Prediction Model of Reserve on Pension Fund for Civil Servants of University Academic Employees in Indonesia Novriana Sumarti, Sendari Novia Sriyanto Putri, Annisa Rahmienda Maha, Hasna Khalishfi Yasyfa
Industrial and Financial Mathematics, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung
Abstract
The pension fund program is a type of long-term planning that aims to provide the living cost of civil servants after retirement. In Indonesia, the pension fund is calculated based on the latest salary of civil servants. As a consequence, there is an imbalance between the accumulated contributions being collected during the working period, and the pension benefits being paid for the rest of life. It results in an increase in the government^s budget burden to meet the sufficient pension fund. In this research, we propose a model to determine the civil servant pension fund that is based on the average salary of the last 10 years using the Accrued Benefit Cost method. We implement the model to determine the reserve of the pension fund of academic employees of Indonesian public universities. The salary increase of the academic staff of public universities is quite unique because it is heavily dependent on individual achievement during the working period. They could retire at age of 65 or 70, which is dependent on their rank. The first step in this study is to build a model to predict the final ranks of employees, which are commonly quite diverse. The regression model will then be used to forecast the percentage of salary increases. The Accrued Benefit Cost method is used to calculate pension benefits, which include salary increases between groups, periodic salary increases, and salary increases based on government regulations. The final step is to forecast the pension fund amount that the government must provide for the public university over the next 20 years, which is taking into account predictions for the final rank of civil servants, salary increase percentages, and potential new civil servants. The data set of new civil servants is created using the Poisson distribution, where the parameter values are obtained by analyzing sample data of civil servants from the past 30 years. In this study, the result of the model shows a significantly lower value than the results of the existing method based on the latest salary. This concludes that the method can reduce the fund amount that must be prepared by the government so that the burden on the state budget could be decreased.
Keywords: Reserve, pension fund, time series, prediction model
