Abstract
The existing literature on modelling provides two main ways of viewing model migration: a modular view, which seeks to decompose models in their constitutive elements, and thus provides a view on what it is that migrates; and a practice-based view, which focuses on modelling as an activity, and understands a model as intricately entangled with its context of use. This article brings together these two sensitivities by focusing on ontologies of modelling. The paper presents a case study of the appropriation of modern finance theory’s ‘no-arbitrage’ models by British actuaries – a process that gradually unfolded at around the turn of the century and led to significant friction within the UK’s insurance industry. We can distinguish two main modelling ontologies: a ‘risk-neutral ontology’, which underpins no-arbitrage models and holds that the value of financial instruments is determined by ‘arbitrage’; and, a ‘real-world ontology’, which assumes that the economic world consists of real probabilities that may be approximated through a combination of archival-statistical methods and expert judgment. The appropriation of the risk-neutral modelling ontology was made possible by the declining legitimacy of actuarial expertise as ‘financial stewards’ of life insurance companies. The risk-neutral modelling ontology provided an ‘objective’ alternative to the traditional actuarial models, which explicitly required actuaries to make ‘prudent’ judgments. Despite the fact that the no-arbitrage modelling was considered an ‘objective’ affair, the valuation models that insurers use today are strongly shaped by political compromises, a result of the ‘rough edges’ of models.
Highlights
Simulation models are increasingly identified as an important interface between academic research and policymaking (Svetlova and Dirksen, 2014; Van Egmond and Bal, 2011; Van Egmond and Zeiss, 2010)
I argue that two modelling ontologies can be distinguished: a ‘risk-neutral ontology’, which underpins the no-arbitrage models developed in modern finance theory; and, a ‘real-world ontology’, which assumes that the economic world consists of real probabilities that may be approximated through a combination of archival-statistical methods and expert judgment
The purpose of this paper was to consider the problem of model migration from the perspective of empirical ontology
Summary
Simulation models are increasingly identified as an important interface between academic research and policymaking (Svetlova and Dirksen, 2014; Van Egmond and Bal, 2011; Van Egmond and Zeiss, 2010). In the development of insurers’ market-consistent models, three main types of rough edges can be distinguished: (1) those due to differences between the objects to be modelled, namely insurance contracts and financial instruments; (2) calibration problems that emerged from the limited range of financial instruments that are traded in financial markets; and (3) the emergence of phenomena that seem to influence the value of an insurance liability, but that cannot be accounted for within the world of the model. These problems concern phenomena that seem to influence the value of insurance liabilities, but that cannot be accounted for in what one of my interviewees described as the ‘theoretically pure’ world of no-arbitrage modelling This is because practical implementations of the ontology of no-arbitrage require idealizations that make the models mathematically, or, in most cases, computationally tractable (see for a list of idealizations identified by actuaries: Sheldon and Smith, 2004: 590). They must do so, in ways that are conceptually defensible even if methodologically rather arbitrary and, according to some, problematic
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
