Implementasi Model Black-Litterman Berbasis Peramalan Return menggunakan Long Short-Term Memory pada Optimasi Portofolio Saham Perbankan IDX30

Afiifah, Alyaa Nabiilah (2026) Implementasi Model Black-Litterman Berbasis Peramalan Return menggunakan Long Short-Term Memory pada Optimasi Portofolio Saham Perbankan IDX30. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5006221045-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Peningkatan literasi dan inklusi keuangan di Indonesia mendorong pertumbuhan jumlah investor di pasar modal sehingga kebutuhan strategi pengelolaan portofolio yang lebih adaptif semakin penting. Sektor perbankan, sebagai kontributor terbesar kapitalisasi IDX30, memiliki peran dominan dalam stabilitas pasar. Posisi strategis ini semakin diperkuat oleh adanya Undang-Undang Pengembangan dan Penguatan Sektor Keuangan (UU P2SK) 2023 menjadikan saham perbankan sebagai pilihan investasi yang strategis. Namun, perbedaan karakteristik antar perusahaan menghadirkan tantangan dalam pembentukan portofolio yang optimal. Penelitian ini membentuk portofolio saham perbankan IDX30 menggunakan model Black-Litterman dengan investor view yang dihasilkan dari peramalan Long Short-Term Memory (LSTM) untuk 20 hari ke depan, dengan tiga skenario parameter ketidakpastian pasar (τ=1/T, 0.01, 0.05). Data penelitian berupa harga penutupan harian BBCA, BBRI, BMRI, dan BBNI periode 1 Januari 2023–31 Juli 2025 agar mencerminkan kondisi pasar yang stabil pascapandemi. Hasil peramalan LSTM menunjukkan expected return sebesar 5.37% (BBCA), 7.21% (BBRI), –4.96% (BMRI), dan 2.13% (BBNI). Integrasi peramalan tersebut ke dalam model Black-Litterman menghasilkan portofolio terbaik pada skenario τ=0.01 dengan komposisi 57.04% BBCA dan 42.96% BBRI. Portofolio ini memiliki Sharpe Ratio –2.33%, yang menunjukkan karakter lebih konservatif dibandingkan dua skenario τ lainnya maupun indeks IDX30 sebagai tolok ukur. Analisis risiko melalui simulasi Monte Carlo menghasilkan estimasi Value at Risk (VaR) untuk tingkat kepercayaan 90%, 95%, dan 99% sebesar –1.63%, –2.18%, dan –3.34% serta Conditional Value at Risk (CVaR) sebesar –2.40%, –2.91%, dan –3.98%. Validitas estimasi risiko dikonfirmasi melalui uji Kupiec yang menunjukkan bahwa estimasi VaR berada dalam tingkat keandalan yang dapat diterima. Hasil penelitian membuktikan bahwa integrasi LSTM dan Black-Litterman mampu menghasilkan portofolio yang relevan dengan kondisi pasar dan dapat mendukung pengambilan keputusan investasi yang lebih terukur.
======================================================================================================================================
The rising levels of financial literacy and inclusion in Indonesia have increased participation in the capital market, strengthening the need for more adaptive portfolio management strategies. The banking sector, as the largest contributor to IDX30 market capitalization, plays a crucial role in supporting market stability. This position is further reinforced by the Financial Sector Development and Reinforcement Law (UU P2SK) 2023, which enhances the resilience and competitiveness of the financial industry, thereby positioning banking stocks as strategic investment instruments. Despite this, variations in the fundamental characteristics and risk exposures of individual banks create challenges in constructing an optimal portfolio. This study develops an IDX30 banking stock portfolio using the Black-Litterman model, with investor views generated through 20-day ahead return forecasts from a Long Short-Term Memory (LSTM) network. Three scenarios of the market uncertainty parameter (τ = 1/T, 0.01, 0.05) are examined. The dataset consists of daily closing prices of BBCA, BBRI, BMRI, and BBNI from 1 January 2023 to 31 July 2025, representing a stable post-pandemic market period. LSTM forecasting results indicate expected returns of 5.37% (BBCA), 7.21% (BBRI), –4.96% (BMRI), and 2.13% (BBNI). Integrating these forecasts into the Black-Litterman framework produces the best-performing portfolio under τ = 0.01, with allocations of 57.04% to BBCA and 42.96% to BBRI. This portfolio achieves a Sharpe Ratio of –2.33%, reflecting a more conservative profile compared with the other τ scenarios and the IDX30 benchmark. Risk assessment using Monte Carlo simulation yields Value at Risk (VaR) estimates of –1.63%, –2.18%, and –3.34% at confidence levels of 90%, 95%, and 99%, respectively, while Conditional Value at Risk (CVaR) values reach –2.40%, –2.91%, and –3.98%. The Kupiec test confirms the reliability of the VaR estimates. The findings demonstrate that integrating LSTM-based forecasts within the Black-Litterman model produces a portfolio aligned with current market conditions and supports more informed and data-driven investment decisions.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Black-Litterman, IDX30, LSTM, Optimasi Portofolio, Saham Perbankan, Banking Stocks, Black-Litterman, IDX30, LSTM, Portfolio Optimization
Subjects:	H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HG Finance > HG4529.5 Portfolio management H Social Sciences > HG Finance > HG4915 Stocks--Prices
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Alyaa Nabiilah Afiifah
Date Deposited:	14 Jan 2026 07:49
Last Modified:	14 Jan 2026 07:49
URI:	http://repository.its.ac.id/id/eprint/129620

Actions (login required)

View Item