Abstract

The economic crisis that hit Indonesia in the mid-1997 has changed the fundamental market factors, including to a spike in the IDR exchange rate, a bearish trend of IDX composite movement, a very high inflation rate, and a sharply falling GDP growth. Various theories show that there is any correlation between fundamental market factors and the level of return and risk in investment at the stock exchange. However it still needs to be proven statistically, especially in Indonesia.

On the other hand, the investment risk can be reduced by a diversification and the stock investment risk can be reduced by a portfolio. However the diversification would still leave a risk, called a systematic risk. Reducing risk by adding the stock collection to the portfolio will create a specific risk reduction pattern. If we can compare the risk reduction pattern between period before the crisis and during the crisis, then it would define whether the fundamental market factors will or will not change the risk profile of stock investment in Indonesia.

his research was conducted by observing the risk reduction of LQ-45 portfolio. It was also observed the portfolio risk reduction of all stocks listed on the Jakarta Stock Exchange (BEJ), with a sample of 150 stocks which had the greatest risk during the observation period.

Using stock index data as an observation model will only result a non stationary data distribution pattern and makes it difficult to be used for the observation. The author therefore used index return data which provides a normal distribution data pattern. The regression model of each LQ-45 stock data uses a multifactor Arbitrage Pricing Theory (APT) model with IDX composite returns and IDR/USD exchange rate returns as the independent variables. These two variables were tested for their correlation significance level with each LQ-45 stock individual IDX return. The author also examined auto regression and model using the correlogram facility (in EViews) to clean the data from serial autocorrelation.

Each LQ-45 stock that already has a regression model sorted from the largest to smallest index. This would become a sequence priority while adding each stock onto the portfolio. The standard deviation for first stock would be the initial risk of portfolio. This stock is then combined with the second stock using the same investment proportion (1/2 each) to form a single portfolio, and the portfolio risk is recorded. Then collected the 3rd stock onto portofolio with the same proportion (1/3 each), and so on until all 45 LQ-45 stocks form a single portfolio using an investment proportion of 1/45 each. All risks are always recorded at every time portofolio created.

These risks are indexed with a value of 100%, which is the initial risk of the portfolio (= risk of the first stock). This risk index is then plotted into a graph of the impact of additional stock collection on the LQ-45 portfolio risk. All listed individual IDX return is also observed statistically for the comparison. The data is also sorted based on its standard deviation and be added sequentially into the portfolio. The portfolio is collected from 150 stocks with the greatest risk, it is a same process like the LQ-45 portfolio observation.

The interesting result of the research is, eventhough there is a residual risk increased during the crisis (the systematic risk from the diversification process), but the minimum stock collections needed to result an effective risk portfolio reduction before the crisis is relatively same as during the crisis period.

These research findings have been presented and evaluated for a master's degree qualification by the academic professionals at the University of Indonesia