PORTFOLIO PERFORMANCE ANALYSIS WITH JENSEN ' S METHOD ON CAPM AND APT MODELS

Dr. Andi Desfiandi 1 , Ita Fionita, MM 1 and Prof. Dr. Hapzi Ali CMA 2 . 1. Institute of Informatics & Business Darmajaya, Bandar Lampung Indonesia. 2. Postgraduate of MercuBuana University, Jakarta Indonesia. ...................................................................................................................... Manuscript Info Abstract ......................... ........................................................................ Manuscript History

The purpose of this research is to analyze the portfolio using CAPM and APT models to predict stock returns LQ 45 in BEI. The issues raised in this study were: 1) Which combination of companies that produce optimal portfolio using the CAPM model; 2) Which company combination that produces an optimal portfolio by using the APT model; 3) How is the comparison of optimal portfolio by using Jensen's Index on the model of CAPM and APT models. Beta in this study using simple linear regression analysis method to the CAPM model and multiple linear regression model for APT. The independent variables in this model is LQ 45, while the dependent variable is the market risk premium, exchange rate, SBI, inflation and GDP. To determine the optimal portfolio return is analyzed using a model CAPM and APT models, whereas for portfolio performance measure used Jensen's index. On the other hand to know the difference between the CAPM model with the APT model used t test. The sample used in this study as many as 31 companies are determined by the criteria 1). Companies listed in the LQ 45 in BEI. 2). The company is listed in the LQ 45 actively traded 2 consecutive years from January 2010 to December 2011. The results of this study indicate that there are significant differences between Jensen's Index on the model of CAPM and APT models. With Jensen's Index score APT model larger than the CAPM model.

…………………………………………………………………………………………………….... Introduction:-
In the stock market almost all investment contain elements of uncertainty or risk. Investors do not know with certainty the results to be obtained from its investments. In such circumstances it is said that these investors face risks in its investments. Given this risk, the investor will suffer or benefit that is not as expected, resulting in the emergence of irregularities, divergence is often called the uncertainty (uncertainty). In theory it is said that a portfolio of high risk will provide high returns as well (high risk, high return), so the greater the risk will be greater the return that would be acceptable, and vice versa. But not all investors agree with the statement, and the main desire of investors is to minimize risk and increase the cost (minimize risk and maximize return). So that they will look for a portfolio that has the lowest possible risk to the optimum return.

ISSN: 2320-5407
Int. J. Adv. Res. 5 (2), 1981-1991 1982 Model balance (analysis tools) will help the understanding of how to determine the relevant risk of an asset, as well as the relationship of risk and expected return on an asset when the market is in equilibrium. Two kinds of balance popular models that can be used in predicting exchange for shares (return) is expected is the Capital Asset Pricing Model (CAPM) and the Arbitrage Pricing Theory (APT). Both models are popular because of the ease of application as well as the assumptions underlying these two models.
Markowitz's portfolio theory put forward in 1952 (Arzac and Bawa, 1977), known as the Markowitz model, it provides a way how to invest efficiently and optimally. By forming the optimal portfolio. The purpose of forming the optimal portfolio is to fulfill the principles of investing. "Obtaining yield (return) at the desired level with minimum risk". To minimize the risk, there should be diversification in investing, is to establish a portfolio or to invest funds not only in one asset alone but to several asset. The problem is how large a proportion of the funds must be invested in each asset in order to obtain the desired level of yield with minimum risk. Bring Arzac research and use two parameters CAPM by behavioral approach safety-first on the return and risk, which found that safety-first approach implies a cautious attitude towards risk, portfolio choice and balance of the market comparable to that implied by the approach of expected utility. Black (1974) suggests that the estimator for the portfolio only the Rt -R. It is exactly the same as tests used for capital asset pricing model, using the world market portfolio to calculate this is consistent with the notion that taxes are what cause the world capital asset pricing model does not continue smoothly.
CAPM (Capital Asset Pricing Model) is a theory that tries to explain how an asset is determined by the market price, or how to determine the level of profit that is deemed worthy of an investment. Sharpe (1964) analyze how the links between return an asset (R i ) with the market portfolio return (R m ) is. By making a simple regression equation in which the dependent variable his return an asset and its independent variable return of the market portfolio will be determined how or the relationship between the return of the market portfolio with the return of certain assets. Cornell Research (1979), was able to show that the CAPM can be used to detect the investor excel in the world of asymmetric information, but they do not arouse demonstration CAPM as a practical tool for measuring performance.
APT (Arbitrage Pricing Theory) is an alternative model of CAPM to assess the financial asset. The financial model developed by Ross (1976) have emerged based on the idea that the financial markets are competitive, the arbitrage will make two assets that share similar characteristics, such as equally (riskless) or without taking the risk of giving expectations of return of the same. APT basically uses the idea also that two investment opportunities that have the same identical characteristics can not be sold at different prices (the law of one price). When assets are sold with the same characteristics different prices there will be a chance to do arbitrage, buying valuable assets cheap and sell at a higher price at the same time so as to obtain profit without risk. APT model emphasizes that the level of expected return (expected return) is heavily influenced by macro-economic factors and not by the unique risks. We can assume that there are factors in arbitrage pricing as specific portfolios tend to be affected by the influence of joint (common influence). If the expected risk premium of each portfolio will be proportionate to the market beta portfolio, the APT and CAPM will give the same result.. Based on the description above, where it is known that the CAPM and APT models have strengths and weaknesses of each to determine which rewards portfolio shares (return) the most optimal and profitable for investors. The authors are interested to analyze the portfolio by using both models to compare measurement models that can be known where the most optimal as an asset assessment tool that can assist investors in selecting a portfolio decision making. In the APT model of macro-economic factors used in this study is inflation, gross domestic product, interest rates and the exchange rate against the US dollar (exchange rate).
The author's intent in doing this Research are:-1. Determine the combination of companies which produce optimal portfolio using the CAPM model. 2. Knowing which combination of companies that produce optimal portfolio by using the APT model. The population used in this study is LQ 45 listed in the Indonesia Stock Exchange. While the sample is determined by using purposive sampling method, which is intended to achieve certain limits or goals expected from this research. For that set of samples with the following criteria: The company whose shares including the shares are actively traded on an ongoing basis and is LQ 45 in Indonesia Stock Exchange 2010-2011 period.
This study uses multiple methods of data collection are as follows:-

Results of Testing Requirements Analysis Data:-
Optimal portfolio Outline Security Market Line (SML) describe the overall market return to longs systematic risk (the risk can not be diversified away) or Beta. With these criteria results obtained from the company's stock 22 9 sector, then pursed again with a selection system that shares a big 3 ratings (sectors) based on market capitalization. That was selected 17 stocks of companies included in the criteria for the establishment of an optimal portfolio as shown in the figure below:  In this study to determine the combined portfolio writer used a tree diagram that is a way to describe the nature of a hierarchical structure in graphical form. Tree (tree diagram) a specific type of diagram that has a network topology (relationship) is unique (AdhiDarmawanSutjiadi, 2003). Each stock companies get the opportunity, and equal opportunities to represent the sector in the combined portfolio, resulting from the 17 stocks incorporated in 9 sectors industry produces 108 combined portfolio as shown in the following table.
1989 Table 2:-The combination of company shares Index jensen's (α) model CAPM dan model APT:-Calculating Jensen's portfolio index by using the following formula: α = Rp -E(R) After E (Rp) of each combination is obtained then the value of the portfolio Jensen'n index (α) of the model CAPM and APT model can be calculated by using the formula above the monthly period so that we will get the value of α portfolio of 108 combinations. Then α portfolio value each combination calculated the mean (average) of its α, so that there is only one (1) α within 1 (one) combined portfolio, and as a result there will be as many as 108 108 α on the combined portfolio. To determine the return combinations where the most optimal portfolio can be seen from an index value of Jensen's (α) of each of the combined portfolio, the greater the value of α, the greater the return portfolio. Jensen's Index (α) the top 10 of the 108 combined portfolio CAPM and APT models shown in the following table.    (α)  1  80  B, C, E, G, H, I, L, O, P  1.208032  2  25  A, C, D, G, H, I, L, O, P  1.207309  3  85  B, C, D, G, H, I, K, M, Q  1.190167  4 26 From Table 3 and Table 4 illustrates Alpha (α) is showing the difference between the actual investment returns with the expected investment returns or the benchmark for the level of market risk (beta) specific. On top ranking calculation results showed that the alpha value of 0.025107 CAPM model, which is a positive alpha value and meaning describe portfolio performance is good or feasible to do investment. While the alpha value of 1.208032 top APT model, demonstrating the value of alpha is positive and that means investment or portfolio performance in good standing and eligible to do investment.  The data indicate that Jensen's Index APT model was higher than Jensen's Index CAPM model with a number of differences in the average standard deviation of 0.53223 to 0.374123. Total average score of Jensen's Index APT model is 0.542808, while the total score of the average Index of Jensen's model of CAMP is 0.011585.

Conclussion:-
Based on the results of the discussion can be drawn conclusions as follows:-1. The combination of the company that produces the optimal portfolio by using the CAPM model is PP London Sumatra Tbk, LIPPOKarawaciTbk, Indo TambangrayaTbk, Holcim Indonesia Tbk, PT Astra International Tbk, United Tractors Tbk, Bank MandiriTbk, Kalbe FarmaTbk and JasaMargaTbk , with an Index score of Jensen's CAPM model by 0.025107. 2. The combination of the company that produces the optimal portfolio by using the APT model is PP London Sumatra Tbk, LippoKarawaciTbk, Indo TambangrayaTbk, Holcim Indonesia Tbk, PT Astra International Tbk, United Tractors Tbk, Bank of the Republic of Indonesia Tbk, Kalbe FarmaTbk and JasaMargaTbk, with an Index score of Jensen's APT model by 1.208032. 3. The results of this study showed no significant difference between Jensen's index CAPM model with Jensen's index APT model, where Jensen's index APT model was higher than Jensen's Index CAPM model with a number of differences in the average standard deviation of 0.53223 to 0.374123. Total average score of Jensen's index APT model is 0.542808, while the average score of Jensen's index CAPM model is 0.011585.