Correlation and dependency modeling
Clearly, companies are investing now in building their EC framework to give their businesses the best possible foundations for long-term advantage. The costs of this investment are high in the short term, but as any architect will confirm, investing in solid, appropriate foundations at the outset is the best indicator of future stability.
Dissatisfaction with this modeling reflects the lack of hard data to analyze how risks are interlinked and how chans of events can occur together. In the absence of data, the majority of respondents are using expert judgment and simplified techniques such as correlation-based approaches to aggregation.
Refinements
Even after after the initial calculation of EC, there are a number of adjustments or refinements that may be required to reflect items that may not be adequately captured in the model. Appropriately allowing for fungibility constraints is critical to ensuring that EC is not understated by taking credit for capital movements that would be difficult to achieve in practice. However, it is a major challenge due to the complex capital structures that companies have typically had in place.
Such arrangements have normally been developed to improve tax efficiency, or as part of historic mergers & acquisitions activity, but they add to the complexity of EC computation. Advanced companies have started to include these constraints in their models. This is not a trivial task. It may require management to think through restrictions on the transferability of capital in different geographies, in extreme scenarios, and also the priority for distributing surplus when more than one entity is in deficit at the same time.
Projection techniques
Much of the historic development of EC modeling has focused on calculating the opening position, with little focus on projected results. However, to effectively use EC in businesses it needs to be incorporated into the business planning and pricing process – both of which require a robust projection methodology. In Europe, projected results are also a key part of the Pillar II ORSA requirements, compelling companies to demonstrate an understanding of the development and emergence of risk over the business horizon.
The issue that respondents recognize is that they generally do not have sophisticated methodologies in place. 56 percent of respondents that currently project EC adopt factor-based approaches, with the complexity of the factors varying significantly between companies.
Furthermore, there is a lack of understanding about how robust projection methodologies are in stress scenarios or how wide the funnel of doubt becomes as you increase the projection period.
Operational risk modeling
The difficulty is there are many areas of subjectivity in the operational risk framework that complicate the modeling. Furthermore, potential losses as a result may not be known for many years, if at all. Compounding this is the fact that most companies have not recorded historic operational losses, meaning there is a lack of data available to calibrate the model. To address these issues, companies are establishing operational loss databases and looking to supplement their own limited data sets with external operation loss data provided by consortiums such as Operational Risk Consortium (ORIC).