude by outlining how each epistemic culture helps to coordinate the market for reinsurance deals. Each risk-type constitutes a particular site of calculative activity for marketization. This enables dispersed market actors to be connected through their collective practices as they construct prices within the market, despite an absence of electronic or physical connections (Knorr Cetina and Bruegger, 2002). We develop a relational, practice-based, framework to further our understanding of market coordination. More specifically, we also address how marketization is possible when information is sparse, rendering the typical calculative practices of a financial markets, such as financial models (Millo and MacKenzie, 2009), less applicable. Namely, we show how such risk-types are rendered tradable through the practices of their particular epistemic culture.