Optimasi Portofolio Menggunakan Metode Hybrid Branch and Bound and Outer Approximation

Aini, Ela Nur (2024) Optimasi Portofolio Menggunakan Metode Hybrid Branch and Bound and Outer Approximation. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002201116 - Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (3MB)

Abstract

Perkembangan investasi yang signifikan menjadikan investasi sebagai tren masyarakat untuk mendapatkan passive income. Salah satu instrumen investasi pada aset keuangan yang paling populer dan cukup dikenal masyarakat adalah saham. Saham merupakan instrumen investasi yang mempunyai tingkat pengembalian (return) yang tinggi, akan tetapi memiliki risiko yang tinggi juga. Oleh karena itu, perlu dibentuk portofolio sebagai langkah untuk mengurangi risiko. Optimasi portofolio adalah suatu teknik untuk menentukan portofolio terbaik dengan mengumpulkan portofolio yang tersedia berdasarkan toleransi risiko yang dapat diterima dan tingkat pengembalian (return) yang diharapkan oleh investor. Dalam penelitian ini, pembentukan portofolio menggunakan model markowitz dengan penambahan kendala yang membatasi risiko non sistematis yang berupa kendala kuadratik, kendala kardinal yang membatasi banyaknya aset yang terpilih, dan kendala semi-continuous yang menentukan keikutsertaan suatu aset sehingga membentuk model Mixed Integer Quadratiqally Constrained Quadratic Program (MIQCQP). Model MIQCQP tersebut akan diselesaikan menggunakan metode Hybrid Branch and Bound and Outer Approximation. Metode Hybrid tersebut merupakan perbaikan dari metode Branch and Bound Nonlinear klasik dengan menerapkan algoritma Outer Approximation pada beberapa submasalah Non Linear Program (NLP) pada pohon pencacahan dan masalah utama pada frekuensi yang telah ditentukan. Kemudian akan dilakukan perbandingan metode Hybrid Branch and Bound and Outer Approximation dengan metode Branch and Bound Nonlinear klasik. Dari hasil simulasi menunjukkan bahwa banyaknya aset yang terpilih dan return dari kedua metode sama besar tetapi dengan proporsi aset yang berbeda. Nilai risiko yang diperoleh pada metode Hybrid Branch and Bound and Outer Approximation lebih kecil daripada metode Branch and Bound, artinya optimasi portofolio menggunakan metode Hybrid Branch and Bound and Outer Approximation lebih optimal daripada metode Branch and Bound. Namun, metode Hybrid Branch and Bound and Outer Approximation memiliki kekurangan dari sisi waktu.

========================================================================================================================

The significant development of investment has invested a trend for people to get passive income. One of the most popular and well-known investment instruments in financial assets is stocks. Stocks are an investment instrument with a high return, rate but also a high risk. Therefore, it is necessary to form a portfolio to reduce risk. Portofolio optimization is a technique to determine the best portfolio by collecting available portfolios based on acceptable risk tolerance and the expected rate of return by investors. In this study, portfolio formation uses the Markowitz model with the addition of constraints that limit non-systematic risks in the form of quadratic constraints, cardinal constraints that limit the number of selected assets, and semi-continuous constraints that determine the participation of an asset so as to form the Mixed Integer Quadratiqally Constrained Quadratic Program (MIQCQP) model. The MIQCQP model will be completed using the Hybrid Branch and Bound and Outer Approximation method. The Hybrid method is an improvement over the classic Branch and Bound Nonlinear method by applying the Outer Approximation algorithm to several Non-Linear Program (NLP) subproblems in the enumeration tree and to the main problem at the predetermined frequency. Then a comparison of the Hybrid Branch and Bound and Outer Approximation method with the classic Branch and Bound Nonlinear method will be carried out. The simulation results show that the number of selected assets and returns from both methods are the same but with different proportions of assets. The risk value obtained in the Hybrid Branch and Bound and Outer Approximation method is smaller than the Branch and Bound method, meaning that portfolio optimization using the Hybrid Branch and Bound and Outer Approximation method is more optimal than the Branch and Bound method. However, the Hybrid Branch and Bound and Outer Approximation method has a shortcoming in terms of time.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Branch and Bound, Hybrid, Mixed Integer Quadratically Constraint Quadratic Program, Optimasi Portofolio, Outer Approximation, Branch and Bound, Hybrid, Mixed Integer Quadratically Constraint Quadratic Program, Outer Approximation, Portofolio Optimization.
Subjects:	H Social Sciences > HG Finance > HG4529.5 Portfolio management H Social Sciences > HG Finance > HG4910 Investments Q Science > QA Mathematics > QA401 Mathematical models.
Divisions:	Faculty of Mathematics, Computation, and Data Science > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Ela Nur Aini
Date Deposited:	06 Aug 2024 03:31
Last Modified:	06 Aug 2024 03:31
URI:	http://repository.its.ac.id/id/eprint/113721

Actions (login required)

View Item