OPTIMALISASI PORTOFOLIO MENGGUNAKAN ANALISIS KLASTER AVERAGE LINKAGE; PORTFOLIO OPTIMIZATION USING CLUSTER ANALYSIS AVERAGE LINKAGE

Rokhman, Awani Dwita, Abdurakhman

2016 | Skripsi | FMIPA

Nowadays, investment portfolio is one of the most favorite business by investors. In order to reach satisfied return on investment, so it needs to optimize the portfolio, which has maximum return and minimum risk. One of to get the optimum portfolio is the mean variance method. However, the using of this methods are still in doubt because it tends to emphasize the error estimation problem without considering the effect of the expected return. Solving the problem, there is research that single linkage clustering method is able to solve the problem but weighted result still obtain negative weight of some holding (short selling). For more real condition, weighted result should be positive (no short selling), this study will have a case studies about mean variance portfolio analysis with average linkage clustering method. The using data is the holding closing price every month from 10 stock LQ-45. From these data will be calculated monthly stock returns and continued with normality test. Afterthat, the mean variance portfolio is weighted using average linkage cluster analysis and single linkage (as a comparison). To find out optimum weighting result, will calculated by sharpe ratio. Case study showed that mean variance portfolio weighting by average linkage clustering get all positive stock value, while the single linkage clustering get some negative stock. Average linkage cluster analysis has a value higher Sharpe ratio (2.33), compared with single linkage cluster analysis (1.74). From this study, we get informed that mean variance portfolio using average linkage clustering is more optimum and more stable than single linkage clustering.

Kata Kunci : optimization portfolio, short selling, cluster analysis, average linkage, sharpe ratio