Article ID: 61d4d678001c5f84ba4a29d75101a11c47e6202b977a8ce5116a70f5b5abae95
Source ID: primary:theinsurer.com
Published At: -
Extraction Method: trafilatura
Body Text
The Insurer news staff had no role in the production of this content. This content is created by the brand marketing unit of The Insurer. Impact Forecasting has strengthened its support behind the open-source catastrophe modelling platform Oasis Loss Modelling Framework (LMF) as part of a drive towards increased transparency in catastrophe modelling. Sustainable Insurer met with the Impact Forecasting team to discuss the importance of transparency and what is driving demand for alternative modelling approaches. The latest step for Aon’s catastrophe model development centre comes as insurers face increasing pressure to ensure a robust risk view in the face of climate-driven volatility and regulatory pressure. Against this backdrop, an open modelling approach adopted by Impact Forecasting provides users with the necessary tools to develop tailored models. Chris Ewing, head of business development at Impact Forecasting, explained the importance of users “being able to see behind the models”. “We think it’s important that our models remain accessible so that individual clients can view the individual components, and can not only understand how frequent and severe the events themselves can be, but can also quantify the impact of those events on different types of structures,” he said. “The Oasis framework is helpful in achieving this because it includes the Open Data Standards (ODS). “Clients can not only see the hazard and vulnerability components of models and understand the algorithms behind them but also customise them – for example by incorporating their own claims data to better reflect their individual loss experience.” Ewing added that demand for greater insight into the models has stemmed from users becoming more technically minded in the past decade, and consequently “the ability to communicate” with the models for both technical and non-technical users is paramount. “Some modelling analysts might want to click a button and receive an answer but for those who want more detail, that’s where we’ve really opened it up to provide access to our model components and developments,” Ewing said. The wider notion of improved transparency is reflected in how Impact Forecasting’s offerings are designed to fit into users’ existing workflows. Muaz Nawaz, business development representative at Impact Forecasting, explained how Impact Forecasting models incorporate Open Exposure Data (OED) from Oasis and this, along with the ability to digest other forms of exposure using tools such as its internal data conversion offering, improves flexibility for users. “Let’s say insurers or reinsurers are primarily using a certain vendor but want to run their data through one of the open models,” Nawaz said. “Using our tools, they can convert their data and run multiple views of risk side by side, which is especially valuable for uncertain perils such as hail.” “By leveraging the Oasis LMF, these views can be run together within Impact Forecasting’s Elements platform, bringing Impact Forecasting and third‑party vendor models into a single, integrated workflow.” In addition, the automated approach means the insurer or reinsurer’s workforce is “freed-up from manual button clicking, and can actually spend their time interrogating the results and understanding the science”, Nawaz added. Another part of the transparency drive stems from Aon’s wider risk management objectives. Ewing said enabling insurance clients to view the components of the model aids in key areas of risk management, such as regulatory risk. “For instance, it can help when a company is using a Solvency II standard formula as part of their process for calculating capital requirements, and they then want to shift to using a partial internal model as they could have capital benefits – maybe with the Impact Forecasting model,” he said. “I think we’re well-placed in terms of sharing our data and insights from the model to help with that journey.”
Metadata (JSON)
{
"score": 9.885
}