CRASSH

Conditional Value at Risk (CoVaR) is a newly proposed macro prudential risk measure to capture the  systemic risk and it is claimed to overcome some problems of traditional VaR that might have caused recent financial crisis. Accordingly, this paper develops a framework for answering the following questions about CoVaR: 1) How can a risk manager form analytical CoVaR that is readily employable? And 2) Under such analytical framework, how does CoVaR capture the systemic risk? The usefulness of the  methodology is illustrated assuming returns follow bivariate log Ornstein-Uhlenbeck process. In this example, VaR and CoVaR measures are derived then tested to assess the consistency of VaR and CoVaR between the normal and crisis periods. The result shows that the analytical CoVaR could be a  good  complementary tool to the existing financial risk infrastructure and therefore would be able to help  regulators and risk managers in governing financial risk.

Construal Level Theory (Trope & Liberman, 2003) relies on the spatial metaphor of distance to describe how people represent events and circumstances: Psychologically “close” events are construed concretely, whereas psychologically “distant” events are construed abstractly. Varying construal level leads to critical consequences in human decision making. For example in economic decision making, people might prefer a smaller reward now to a bigger reward later because the “now” option is close and concrete, whereas the “later” option is far and abstract. We propose that such inter-temporal choices constitute a conflict because they involve comparing options that are incongruent regarding their level of construal. Thus, framing options within the same level of distance should encourage more adaptive, patient decisions. In Studies 1 and 2 we made the “later” option more close and concrete by inducing participants to mentally simulate it. While participants in the control conditions made choices between receiving a cheque (e.g., £460) right now or a somewhat larger amount later (e.g., £500), participants in the Congruent Construal conditions considered receiving a travel voucher for Paris either now or later, and those participants made more patient choices. Based on the same rationale of equating construal level, in Studies 3-4 participants made intertemporal choices with “now” and “later” options metaphorically phrased as distant and abstract. For instance, instead of making choices for themselves, participants made decisions for other people, and this way of framing again resulted in more patient choices. Overall, the current studies show that economic decision making can be improved by providing people with the appropriate metaphorical distance and its corresponding construal level.  

This paper will argue that within the UK discussions of uncertainty regarding climate change science by scientists, policymakers, and the public are fundamentally and necessarily incomplete. This is not the result of malice or ignorance, but is instead an inevitable outcome of the use of advanced statistical analysis within climate science in the form of, for example, models of physical laws or simulations of global climate. Such a form of analysis lends itself to quantified summarisation which appears unobjectionable but in fact strips out important 'local' qualitative knowledge held by those intimately involved in the construction of such analyses. This knowledge is particularly important for ensuring informed consideration of the uncertainty surrounding such models and simulations. This loss of knowledge can be seen to occur within the climate science research community itself and becomes more severe when such knowledge is interacted with by policymakers and the public. Recognising this intrinsic absence permits a critical evaluation of the policy process, its outputs and justificatory narratives.