Application of High-Performance Computing for Calculating Reserves using the Cape Cod Method
S. R. Pranav Sai1, Ajay Singh Pawar2, Satya Sai Mudigonda3, Pallav Kumar Baruah4
1S.R.Pranav Sai, Doctorial Research Scholar, Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, Puttaparthi, India.
2Ajay Singh Pawar, Master’s Student, Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, Puttaparthi, India.
3Satya Sai Mudigonda, Senior Tech Actuarial Consultant, Hon. Professor in Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning Puttaparthi, India.
4Dr.Pallav Kumar Baruah, Associate Professor, Department of Mathematics and Computer Science, Sri Sathya Sai Institute of Higher Learning, Puttaparthi, India.
Manuscript received on September 12, 2019. | Revised Manuscript received on October 20, 2019. | Manuscript published on October 30, 2019. | PP: 54-57 | Volume-9 Issue-1, October 2019 | Retrieval Number: A1032109119/2019©BEIESP | DOI: 10.35940/ijeat.A1032.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Calculation of reserves is one significant step in the strategic perspective for an insurance company. This is done on a periodic manner in order to have a better understanding regarding the future liabilities of the company. The time required for these calculations will grow quadratically as the size of the input increases. These computations are done several times by taking different factors in view. Thus, these computations become costly with regard to time. One of the many actuarial techniques for calculating reserves namely the Cape Cod method was used to calculate the reserves. HPC was applied for the same in order to reduce the time required to calculate a company’s reserves. We applied HPC to calculate Projected Ultimate Claims based on Cape Cod method. Method takes year wise input of reported claims, incurred claims and earned premiums and outputs Projected Ultimate claims for each year. Making use of the independence which exists in computation of output across years, we parallelized the computations across years. The parallelization strategy proposed gives the speed up to 66194X compared to the serial implementation of the same.
Keywords: Reserves, Insurance, Development triangle, Cape Cod method, Actuarial analysis, Development year, Accident year, Claims incurred, Claims pair, GPU, CUDA.