PREDICTIVE BLEND: FUNDAMENTAL INDEXING WITH MARKOWITZ MEAN VARIANCE PORTOFOLIO IN INDONESIAN STOCK MARKET

• Main Author: Christian, David
• Format: Theses
• Language: Indonesia
• Subjects: Manajemen umum
• Online Access: https://digilib.itb.ac.id/gdl/view/60929
• Institution: Institut Teknologi Bandung

Portfolio selection has been extensively studied in field of business and economics. Many methods have been developed to construct a well-diversified portfolio that is expected to result in higher investment return with minimum risk. One of the most foundational works contributing to modern portfolio selection is the Markowitz mean variance optimization approach. The Markowitz approach heavily relies on past stock price performance, both in term of correlation structure and the return, to predict the future outcome. More recent studies, such as done by Hong and Wu (2016), shows that the stock past return is not always a good source of information about its future performance. In period with more volatility, it is better to rely more on the stock fundamentals rather than the past price movement, thus investors are tasked to look for another approach in constructing a well-diversified portfolio that rely more on the fundamentals, in times of volatile market. Another approach to construct a well-diversified portfolio is fundamental indexing by Arnott and Hsu (2005). It relies on stocks fundamental performance, choose the stocks, and based the weight of the stocks in the portfolio, upon several metrics of the fundamental performance. In times of volatility, this approach is expected to perform better than approaches that are reliant on past stock performance such as the Markowitz’s. Previous works by Pysarenko et al (2019) tried to combine Markowitz mean-variance optimization with the fundamental indexing with a good result. The Predictive Blend, as researched by Pysarenko et al (2019) using S&P500 data in the U.S. stock market, shows the newly combined portfolio performs better, not only in term of return, but also Sharpe ratio. This research is aimed to see if the approach by Pysarenko et al will also perform with similar result in Indonesian stock market. The author of this research constructed both Markowitz portfolio and the Fundamental Indexing portfolio independently, then using Buffet ratio to weight, combined both portfolio into a newly blended portfolio, test out-of-sample the new portfolio in term of return and then compare it to the Indonesian LQ45 benchmark index. The result shows that the new combined portfolio returns annually on average 43.89% higher than the benchmark index. This research benchmarked the new portfolio performance by investment return alone. It is expected that this research will be a source of future work concerning the blending of fundamental indexing and Markowitz approach in Indonesian stock market that also benchmark by Sharpe ratio.