string(33) ‘ model is supported by the data\. ‘

International Research Journal of Finance and Economics ISSN 1450-2887 Concern 4 (2006) © EuroJournals Publishing, Incorporation. 2006 http://www. eurojournals.

com/finance. htm Tests the Capital Asset Pricing Unit (CAPM): The truth of the Growing Greek Securities Market Grigoris Michailidis College or university of Macedonia, Economic and Social Savoir Department of Applied Informatics Thessaloniki, Portugal E-mail: [email, protected] grms Tel: 00302310891889 Stavros Tsopoglou University of Macedonia, Financial and Social Sciences Division of Applied Informatics Thessaloniki, Greece Email-based: [email, protected] r Tel: 00302310891889 Demetrios Papanastasiou School of Miscuglio, Economic and Social Savoir Department of Applied Informatics Thessaloniki, Greece E-mail: [email, protected] grms Tel: 00302310891878 Eleni Mariola Hagan Institution of Organization, Iona School New Rochelle Abstract The content examines the main city Asset Pricing Model (CAPM) for the Greek stock exchange using weekly stock results from 100 companies listed on the Athens stock exchange for the period of January 1998 to December 2002.

In order to shift away the firm-specific part of returns therefore enhancing the precision with the beta quotes, the investments where arranged into portfolios. The studies of this article are not supportive of the theory’s standard statement that higher risk (beta) is associated with higher levels of return. The model really does explain, however , excess comes back and thus lends support towards the linear framework of the CAPM equation. The CAPM’s conjecture for the intercept is that it should equal zero and the slope should certainly equal the surplus returns available portfolio.

The results with the study refute the above speculation and offer data against the CAPM. The testing conducted to examine the non-linearity of the romance between go back and betas support the hypothesis that the expected return-beta relationship is definitely linear. In addition , this daily news investigates perhaps the CAPM effectively captures all-important determinants of returns such as the residual Foreign Research Record of Financing and Economics , Issue 4 (2006) variance of stocks. The results illustrate that left over risk is without effect on the expected returns of portfolios.

Tests may possibly provide proof against the CAPM but they usually do not necessarily comprise evidence in support of any alternative unit (JEL G11, G12, and G15). Keywords: CAPM, Athens Stock Exchange, stock portfolio returns, beta, risk free rate, stocks JEL Classification: F23, G15 79 I. Advantages Investors and financial experts have paid extensive attention during the last few years for the new value markets that have emerged around the globe. This new interest has unquestionably been sparked by the significant, and in some cases incredible, returns provided by these market segments.

Practitioners all over the world use a variety of models in their stock portfolio selection process and their make an attempt to assess the risk exposure to several assets. One of the important improvements in modern day capital theory is the capital asset pricing model (CAPM) as manufactured by Sharpe [1964], Lintner [1965] and Mossin [1966]. CAPM suggests that excessive expected comes back are associated with high levels of risk. Simply stated, CAPM �vidence that the expected return with an asset over a risk-free level is linearly related to the non-diversifiable risk as assessed by the asset’s beta.

Even though the CAPM has become predominant in empirical work over the past 30 years and is the foundation of modern portfolio theory, amassing research has increasingly cast uncertainty on their ability to describe the actual moves of advantage returns. The purpose of this article is to measure thoroughly in case the CAPM is true in the capital market of Greece. Tests are executed for a period of five years (1998-2002), which can be characterized by strong return movements (covering historically high earnings for the Greek Wall street game as well as significant decrease in property returns over the examined period).

These industry return features make it possible with an empirical investigation of the charges model in differing economic conditions hence obtaining a conclusion under various stock go back volatility. Existing financial literature on the Athens stock exchange is rather scanty in fact it is the goal of this kind of study to widen the theoretical analysis of this industry by using contemporary finance theory and to give useful ideas for foreseeable future analyses with this market. II. Empirical evaluation of the unit and competitive studies from the model’s quality 2 . 1 .

Empirical appraisal of CAPM Since its introduction in early 60s, CAPM continues to be one of the most tough topics in financial economics. Almost any manager who would like to undertake a project must justify his decision partly based upon CAPM. The reason is that the version provides the means for a firm to calculate the return that its investors demand. It was the initially successful try to show how you can assess the risk of the cash flows of a potential investment task, to calculate the project’s cost of capital and the anticipated rate of return that investors will demand if they happen to be to invest in the project.

The model was developed to explain right after in the risk premium throughout assets. Based on the theory these differences are due to differences in the riskiness of the earnings on the property. The style states the fact that correct way of measuring the riskiness of an property is their beta and that the risk superior per device of riskiness is the same across most assets. Provided the risk cost-free rate as well as the beta of an asset, the CAPM anticipates the predicted risk superior for a property. The theory itself has been belittled for more than 30 years and has established a great academic debate regarding its convenience and quality.

In general, the empirical tests of CAPM has two broad purposes (Baily et al, [1998]): (i) to try whether or not the hypotheses should be rejected (ii) to supply information which could aid financial decisions. To achieve (i) assessments are carried out which could probably at least reject the model. The model goes by the test in case it is not possible to reject the hypothesis it is true. Ways of statistical examination need to be applied in order to attract reliable a conclusion on whether the 80 Foreign Research Record of Fund and Economics , Issue 4 (2006) model is usually supported by your data.

To accomplish (ii) the scientific work uses the theory as being a vehicle intended for organizing and interpreting the information without searching for ways of rejecting the theory. This type of approach is found in the area of portfolio decision-making, in particular based on the selection of resources to the bought or offered. For example , buyers are advised to buy or sell assets that according to CAPM will be underpriced or perhaps overpriced. In such a case empirical research is needed to assess the assets, determine their riskiness, analyze all of them, and place all of them into their particular categories.

A second illustration with the latter method appears in corporate fund where the approximated beta coefficients are used in assessing the riskiness of various investment assignments. It is then possible to calculate “hurdle rates” that projects need to satisfy if they are to be performed. This part of the paper concentrates on tests in the CAPM since its introduction inside the mid 1950’s, and explains the benefits of competing studies that attempt to measure the usefulness in the capital advantage pricing version (Jagannathan and McGrattan [1995]). 2 . installment payments on your

The classic support of the theory The model was developed in the early 1960’s by Sharpe [1964], Lintner [1965] and Mossin [1966]. In its basic form, the CAPM predicts that the anticipated return by using an asset over a risk-free charge is linearly related to the non-diversifiable risk, which is scored by the asset’s beta. One of the earliest scientific studies that found supporting evidence for CAPM is that of Black, Jensen and Scholes [1972]. Using month to month return data and portfolios rather than person stocks, Black et ing tested whether the cross-section of expected earnings is thready in beta.

By incorporating securities in to portfolios you can diversify away most of the firm-specific component of the returns, thereby enhancing the precision from the beta estimates and the anticipated rate of return in the portfolio investments. This approach mitigates the statistical problems that occur from measurement errors in beta estimates. The authors found that the data are consistent with the predictions of the CAPM i. electronic. the connection between the typical return and beta is incredibly close to linear and that portfolios with substantial (low) betas have high (low) normal returns.

An additional classic scientific study that supports the theory is that of Fama and McBeth [1973], they evaluated whether there exists a positive geradlinig relation between average comes back and beta. Moreover, the authors looked at whether the square-shaped value of beta and the volatility of asset earnings can make clear the residual deviation in normal returns across assets which are not explained by beta alone. 2 . 3. Issues to the quality of the theory In the early 1980s several studies recommended that there was deviations in the linear CAPM riskreturn trade-off due to different variables that affect this tradeoff.

The goal of the above research was to discover the components that CAPM was missing in explaining the risk-return trade-off and to recognize the variables that created those deviations. Banz [1981] tested the CAPM by simply checking if the size of organizations can make clear the residual variance in normal returns around assets that remain unexplained by the CAPM’s beta. This individual challenged the CAPM simply by demonstrating that firm size does describe the get across sectional-variation in average results on a particular collection of property better than beta.

The author concluded that the average returns on shares of tiny firms (those with low market ideals of equity) were greater than the average results on stocks of large companies (those with high market values of equity). This kind of finding is becoming known as the size effect. The study has been widened by evaluating different sets of parameters that might affect the riskreturn tradeoff. In particular, the income yield (Basu [1977]), leverage, and the proportion of a business book benefit of collateral to the market value (e. g.

Stattman [1980], Rosenberg, Reid and Lanstein [1983] and Chan, Hamao, Lakonishok [1991]) have all been utilized in tests the quality of CAPM. International Research Journal of Finance and Economics , Issue 5 (2006) 81 The general reaction to Banz’s [1981] findings, that CAPM can be missing a lot of aspects of actuality, was to support the view that although the info may recommend deviations via CAPM, these deviations are not so important about reject the theory. However , this kind of idea has become challenged simply by Fama and French [1992].

That they showed that Banz’s conclusions might be financially so important it raises severe questions about the quality of the CAPM. Fama and French [1992] used the same procedure as Fama and McBeth [1973] but arrived at very different results. Fama and McBeth find a positive connection between returning and risk while Fama and People from france find no relation at all. 2 . 4. The educational debate carries on The Celebridad and French [1992] research has alone been belittled. In general the studies addressing the Reputaci�n and France challenge more often than not take a nearer look at the data used in the study.

Kothari, Shaken and Sloan [1995] argue that Fama and French’s [1992] findings depend essentially on how the statistical findings will be interpreted. Amihudm, Christensen and Mendelson [1992] and Dark [1993] support the view the data are too noisy to invalidate the CAPM. In fact , they show that when an even more efficient statistical method is employed, the believed relation between average come back and beta is confident and significant. Black [1993] suggests that the size effect observed by Banz [1981] can simply be an example period effect i. at the. the size impact is noticed in some intervals and not in others.

Despite the above criticisms, the general reaction to the Reputaci�n and French [1992] studies has been to pay attention to alternative advantage pricing models. Jagannathan and Wang [1993] argue that this isn’t always necessary. Rather they demonstrate that the deficiency of empirical support for the CAPM could possibly be due to the inappropriateness of simple assumptions built to facilitate the empirical examination. For example , most empirical tests of the CAPM assume that the return on broad stock market indices is a great proxy to get the return on the market stock portfolio of all property in the economy.

Nevertheless , these types of market indexes will not capture every assets throughout the economy such as human being capital. Other empirical data on inventory returns is founded on the argument that the volatility of stock returns is consistently changing. When one views a time-varying return circulation, one need to refer to the conditional suggest, variance, and covariance that change according to currently available details. In contrast, the typical estimates of return, difference, and average squared deviations over a test period, offer an unconditional estimation because they treat difference as continuous over time.

One of the most widely used model to estimation the conditional (hence time- varying) variance of stocks and shares and stock index earnings is the generalized autoregressive conditional heteroscedacity (GARCH) model initiated by Robert. F. Engle. To summarize, each of the models over aim to improve the empirical testing of CAPM. There have also been numerous adjustments to the designs and perhaps the earliest or perhaps the subsequent alternate models validate or not really the CAPM is however to be determined. III. Sample selection and Data several. 1 . Sample Selection The study covers the time from January 1998 to December 2002.

This time period was chosen because it is seen as a intense return volatility with historically high and low returns pertaining to the Greek stock market. The selected sample consists of 100 stocks and options that are within the formation with the FTSE/ASE 20, FTSE/ASE Core 40 and FTSE/ASE Small Cap. These kinds of indices are made to provide real-time measures of the Athens Stock Exchange (ASE). The above mentioned indices happen to be formed subject to the following criteria: (i) The FTSE/ASE 20 index is the large cap index, containing the twenty largest blue chip corporations listed in the ASE. 82 International Analysis Journal of Finance and Economics , Issue 4 (2006) ii) The FTSE/ASE Mid 45 index is definitely the mid cover index and captures the performance of the next forty companies in size. (iii) The FTSE/ASE Small Cap index is the tiny cap index and reflects the performance of the subsequent 80 businesses. All investments included in the directories are bought and sold on the ASE on a ongoing basis throughout the full Athens stock exchange trading-day, and are chosen according to prespecified fluid criteria set by the ASE Advisory Committee1. For the purpose of the study, 100 stocks and options were selected from the pool of investments included in the aforementioned indices.

Every single series contains 260 findings of the every week closing prices. The selection was performed on the basis of the trading volume and excludes stocks that have been traded irregularly or experienced small trading volumes. 3. 2 . Data Selection The analysis uses every week stock returns from 75 companies listed on the Athens stock market for the period of January 1998 to December 2002. The data will be obtained from MetaStock (Greek) Data Base. To be able to obtain better estimates with the value in the beta agent, the study utilizes weekly share returns. Comes back calculated using a longer time frame (e. g. onthly) may possibly result in adjustments of beta over the examined period presenting biases in beta quotes. On the other hand, higher frequency data such as daily observations covering a short and stable time span can result in the application of very loud data and so yield ineffective estimates. Almost all stock results used in the analysis are modified for payouts as needed by the CAPM. The ASE Composite Share index is employed as a proxy server for industry portfolio. This index is known as a market value measured index, is definitely comprised of the 60 many highly capitalized shares from the main industry, and shows general styles of the Ancient greek language stock market.

Furthermore, the 3-month Greek Treasury Bill is used as the proxy pertaining to the free of risk asset. The yields had been obtained from the Treasury A genuine and Expenses Department from the National Lender of Greece. The deliver on the 3-month Treasury costs is specifically chosen since the benchmark that better reflects the short-term changes in the Greek monetary markets. IV. Methodology The first step was to calculate a beta coefficient for every single stock employing weekly returns during the period of January 1998 to December 2002. The beta was believed by regressing each stock’s weekly go back against the industry index according to the following formula: Rit , R ft = a i +? ( Rmt , L ft ) + eit (1) in which, Rit is a return about stock my spouse and i (i=1…100), L ft is the rate of return on a risk-free advantage, Rmt is a rate of return available index,? i actually is the approximate of beta for the stock i, and eit is the matching random disruption term inside the regression formula. [Equation 1 could also be expressed applying excess go back notation, where ( Rit , R ft ) = rit and ( Rmt , Rft ) = rmt ]

In spite of the truth that each week returns were used to steer clear of short-term noises effects the estimation analysis tests pertaining to equation (1) indicated, in numerous occasions, departures from the linear assumption. www. ase. grms International Study Journal of Finance and Economics , Issue 5 (2006) 83 In such cases, formula (1) was re-estimated providing for EGARCH (1, 1) form to comfort with misspecification. The next measure was to calculate average profile excess returns of stocks and options ( rpt ) bought according with their beta coefficient computed simply by Equation 1 ) Let, rpt =? 3rd there�s r i =1 k it k (2) where, e is the volume of stocks included in each profile (k=1…10), s is the number of portfolios (p=1…10), rit is definitely the excess go back on stocks and options that form each profile comprised of e stocks every.

This procedure made 10 equally-weighted portfolios comprised of 10 shares each. Simply by forming portfolios the propagate in betas across portfolios is maximized so that the a result of beta in return could be clearly evaluated. The most obvious way to form portfolios is to rank stocks into portfolios by the true beta. But , all that is available is definitely observed beta. Ranking in to portfolios simply by observed beta would introduce selection prejudice. Stocks with high-observed beta (in the greatest group) can be more likely to have a positive way of measuring error in estimating beta.

This would bring in a positive prejudice into beta for high-beta portfolios and would present a negative prejudice into an estimate of the intercept. (Elton and Gruber [1995], p. 333). Combining securities into portfolios diversifies away a lot of the firm-specific element of returns thereby enhancing the precision with the estimates of beta plus the expected level of come back on the portfolios on investments. This minimizes statistical conditions that arise coming from measurement error in the beta estimates. The next equation utilized to estimate portfolio betas: rpt sama dengan a s +? g? mt + e pt (3) where, rpt is the average extra portfolio return,? p is a calculated stock portfolio beta. The study continues by simply estimating the ex-post Protection Market Collection (SML) by simply regressing the portfolio earnings against the portfolio betas attained by Equation 3. The relation reviewed is the subsequent: rP sama dengan? 0 +? 1? G + elizabeth P (4) where, rp is the common excess come back on a collection p (the difference between return for the portfolio and the return on the risk-free asset),? p can be an estimate of beta of the portfolio l,?

1 is a market price of risk, the danger premium for bearing one unit of beta risk,? is the zero-beta rate, the expected return on an property which has a beta of actually zero, and elizabeth p is definitely random disruption term inside the regression equation. In order to check for nonlinearity between total portfolio earnings and betas, a regression was run on average stock portfolio returns, calculated portfolio beta, and beta-square from formula 3: two rp =? 0 &? 1? p +? 2? p & e l (5) Finally in order to look at whether the residual variance of stocks influences portfolio results, an additional term was included in equation your five, to test intended for the informative power of nonsystematic risk: a couple of rp =? +? one particular? p +? 2? s +? three or more? RVp + e s (6) where 84 Foreign Research Log of Finance and Economics , Concern 4 (2006) RV p is the residual variance of portfolio returns (Equation 3), RV g =? a couple of (e pt ). The estimated guidelines allow all of us to test several hypotheses regarding the CAPM. The tests are: i)? 3 = zero or left over risk would not affect come back, ii)? 2 = zero or you will find no non-linearities in the security market range, iii)? one particular >, 0 that is, we have a positive price of risk in the capital markets (Elton and Gruber [1995], p. 336).

Finally, the above mentioned analysis was also executed for each 12 months separately (1998-2002), by changing the stock portfolio compositions in respect to every year estimated betas. V. Scientific results and Interpretation of the findings The first area of the methodology essential the appraisal of betas for individual stocks and shares by using findings on prices of return for a collection of dates. Useful feedback can be produced from the effects of this process, for the assets employed in this study. The range from the estimated stock betas is definitely between 0. 0984 the minimum and 1 . 4369 the maximum which has a standard deviation of zero. 240 (Table 1). Almost all of the beta coefficients for individual stocks and shares are statistically significant for a 95% level and all estimated beta coefficients happen to be statistical significant at a 90% level. For a more accurate estimation of betas a great EGARCH (1, 1) model was used anywhere it was important, in order to appropriate for non-linearities. Table one particular: Stock beta coefficient estimations (Equation 1)

Stock name beta Stock name beta Stock term OLYMP. 0984 THEMEL. 8302 PROOD EYKL. 4192 AIOLK. 8303 ALEK MPELA. 4238 AEGEK. 8305 EPATT MPTSK. 5526 AEEXA. 8339 SIDEN FOIN. 5643 SPYR. 8344 GEK GKOYT. 862 SARANT. 8400 ELYF PAPAK. 6318 ELTEX. 8422 MOYZK ABK. 6323 ELEXA. 8427 TITK MYTIL. 6526 MPENK. 8610 NIKAS FELXO. 6578 HRAKL. 8668 ETHENEX ABAX. 6874 PEIR. 8698 IATR TSIP. 6950 BIOXK. 8747 METK AAAK. 7047 ELMEK. 8830 ALPHA EEEK. 7097 LAMPSA. 8848 AKTOR ERMHS. 7291 MHXK. 8856 INTKA LAMDA. 7297 DK. 8904 MAIK OTE. 7309 FOLI. 9005 PETZ MARF. 7423 THELET. 9088 ETEM MRFKO. 7423 ATT. 9278 FINTO KORA. 7520 ARBA. 9302 ESXA RILK. 7682 KATS. 9333 BIOSK LYK. 7684 ALBIO. 9387 XATZK ELASK. 7808 XAKOR. 9502 KREKA NOTOS. 8126 SAR. 9533 ETE KARD. 8290 NAYP. 577 SANYO Origin: Metastock (Greek) Data Base and measurements (S-PLUS) beta. 9594. 9606. 9698. 9806. 9845. 9890. 9895. 9917. 9920 1 . 0059 1 . 0086 1 ) 0149 1 . 0317 1 ) 0467 1 . 0532 1 ) 0542 1 . 0593 1 . 0616 1 . 0625 1 ) 0654 1 ) 0690 1 . 0790 1 . 0911 1 ) 1127 1 ) 1185 Share name EMP NAOYK ELBE ROKKA SELMK DESIN ELBAL ESK TERNA KERK POYL EEGA KALSK GENAK FANKO PLATH STRIK EBZ ALLK GEBKA AXON RINTE KLONK ETMAK ALTEK beta 1 . 1201 1 . 1216 1 ) 1256 1 . 1310 1 . 1312 1 ) 1318 1 . 1348 1 . 1359 1 ) 1392 1 . 1396 1 . 1432 1 ) 1628 1 . 1925 1 ) 1996 1 . 2322 1 ) 2331 1 ) 2500 1 ) 2520 1 . 2617 1 ) 2830 1 ) 3030 1 . 3036 1 . 3263 1 . 3274 1 ) 4369

The content argues that particular hypotheses may be tested regardless of whether a single believes in the validity of the simple CAPM or in different other version of the theory. Firstly, the theory indicates that higher risk (beta) is connected with a higher level of return. Nevertheless , the benefits of the examine do not International Research Log of Financing and Economics , Concern 4 (2006) 85 support this speculation. The beta coefficients in the 10 portfolios do not suggest that larger beta portfolios are related with higher earnings. Portfolio 15 for example , the best beta collection (? = 1 . 2024), yields negative portfolio earnings.

In contrast, stock portfolio 1, the lowest beta stock portfolio (? sama dengan 0. 5474) produces great returns. These contradicting benefits can be somewhat explained by the significant fluctuations of stock returns over the period examined (Table 2). Desk 2: Average excess portfolio returns and betas (Equation 3) rp beta (p) a10. 0001. 5474 b10. 0000. 7509 c10 -. 0007. 9137 d10 -. 0004. 9506 e10 -. 0008. 9300 f10 -. 0009. 9142 g10 -. 0006 1 ) 0602 h10 -. 0013 1 . 1066 i10 -. 0004 1 . 1293 j10 -. 0004 1 . 2024 Average Rf. 0014 Average rm=(Rm-Rf). 0001 Source: Metastock (Greek) Info Base and calculations (S-PLUS) Portfolio Var.

Error. 0012. 0013. 0014. 0014. 0009. 0010. 0012. 0019. 0020. 0026 R2. 4774. 5335. 5940. 6054. 7140. 6997. 6970. 6057. 6034. 5691 In order to test out the CAPM hypothesis, you need to find the counterparts towards the theoretical beliefs that must be used in the CAPM equation. From this study the yield for the 3-month Ancient greek language Treasury Expenses was used while an approximation of the free of risk rate. Intended for the Ur m, the ASE Amalgamated Share index is taken as the best approximation for the market portfolio. The fundamental equation employed was rP =? zero +? 1? P & e S (Equation 4) where? is a expected excessive return on a zero beta portfolio and? 1 is a market price of risk, the difference between the anticipated rate of return in the marketplace and a zero beta portfolio. One way for enabling the possibility that the CAPM does not hold authentic is to add an intercept in the evaluation of the SML. The CAPM considers the intercept is definitely zero for each asset. Hence, a test out can be constructed to examine this hypothesis. To be able to diversify aside most of the firm-specific part of comes back, thereby enhancing the accuracy of the beta estimates, the securities had been previously mixed into portfolios.

This approach mitigates the record problems that come up from way of measuring errors in individual beta estimates. These portfolios were created for many reasons: (i) the randomly influences in individual stocks and options tend to end up being larger compared to those upon suitably made portfolios (hence, the intercept and beta are easier to estimate pertaining to portfolios) and (ii) the tests to get the intercept are easier to implement intended for portfolios since by building their predicted coefficients are much less likely to be linked to one another than the shares of individual businesses.

The quality value of the believed correlation coefficient between the intercept and the slope indicates the model utilized explains excessive returns (Table 3). 86 International Analysis Journal of Finance and Economics , Issue some (2006) Table 3: Stats of the evaluation of the SML (Equation 4) Coefficient? 0 Value. 0005 t-value (. 9011) p-value. 3939 Left over standard error:. 0004 on 8 degrees of freedom Multiple R-Squared:. 2968 F-statistic: several. 3760 in 1 and 8 degrees of freedom, the p-value is usually. 1034 Relationship of Rapport? 0,? you =. 9818? 1 -. 0011 (-1. 8375). 1034

However , the simple fact that the intercept has a benefit around actually zero weakens these explanation. The results on this paper is very much inconsistent with the zero beta version from the CAPM for the reason that intercept of the SML is usually not higher than the interest price for risk free-bonds (Table 2 and 3). In the estimation of SML, the CAPM’s prediction pertaining to? 0 is that it should be equal to zero. The calculated benefit of the intercept is small (0. 0005) but it is usually not drastically different from no (the tvalue is not greater than 2) Hence, depending on the intercept criterion alone the CAPM hypothesis are unable to clearly be rejected.

In accordance to CAPM the SLM slope ought to equal the surplus return available portfolio. The surplus return available portfolio was 0. 0001 while the approximated SLM incline was – 0. 0011. Hence, these result also indicates there is evidence against the CAPM (Table 2 and 3). To be able to test for nonlinearity between total stock portfolio returns and betas, a regression was run among average profile returns, calculated portfolio betas, and the square of betas (Equation 5). Results present that the intercept (0. 0036) of the equation was greater than the free of risk interest rate (0. 014),? 1 was bad and different via zero while? 2, the coefficient of the square beta was really small (0. 0041 with a t-value not higher than 2) and thus consistent with the hypothesis that the anticipated return-beta marriage is thready (Table 4). Table 4: Testing intended for Non-linearity (Equation 5) Agent? 0 Worth. 0036 t-value (1. 7771) p-value 0. 1188 Recurring standard mistake:. 0003 in 7 examples of freedom Multiple R-Squared:. 4797 F-statistic: a few. 2270 about 2 and 7 degrees of freedom, the p-value can be. 1016? one particular -. 0084 (-1. 8013) 0. 1147? 2 . 0041 (1. 5686) 0. 1607

According to the CAPM, expected returns vary across assets because the assets’ betas are very different. Hence, one way to investigate whether CAPM sufficiently captures all-important aspects of the risk-return tradeoff is to evaluation whether other asset-specific qualities can make clear the crosssectional differences in average returns that cannot be related to cross-sectional differences in beta. To do this task the residual variance of portfolio returns was added as one more explanatory adjustable (Equation 6). The pourcentage of the residual variance of portfolio earnings? 3 can be small and not really statistically unlike zero.

It is therefore safe in conclusion that recurring risk has no affect for the expected come back of a protection. Thus, when portfolios are being used instead of individual stocks, recurring risk no more appears to be important (Table 5). International Analysis Journal of Finance and Economics , Issue four (2006) Stand 5: Tests for nonsystematic risk (Equation 6) Agent? 0? 1 Value. 0017 -. 0043 t-value (. 5360) (-. 6182) p-value 0. 6113 0. 5591 Residual regular error:. 0003 on 6th degrees of freedom Multiple R-Squared:. 5302 F-statistic: 2 . 2570 on a few and six degrees of freedom, the p-value is. 1821? 2 . 0015 (. 3381) 0. 7468? 3. 3503 (. 8035) 0. 523 87 Considering that the analysis for the entire five-year period would not yield good evidence for the CAPM we evaluated whether a similar approach in yearly info would provide more supportive evidence. All models had been tested independently for each with the five-year period and the results were statistically better for some years but still did not support the CAPM hypothesis (Tables 6, 7 and 8).

Table 6: Stats of the evaluation SML (yearly series, Equation 4) 98 1999 2k 2001 2002 Coefficient? 0? 1? 0? 1? 0? 1? 0? 1? zero? 1 Worth. 0053. 0050. 0115. 0134 -. 0035 -. 0149. 0000 -. 0057 -. 0017 -. 0088 t-value (3. 7665) (2. 231) (2. 8145) (4. 0237) (-1. 9045) (-9. 4186) (. 0025) (-2. 4066) (-. 8452) (-5. 3642) Std. Mistake. 0014. 0022. 0041. 0033. 0019. 0016. 0024. 0028. 0020. 0016 p-value. 0050. 0569. 2227. 0038. 0933. 0000. 9981. 0427. 4226. 0007 Table 7: Screening for Non-linearity (yearly series, Equation 5) 1998 Coefficient? 0? one particular? 2? 0? 1? a couple of? 0? 1? 2? zero? 1? two? 0? 1? 2 Worth. 0035. 0139 -. 0078. 0030 -. 0193. 0135 -. 0129. 0036 -. 0083. 0092 -. 0240. 0083 -. 0077. 0046 -. 0059 t-value (1. 7052) (1. 7905) (-1. 1965) (2. 1093) (-. 7909) (1. 3540) (-3. 5789) (. 5435) (-2. 8038) (1. 2724) (-1. 7688) (1. 3695) (-2. 9168) (. 139) (-2. 7438) Std. Error. 0020. 0077. 0065. 0142. 0243. 0026. 0036. 0067. 0030. 0072. 0136. 0060. 0026. 0050. 0022 p-value. 1319. 1165. 2705. 0729. 4549. 0100. 0090. 6037. 0264. 2439. 1202. 2132. 0224. 3911. 0288 1999 2150 2001 2002 88 Worldwide Research Journal of Financial and Economics , Concern 4 (2006) Table 8: Testing intended for Non-Systematic risk (yearly series, Equation 6) 1998 Agent? 0? one particular? 2? 3? 0? 1? 2? several? 0? one particular? 2? several? 0? 1? 2? three or more? 0? 1? 2? 3 Value. 0016. 0096 -. 0037 a few. 0751. 0017 -. 0043. 0015. 3503 -. 0203. 0199 -. 0185 installment payments on your 2673. 0062 -. 0193. 0053 1 ) 7024 -. 0049. 500 -. 0026 -5. 1548 t-value (. 7266) (1. 2809) (-. 5703) (. 5862) (1. 4573) (-. 0168) (. 0201) (2. 2471) (-4. 6757) (2. 2305) (-3. 6545) (2. 2673) (. 6019) (-1. 0682) (. 5635) (. 4324) (-. 9507) (. 0054) (-. 4576) (-. 6265) Std. Error. 0022. 0075. 0065 1 . 9615. 0125. 0211. 0099 1 ) 4278. 0043. 0089. 0051. 9026. 0103. 0181. 0094 3. 9369. 0052. 0089. 0058 8. 2284 p-value. 4948. 2475. 5892. 1680. 1953. 9846. 9846. 0657. 0034. 0106. 0106. 0639. 5693. 3265. 5935. 6805. 3785. 9959. 6633. 5541 1999 2000 2001 2002 VI. Ending Remarks This article examined the validity in the CAPM pertaining to the Ancient greek language stock market.

The analysis used regular stock results from 75 companies on the Athens stock exchange from January 1998 to December 2002. The studies of the document are not supporting of the theory’s basic hypothesis that the upper chances (beta) can be associated with a higher level of come back. In order to diversify away most of the firm-specific component to returns thus enhancing the precision of the beta estimates, the investments where combined into portfolios to reduce the record problems that come up from dimension errors in individual beta estimates. The model really does explain, nevertheless , excess results.

The results obtained loan support for the linear structure of the CAPM equation as being a good justification of security returns. The high value in the estimated correlation coefficient involving the intercept and the slope signifies that the unit used, explains excess comes back. However , the very fact that the intercept has a benefit around no weakens the above explanation. The CAPM’s prediction for the intercept is the fact it should be comparable to zero and the slope ought to equal the surplus returns available portfolio. The findings in the study contradict the above hypothesis and reveal evidence resistant to the CAPM.

The inclusion with the square in the beta coefficient to test pertaining to nonlinearity in the relationship among returns and betas signifies that the studies are in line with the hypothesis and the expected returnbeta relationship can be linear. In addition , the checks conducted to check into whether the CAPM adequately captures all-important facets of reality by simply including the left over variance of stocks implies that the residual risk does not have any effect on the expected return on portfolios. The lack of solid evidence in favour of CAPM necessitated the study of every year data to check the quality of the model.

The results from this procedure provided better statistical results for some years but still would not support the CAPM hypothesis. The results of the assessments conducted on data through the Athens stock exchange for the time of January 1998 to December 2002 do not apparently clearly deny the CAPM. This does not imply that the data do not support CAPM. As Dark-colored [1972] points out these outcomes can be described in 2 different ways. First, way of measuring and version specification problems arise because of the use of a proxy rather than the actual industry International Analysis Journal of Finance and Economics , Issue some (2006) fifth there�s 89 ortfolio. This error biases the regression line believed slope to zero and its estimated intercept away from no. Second, if no free of risk asset is out there, the CAPM does not forecast an intercept of zero.