Triatmojo, Karuniawan Bangun (2026) Penentuan Premi Asuransi Kelapa Sawit Menggunakan Generalized Linear Models Dengan Insentif Karbon Sebagai Pengurang Premi. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5002221057-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5002221057-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy |
Abstract
Produksi kelapa sawit memiliki kerentanan tinggi terhadap variabilitas curah hujan, sehingga asuransi diperlukan sebagai instrumen mitigasi risiko produksi di sektor perkebunan. Keterbatasan data klaim aktual menjadi kendala dalam penentuan premi berbasis risiko, sehingga penelitian ini menggunakan pendekatan production shortfall untuk mengonstruksi nilai kerugian produksi sebagai proksi klaim. Penelitian ini bertujuan menentukan premi asuransi kelapa sawit menggunakan pendekatan Generalized Linear Models (GLMs) dengan insentif karbon diperlakukan sebagai pengurang premi komersial pada tahap akhir penentuan premi. Menggunakan data bulanan kelapa sawit periode 2019–2024 sebanyak 72 observasi dengan variabel utama curah hujan dan produksi kelapa sawit. Kerugian produksi didefinisikan sebagai selisih antara threshold produksi normal (kuantil 25% produksi historis sebesar 360.161,2 ton) dan produksi aktual, dengan 18 dari 72 observasi teridentifikasi mengalami kerugian dan rata-rata severity historis sebesar 126.039,6 ton. Metode frequency-severity digunakan, dengan model frequency berupa GLMs Binomial berfungsi hubung logit untuk peluang terjadinya loss, dan model severity berupa GLMs Gamma berfungsi hubung log untuk besarnya kerugian saat loss terjadi. Hasil penelitian menunjukkan bahwa curah hujan berpengaruh signifikan terhadap peluang terjadinya loss, namun tidak signifikan terhadap besarnya severity, dengan expected loss berbasis GLMs sebesar 31.509,9 ton. Berdasarkan harga TBS Rp3.880/kg, diperoleh premi murni Rp122.258.424.707 dan premi komersial Rp146.710.109.649 (rentang bootstrap Rp111.861.390.981–Rp182.949.617.177). Lima skenario insentif karbon (2%, 4%, 6%, 8%, 10%) diterapkan, dipilih skenario dasar 6% (K3) setara Rp8.802.606.579 atau 149.704,20 tCO₂e sehingga premi turun menjadi Rp137.907.503.070. Lima skenario deductible-limit diterapkan untuk memperoleh struktur premi yang lebih proporsional, dengan skenario terbaik D3×K3 (deductible 50.000 ton, limit 200.000 ton, insentif karbon 6%) menghasilkan premi akhir Rp34.669.264.416 (interval kepercayaan 95%: Rp19.601.552.559–Rp51.063.309.410). Hasil penelitian menunjukkan GLMs dapat digunakan sebagai dasar penentuan premi asuransi kelapa sawit berbasis risiko produksi, sedangkan insentif karbon berperan sebagai pengurang premi tanpa mengubah struktur risiko dasar.
Kata kunci: Asuransi Kelapa Sawit, Premi, GLMs, Production Shortfall, Insentif Karbon.
====================================================================================================================================
Oil palm production is highly vulnerable to rainfall variability, making insurance necessary as an instrument for mitigating production risk in the plantation sector. The unavailability of actual claim data poses a challenge in determining risk-based premiums, so this study uses the production shortfall approach to construct production loss values as a claim proxy. This study aims to determine oil palm insurance premiums using the Generalized Linear Models (GLMs) approach, with carbon incentives treated as a deduction from the commercial premium in the final stage of premium determination. Monthly oil palm data from 2019 to 2024, comprising 72 observations, were used, with rainfall and oil palm production as the main variables. Production loss is defined as the difference between the normal production threshold (the 25th percentile of historical production, amounting to 360,161.2 tons) and actual production, resulting in 18 out of 72 observations being identified as loss events, with an average historical severity of 126,039.6 tons. The frequency-severity approach was applied, with the frequency model developed using a Binomial GLMs with a logit link function to estimate the probability of loss occurrence and the severity model developed using a Gamma GLMs with a log link function to estimate the magnitude of loss when it occurs. The results show that rainfall is significant in explaining the probability of loss occurrence but not significant in explaining the magnitude of severity, with a GLMs-based expected loss of 31,509.9 tons. Based on a fresh fruit bunch (FFB) price of Rp3,880/kg, the pure premium is Rp122,258,424,707 and the commercial premium is Rp146,710,109,649 (bootstrap range: Rp111,861,390,981–Rp182,949,617,177). Five carbon incentive scenarios (2%, 4%, 6%, 8%, and 10%) were evaluated, with the 6% scenario (K3) selected as the baseline, equivalent to Rp8,802,606,579 or 149,704.20 tCO₂e, reducing the premium to Rp137,907,503,070. Five deductible-limit scenarios were also applied to obtain a more proportional premium structure. The best-performing scenario, D3×K3 (a deductible of 50,000 tons, a limit of 200,000 tons, and a 6% carbon incentive), resulted in a final premium of Rp34,669,264,416 (95% confidence interval: Rp19,601,552,559–Rp51,063,309,410). The findings indicate that GLMs can be used as a basis for determining oil palm insurance premiums based on production risk, while carbon incentives can serve as a premium deduction without altering the underlying risk structure.
Keywords: Oil Palm Insurance, Premium, GLMs, Production Shortfall, Carbon Incentive.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Kata kunci: Asuransi Kelapa Sawit, Premi, GLMs, Production Shortfall, Insentif Karbon. Keywords: Oil Palm Insurance, Premium, GLMs, Production Shortfall, Carbon Incentive |
| Subjects: | Q Science Q Science > QA Mathematics Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Karuniawan Bangun Triatmojo |
| Date Deposited: | 30 Jul 2026 08:35 |
| Last Modified: | 30 Jul 2026 08:35 |
| URI: | http://repository.its.ac.id/id/eprint/139590 |
Actions (login required)
 |
View Item |