PREDIKSI REALISASI ANGGARAN DENGAN SARIMA

Abstract

Economic development governance in Indonesia runs inclusive economic development. The National Development Planning Agency (Bappenas) defines one of the three main pillars of inclusive economic development in Indonesia as high economic growth. In promoting high economic growth through efficient and effective spending, quality budget management is necessary. One of the challenges in budget management is uncertainty in estimating budget spending or determining the fund withdrawal plan (RPD). From the latest regulation, the Peraturan Direktur Jenderal Perbendaharaan Kementerian Keuangan Nomor Per-5/PB/2024 tentang Petunjuk Teknis Penilaian Indikator Kinerja Pelaksanaan Anggaran (IKPA) Belanja Kementerian Negara/Lembaga, work units are encouraged to maintain the quality of budget planning for Ministries/Agencies based on the average monthly performance value deviation of halaman III DIPA. From this condition, where work units do not yet have a reference in estimating budget spending and it affects the performance of budget spending and IKPA, we propose a model to estimate budget spending using the Seasonal Autoregressive Integrated Moving Average (SARIMA). We developed SARIMA method using budget spending data from the Ministry of Communication and IT (Kemkominfo) obtained from the API OMSPAN since 2016. We used Python as programming tool. SARIMA results are a comparison between predictions and actual spendings of several work units of Kemkominfo in 2024. These budget spending prediction results are expected to give benefit for work unit head and institutions by providing an estimate of budget spending until the end of the year, thus enabling the formulation of strategic budget management policies. For the Ministry of Finance (Kemenkeu), this modeling can help estimate budget spending to spend efficient and effectively. This SARIMA modeling can be further developed in next research using data from SAKTI to generate RPD data in more detail.