بررسي و مقايسه مدلهاي قيمت گذاري دارايي سرمايه اي با رويكردهاي متفاوت به ريسك در بورس اوراق بهادار تهران

A Study on Capital Asset Pricing Model with Different Approaches to Risk in Tehran Stock Exchange

هر سرمايه گذار با هدف كسب نرخ بازده مورد انتظار اقدام به سرمايه‌گذاري مي كند و هدف وي كسب بيشترين بازده از سرمايه گذاري خود مي باشد؛ اما با توجه به اين اصل اقتصاد مالي كه به طور معمول بازده با ريسك رابطه اي مثبت دارد، لذا كسب هر ميزان بازده اي بيشتر از نرخ بازده بدون ريسك، با مقداري ريسك همراه است كه معمولاً با افزايش بازده اين ريسك نيز افزايش مي يابد. در اين راستا دست يابي به بهترين رويه هاي اندازه گيري ريسك در هر بازاري مي تواند براي سرمايه گذاران و سياست‌گذاران بسيار مفيد فايده باشد. اين تحقيق در صدد آن است كه بهترين رويه‌هاي اندازه‌گيري ريسك را در بازار ايران به دست آورد. در اين راستا مدل قيمت گذاري دارايي‌هاي سرمايه گذاري ساده با مدل‌هايي كه در آنها عدم تقارن در توزيع بازدهي دارايي‌ها در نظر گرفته شده است مورد مقايسه قرار مي گيرد. با بررسي و مقايسه بيست شركت از شركت‌هاي پر معامله بورس اوراق بهادار تهران در بازه زمانيفروردين 1385 تا پايان اسفند 1390 كه به‌صورت ماهانه مورد آزمون قرار گرفت اين نتيجه به دست آمد كه در دوره مورد بررسي، مدل قيمت‌‌گذاري دارايي‌هاي سرمايه‌اي(CAPM)، مدل مناسب تري از مدل‌هاي تغييرات بخش پايين تر قيمت‌گذاري دارايي‌هاي سرمايه‌اي (LPMCAPM) و مدل واكنش نامتقارن (ARM) در بازار بورس ايران است.

In any market, knowing the best method of measuring risk can be very useful for investors and policymakers. In this regard, twenty of best seller corporations of Tehran Stock Exchange (TSE) were studied using monthly historical data (from April 2004 to March 2011). Several variations of the capital asset pricing model (CAPM), such as Lower Partial MomentCapital Asset Pricing Model (LPMCAPM), Asymmetric Response Model (ARM), and traditional CAPM were empirically tested. The results show that the traditional Capital Asset Pricing Model is the best model at the present period. JEL Classification: G12, G32 Theoretical Framework A plethora of empirical tests of the CAPM that implicitly assume the meanvariance based preference of the investors have been performed. However, statistical tractability of meanvariance analysis based on multivariate normality is a more important consideration in the development of the theory than the explicit recognition of investor preferences. Beginning from about the last quarter of the twentieth century, alternative theories based on different perceptions of systematic risk have challenged the dominance of the meanvariance notion of riskreturn relationship. The most prominent of these is the asset pricing theory which recognizes risk as the deviation below a target rate of return. Downside risk measures and the associated asset pricing models are motivated by economic and statistical considerations; investor psychology is consistent with asymmetric treatment of the variations in the returns and empirical return distributions appear to be nonnormal. Bawa and Lindenberg (1977) developed an asset pricing model, which we refer to as the meanlower partial moment (MLPM) model, based on downside risk. In the MLPM model, risk is defined as the deviation below the riskfree rate. For normal and studenttdistributions of returns, the MLPM model reduces to the conventional CAPM. In the MLPM model, the downside beta simply replaces the CAPM beta. Bawa and Lindenberg (1977) argue that their model explains the data at least as well as the CAPM does. Harlow and Rao (1989) developed an asset pricing model in the downside framework that is more general in that the risk is defined as the deviation below an arbitrary target rate. Methodology Twenty of best seller corporations of Tehran Stock Exchange (TSE) were studied using monthly historical data (from April 2004 to March 2011). Several variations of the capital asset pricing model (CAPM), such as Lower Partial MomentCapital Asset Pricing Model (LPMCAPM), Asymmetric Response Model (ARM) and traditional CAPM, were empirically tested. Results and Discussion The results show that, traditional Capital Asset Pricing Model, is better than Lower Partial MomentCapital Asset Pricing Model (LPMCAPM) and Asymmetric Response Model (ARM) in our selected period (from April 2004 to March 2011). Conclusions and Suggestions The results show that despite some empirical tests in recent years, it seems that by using long samples, traditional CAPM model could be a reliable test. Finally, authors suggest that to achieve more reliable results in CAPM models, researchers have to consider environmental conditions to use the best model.