Matching Items (5)
Filtering by
- All Subjects: Actuarial
- Creators: Zicarelli, John
- Creators: Pan, Rong
Description
The use of generalized linear models in loss reserving is not new; many statistical models have been developed to fit the loss data gathered by various insurance companies. The most popular models belong to what Glen Barnett and Ben Zehnwirth in "Best Estimates for Reserves" call the "extended link ratio family (ELRF)," as they are developed from the chain ladder algorithm used by actuaries to estimate unpaid claims. Although these models are intuitive and easy to implement, they are nevertheless flawed because many of the assumptions behind the models do not hold true when fitted with real-world data. Even more problematically, the ELRF cannot account for environmental changes like inflation which are often observed in the status quo. Barnett and Zehnwirth conclude that a new set of models that contain parameters for not only accident year and development period trends but also payment year trends would be a more accurate predictor of loss development. This research applies the paper's ideas to data gathered by Company XYZ. The data was fitted with an adapted version of Barnett and Zehnwirth's new model in R, and a trend selection algorithm was developed to accompany the regression code. The final forecasts were compared to Company XYZ's booked reserves to evaluate the predictive power of the model.
ContributorsZhang, Zhihan Jennifer (Author) / Milovanovic, Jelena (Thesis director) / Tomita, Melissa (Committee member) / Zicarelli, John (Committee member) / W.P. Carey School of Business (Contributor) / School of Mathematical and Statistical Sciences (Contributor) / Barrett, The Honors College (Contributor)
Created2018-05
Description
Catastrophe events occur rather infrequently, but upon their occurrence, can lead to colossal losses for insurance companies. Due to their size and volatility, catastrophe losses are often treated separately from other insurance losses. In fact, many property and casualty insurance companies feature a department or team which focuses solely on modeling catastrophes. Setting reserves for catastrophe losses is difficult due to their unpredictable and often long-tailed nature. Determining loss development factors (LDFs) to estimate the ultimate loss amounts for catastrophe events is one method for setting reserves. In an attempt to aid Company XYZ set more accurate reserves, the research conducted focuses on estimating LDFs for catastrophes which have already occurred and have been settled. Furthermore, the research describes the process used to build a linear model in R to estimate LDFs for Company XYZ's closed catastrophe claims from 2001 \u2014 2016. This linear model was used to predict a catastrophe's LDFs based on the age in weeks of the catastrophe during the first year. Back testing was also performed, as was the comparison between the estimated ultimate losses and actual losses. Future research consideration was proposed.
ContributorsSwoverland, Robert Bo (Author) / Milovanovic, Jelena (Thesis director) / Zicarelli, John (Committee member) / School of Mathematical and Statistical Sciences (Contributor) / Barrett, The Honors College (Contributor)
Created2018-05
Description
Yield is a key process performance characteristic in the capital-intensive semiconductor fabrication process. In an industry where machines cost millions of dollars and cycle times are a number of months, predicting and optimizing yield are critical to process improvement, customer satisfaction, and financial success. Semiconductor yield modeling is essential to identifying processing issues, improving quality, and meeting customer demand in the industry. However, the complicated fabrication process, the massive amount of data collected, and the number of models available make yield modeling a complex and challenging task. This work presents modeling strategies to forecast yield using generalized linear models (GLMs) based on defect metrology data. The research is divided into three main parts. First, the data integration and aggregation necessary for model building are described, and GLMs are constructed for yield forecasting. This technique yields results at both the die and the wafer levels, outperforms existing models found in the literature based on prediction errors, and identifies significant factors that can drive process improvement. This method also allows the nested structure of the process to be considered in the model, improving predictive capabilities and violating fewer assumptions. To account for the random sampling typically used in fabrication, the work is extended by using generalized linear mixed models (GLMMs) and a larger dataset to show the differences between batch-specific and population-averaged models in this application and how they compare to GLMs. These results show some additional improvements in forecasting abilities under certain conditions and show the differences between the significant effects identified in the GLM and GLMM models. The effects of link functions and sample size are also examined at the die and wafer levels. The third part of this research describes a methodology for integrating classification and regression trees (CART) with GLMs. This technique uses the terminal nodes identified in the classification tree to add predictors to a GLM. This method enables the model to consider important interaction terms in a simpler way than with the GLM alone, and provides valuable insight into the fabrication process through the combination of the tree structure and the statistical analysis of the GLM.
ContributorsKrueger, Dana Cheree (Author) / Montgomery, Douglas C. (Thesis advisor) / Fowler, John (Committee member) / Pan, Rong (Committee member) / Pfund, Michele (Committee member) / Arizona State University (Publisher)
Created2011
Description
A factor accounting for the COVID-19 pandemic was added to a generalized linear model to more accurately predict unpaid claims. COVID-19 has affected not just healthcare, but all sectors of the economy. Because of this, whether or not an automobile insurance claim is filed during the pandemic needs to be taken into account while estimating unpaid claims. Reserve-estimating functions such as glmReserve from the “ChainLadder” package in the statistical software R were experimented with to produce their own results. Because of their insufficiency, a manual approach to building the model turned out to be the most proficient method. Utilizing the GLM function, a model was built that emulated linear regression with a factor for COVID-19. The effects of such a model are analyzed based on effectiveness and interpretablility. A model such as this would prove useful for future calculations, especially as society is now returning to a “normal” state.
ContributorsKossler, Patrick (Author) / Zicarelli, John (Thesis director) / Milovanovic, Jelena (Committee member) / Barrett, The Honors College (Contributor) / School of Mathematical and Statistical Sciences (Contributor)
Created2022-05
Description
The objective of this study is to build a model using R and RStudio that automates ratemaking procedures for Company XYZ’s actuaries in their commercial general liability pricing department. The purpose and importance of this objective is to allow actuaries to work more efficiently and effectively by using this model that outputs the results they otherwise would have had to code and calculate on their own. Instead of spending time working towards these results, the actuaries can analyze the findings, strategize accordingly, and communicate with business partners. The model was built from R code that was later transformed to Shiny, a package within RStudio that allows for the build-up of interactive web applications. The final result is a Shiny app that first takes in multiple datasets from Company XYZ’s data warehouse and displays different views of the data in order for actuaries to make selections on development and trend methods. The app outputs the re-created ratemaking exhibits showing the resulting developed and trended loss and premium as well as the experience-based indicated rate level change based on prior selections. The ratemaking process and Shiny app functionality will be detailed in this report.
ContributorsGilkey, Gina (Author) / Zicarelli, John (Thesis director) / Milovanovic, Jelena (Committee member) / Barrett, The Honors College (Contributor) / School of Mathematical and Statistical Sciences (Contributor)
Created2022-05