Optimasi Alokasi Dana Pada Reksa Dana Indonesia Menggunakan Integer Linear Programming Dengan Batasan Kardinalitas

Simbolon, Felix Benaya (2026) Optimasi Alokasi Dana Pada Reksa Dana Indonesia Menggunakan Integer Linear Programming Dengan Batasan Kardinalitas. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221008-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (1MB)

Abstract

Pertumbuhan investasi reksa dana di Indonesia yang mencapai nilai dana kelolaan sebesar Rp 632 triliun dengan 12,8 juta investor pada akhir tahun 2024 mendorong kebutuhan akan pendekatan kuantitatif yang mampu menghasilkan portofolio optimal bagi investor ritel. Investor umumnya membatasi jumlah reksa dana dalam portofolionya agar mudah dikelola, sehingga proses alokasi dana menjadi permasalahan optimasi yang kompleks dan memerlukan pemodelan matematis yang tepat. Penelitian ini bertujuan untuk memformulasikan dan mengimplementasikan model Integer Linear Programming (ILP) dengan batasan kardinalitas guna mengoptimalkan alokasi dana pada reksa dana campuran Indonesia. Data yang digunakan berupa Nilai Aktiva Bersih (NAB) bulanan dari 19 produk reksa dana campuran yang terdaftar di Otoritas Jasa Keuangan (OJK) pada periode Januari 2023 hingga Desember 2024. Risiko portofolio diukur menggunakan Mean Absolute Deviation (MAD) sebagai fungsi tujuan yang diminimalkan, dengan variabel biner zi ∈ {0, 1} yang mengontrol pemilihan aset secara diskret. Model diselesaikan menggunakan solver CBC melalui pustaka Python PuLP pada tiga skenario batasan kardinalitas, yaitu K = 3, K = 5, dan K = 7. Hasil penelitian menunjukkan bahwa skenario K = 5 menghasilkan Sharpe Ratio tertinggi sebesar 16,56, dengan portofolio optimal terdiri atas TRIM Syariah Berimbang, Danapati Balance Fund, TRIM Kombinasi 2, Sucorinvest Anak Pintar, dan Trimegah Balanced Absolute Strategy Kelas A. Perbandingan dengan benchmark K = 19 membuktikan bahwa penambahan jumlah aset tidak selalu meningkatkan efisiensi portofolio. Penelitian ini menyimpulkan bahwa metode ILP dengan batasan kardinalitas efektif dalam menghasilkan portofolio reksa dana yang optimal dan realistis, dengan K = 5 sebagai titik kardinalitas optimal yang memberikan keseimbangan terbaik antara diversifikasi dan kinerja investasi.
==============================================================================================================================================
The significant growth of mutual fund investment in Indonesia, which reached total assets under management of IDR 632 trillion with 12.8 million investors by the end of 2024, has driven the need for a quantitative approach capable of producing optimal portfolios for retail investors. Investors typically limit the number of mutual funds in their portfolios to ensure manageability, making the fund allocation process a complex optimization problem that requires proper mathematical modelling. This study aims to formulate and implement an Integer Linear Programming (ILP) model with cardinality constraints to optimize fund allocation in Indonesian balanced mutual funds. The data used consists of monthly Net Asset Values (NAV) from 19 balanced mutual fund products registered with the Financial Services Authority (OJK) over the period January 2023 to December 2024. Portfolio risk is measured using Mean Absolute Deviation (MAD) as the minimized objective function, with binary variables zi ∈ {0, 1} controlling discrete asset selection. The model was solved using the CBC solver through the Python PuLP library under three cardinality constraint scenarios: K = 3, K = 5, and K = 7. The results indicate that the K = 5 scenario yields the highest Sharpe Ratio of 16.56, with the optimal portfolio comprising TRIM Syariah Berimbang, Danapati Balance Fund, TRIM Kombinasi 2, Sucorinvest Anak Pintar, and Trimegah Balanced Absolute Strategy Class A. Comparison with the K = 19 benchmark demonstrates that increasing the number of assets does not always improve portfolio efficiency. This study concludes that the ILP method with cardinality constraints is effective in generating optimal and realistic mutual fund portfolios for Indonesian retail investors, with K = 5 as the optimal cardinality point providing the best balance between diversification and investment performance.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Integer Linear Programming, Reksa Dana Campuran, Batasan Kardinalitas, Mean Absolute Deviation, Optimasi Portofolio, Integer Linear Programming, Balanced Mutual Fund, Cardinality Constraint, Mean Absolute Deviation, Portfolio Optimization
Subjects:	Q Science > QA Mathematics > QA278.5 Principal components analysis. Factor analysis. Correspondence analysis (Statistics)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Felix Benaya Simbolon
Date Deposited:	03 Aug 2026 09:21
Last Modified:	03 Aug 2026 09:21
URI:	http://repository.its.ac.id/id/eprint/143561

Actions (login required)

View Item