Advanced Financial Modeling - Assignment Example

Advanced Financial Modeling Student Name: Student ID: Section A A1. Theoretical assumptions and implications of the Capital Assets Pricing Model (CAPM) According to Lintner (1965) and Sharpe (1964), the following assumptions are made in the CAPM model 1. Perfect information All investors can access the same information Analysis of information is done in the same manner 2. Markets are smooth without friction There are no taxes No transaction costs/No commissions 3. Security markets compete perfectly Many investors who compete perfectly (price takers) 4. Myopic investors All investors have one and same holding period 5. All investors rationally optimize mean and variance Everyone uses Markowitz portfolio selection criterion 6. Investments are limited to publicly traded assets with unlimited lending at the risk-free rate 7. Investors can borrow and lend at the one risk free rate 8. Investors agree on the prospects of various investments such as expected values, variance, correlation and standard deviations implying homogeneity of beliefs about investments 9. Any investor can short any asset and hold any fraction of the asset 10. All investors have the same planning horizon 11. Common pure rate of interest implying that all investors are able to borrow or lend funds on equal terms Implications 1. Investors will combine a risk free asset with a market portfolio of risky assets and make investments in risky assets in accordance with their market value. 2. Investors should expect returns from their investment according to the risk. 3. In the event that the investors cannot diversify on a risk, they are compensated. 4. All efficient combinations will be perfectly correlated A2. How valid the CAPM Assumptions are and extent to which breach of assumptions invalidate the CAPM model The CAPM assumptions are mainly valid based on the investor behavior and for a single risk factor. When firms are not sorted on metrics such as price/book or price/earnings, the investors’ subjective reactions may not be exposed. Under these circumstances, they won’t over-forecast past performance that would otherwise lead to increased stock prices for high price/earnings firms and too low for low P/E firms. Under the circumstances, the assumptions of CAPM would be breached. The assumption of a single risk factor then limits the validity of the CAPM model (Schmidt, 2008). A3. Empirical test of CAPM This part of assignment refers to the appendix 1 A. The model output for shop all years shows that i.e the model value p>0.05 implies statistical insignificance of value. The same holds for technology stock. As for the 10 year period regression, third 10 year regression shows significant only for the shop. The rest show insignificant Looking at the data profile, the is significant if the mean stock difference is closer to the market free rate. The model shows on average that the market condition accounts for over 75% of the stock price. From the model, implies that at the point of equilibrium (when market value=stock value), then the CAPM’s equilibrium state, through the assumptions shows that all the investors will see their respective alternatives (on the indifference curve) in the same manner as shown below Figure 1: CAPM under market equilibrium (Sharpe). This is in agreement with the CAPM model under the assumptions in equilibrium state. A4. Fama French three factor model:) Background and key features The Fama French model is used to explain differences in the returns of diversified stocks. It compares a portfolio to 3 distinct risks found in the equity market to assist in decomposing returns. According to this model, portfolio’s beta explains for only about 70% of actual returns and the remaining 30% is accounted for by factors not related to beta. Due to this, the model is known to account for up to 95% of returns for a cross-section of equity portfolios of different sizes and styles (Lohrmann, 2015). The key features of the model include: 1. The stock risk/return is useful in explaining the return of a portfolio 2. The return of any stock portfolio can be explained almost entirely by only including an additional two factors which include the Market Cap/(SMB) and book/market (HML) ratio (value). A portfolio with a large size (SM) and a high book/market ratio will have a higher expected return than that of a low book/market ratio. A5. Empirical test for Fama French model , , , For both the Shop and technology industries, the model constant value is not significant (p>0.0500), hence the null hypothesis is true. i.e . However, the model results show that and values are statistically significant hence the null hypotheses for the coefficients is not true. For a detailed model result see the appendix 1B for Fama French model for both the shop and technology industries under the coefficients table 1B. For comparison purposes, intercept is 0 for both cases CAPM and Fama French. However, comparing the model summary for the two models in the cases of shop and technology, it can be seen that there is a difference in what is accounted for by the particular model. Generally shops are smaller stock values and the tech is larger stock values thus the difference between the stock and risk free rate is higher for technology than for shop industry. The CAPM model accounts approximately 60% for variation in shop stock value and 57% for technology. On the other hand, Fama French model accounts for approximately 78% for shop and approximately 82% for technology. From the model results, it can be seen for both small and large stocks that the Fama French model accounts more for a given variation from the free rate as compared to the CAPM model. The CAPM is seen to predict smaller stocks more accurately compared to bigger ones. The Fama French model, predicts bigger stocks more accurately than it would for smaller ones. However, it should be noted that in either cases, the Fama French model is more accurate compared to the CAPM (i.e either small or large stocks). From this it can be seen that increasing size Market Cap increases the part accounted for the model. It confirms further that depending size and style, Fama French three factor model can account for a very large portion (up to 95%) of the expected returns (Lohrmann, 2015). Section B: Factors affecting wages in the US B1. Descriptive statistics Table 9: Descriptive Statistics for wages and factors affecting the wages N Minimum Maximum Mean Std. Deviation WAGE 534 2 67 13.58 7.717 OBS. 534 1 534 267.50 154.297 EDUCATION 534 2 18 13.02 2.615 EXPERIENCE 534 0 55 17.82 12.380 GENDER 534 0 1 .46 .499 AGE 534 18 64 36.83 11.727 MARRIED 534 0 1 .66 .476 UNION 534 0 1 .18 .384 Valid N (listwise) 534 B2. Regression model 1: wages and education Table: Model Summary and regression Model R R Square Adjusted R Square Std. Error of the Estimate 1 .381a .145 .143 7.143 Figure 2: Graph showing relationship between wage and education From the tables and the graph, R squared is 0.145 implying that education accounts for 14.5% of the wages earned. This implies that wages increase by 14.5% for every additional year of education. B3. Collective impact of education and experience Table: Coefficients for education and experience Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -7.224 1.833 -3.942 .000 EDUCATION 1.384 .122 .469 11.309 .000 EXPERIENCE .156 .026 .251 6.040 .000 Table: Model Summary for combined impact of education and experience on wages Model R R Square Adjusted R Square Std. Error of the Estimate 1 .447a .200 .197 6.916 From the tables, it can be concluded that a combined effect of education and experience accounts for 20% (R squared=0.200) of wages. It means wages will increase by 20% for a combined increase of education and experience by one year. This is 6% more than the first model meaning a combination of education and experience lead to increased wages more than education along. B4. Regression model 3 for education, experience and gender on wages Table: Coefficients of education, experience and gender Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -6.104 1.783 -3.423 .001 EDUCATION 1.406 .118 .477 11.868 .000 EXPERIENCE .169 .025 .270 6.716 .000 GENDER -3.552 .583 -.230 -6.091 .000 Table: Model Summary- predictions Model R R Square Adjusted R Square Std. Error of the Estimate 1 .502a .252 .248 6.692 From the tables, gender affects the wages. The numerical values assigned to gender are dummy. Gender is a dummy variable. A dummy variable is a numerical variable used in regression analysis to represent subgroups of the study sample. The dummy variable is used to distinguish different treatment groups. It can be described as a categorical variable or simply put, it’s a qualitative variable. Since the numerical assignment is with due consideration and/or a measure of the wage weight relative to gender, the index is a measure of the gender impact level. To this effect, it is expected that value 1(female) receives less wages. Therefore, in this case, since the wage coefficient shows a negative relationship, it is expected that wage value 1 gets less wages than age value 0 (male). In this regard, it is considered that male gender would not have any impact on wages in terms of gender but being a female reduces value to gender in terms of wages i.e all factors notwithstanding, a female would get less wages than male. B5. Correlation matrix Table: Correlation matrix for factors affecting wages WAGE EDUCATION EXPERIENCE GENDER AGE WAGE Pearson Correlation 1 .381** .085* -.208** .175** Sig. (2-tailed) .000 .049 .000 .000 N 534 534 534 534 534 EDUCATION Pearson Correlation .381** 1 -.353** .002 -.150** Sig. (2-tailed) .000 .000 .963 .001 N 534 534 534 534 534 EXPERIENCE Pearson Correlation .085* -.353** 1 .075 .978** Sig. (2-tailed) .049 .000 .082 .000 N 534 534 534 534 534 GENDER Pearson Correlation -.208** .002 .075 1 .079 Sig. (2-tailed) .000 .963 .082 .068 N 534 534 534 534 534 AGE Pearson Correlation .175** -.150** .978** .079 1 Sig. (2-tailed) .000 .001 .000 .068 N 534 534 534 534 534 MARRIED Pearson Correlation .104* -.036 .271** .011 .279** Sig. (2-tailed) .017 .413 .000 .796 .000 N 534 534 534 534 534 UNION Pearson Correlation .163** -.024 .118** -.157** .119** Sig. (2-tailed) .000 .582 .006 .000 .006 N 534 534 534 534 534 Correlations MARRIED UNION WAGE Pearson Correlation .104 .163** Sig. (2-tailed) .017 .000 N 534 534 EDUCATION Pearson Correlation -.036** -.024 Sig. (2-tailed) .413 .582 N 534 534 EXPERIENCE Pearson Correlation .271* .118** Sig. (2-tailed) .000 .006 N 534 534 GENDER Pearson Correlation .011** -.157 Sig. (2-tailed) .796 .000 N 534 534 AGE Pearson Correlation .279** .119** Sig. (2-tailed) .000 .006 N 534 534 MARRIED Pearson Correlation 1* .093 Sig. (2-tailed) .031 N 534 534 UNION Pearson Correlation .093** 1 Sig. (2-tailed) .031 N 534 534 From the correlation matrix table above, it can be concluded certainly that the wage of a person is significantly determined by both the factors under study which include education, experience, gender, age, marital status and union membership. Except for gender which is negatively correlated with wages ( where females get less by virtue of gender value), all other factors have a positive correlation with wages, meaning that a unit increase in a given factor is accompanied by an increase in wages by a value equal to the correlation factor of the value. From the table, it can also be seen how the factors interact with each other. Education has a significantly negative correlation with age and experience. That is, higher education is seen amongst less experienced and younger persons. Experience positively correlates with age, marital status (married are more experienced), and union. Gender has a positive correlation with only marital status (where female are more likely to be of married status than male). Age correlates with all other factors except for gender. In short, it is good to note that although, at individual level, the factors are determined to be having effect on the wages, they are also interrelated differently where one factor can influenced by others and/or not the others. B6. Regression model for all the factors (age, education, gender, married, experience and union) Table: Coefficients of factors affecting wages Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -2.593 10.197 -.254 .799 EDUCATION 2.050 1.672 .695 1.226 .221 EXPERIENCE .813 1.673 1.305 .486 .627 GENDER -3.275 .588 -.212 -5.567 .000 AGE -.663 1.671 -1.007 -.396 .692 MARRIED .779 .632 .048 1.232 .219 UNION 2.169 .767 .108 2.828 .005 Table: Model Summary for thepart accounted for by the model Model R R Square Adjusted R Square Std. Error of the Estimate 1 .516a .266 .258 6.648 This model only accounts for approximately 27% of the total wages. First of all, it is good enough to say that the model cannot be applied to predict the wages of a particular person effectively. It means that some of the high impact factors to determine the wages have been left out. One more thing that makes this model less effective is the fact that in the absence of all other factors, the wage of a person is in negatives and the woman has even more negative wages. This may not be empirically true since the minimum wage of a person is 0 (i.e for a man who is not working) (Doing business in 2011, 2010). B7 Important factors and problems that need to be considered when designing a financial model When developing a financial model, the key factors to consider include validity of the data being used and possible sources of error. A consideration of the sources implies that all factors that could affect a dependent variable should all be involved. One other major factor to consider is the source of error. The errors/factors in this respect include the following: 1. Formula errors which are the easiest errors to make 2. The assumption/input error. This can be identified by making sure the assumptions are all clearly documented within the model. 3. Logic errors Which are hard to identify. It particularly comes by when the formulas and assumptions are correct, yet the model behavior shows inaccuracies for a given test situation. A model is developed based information systems that require key aspects. Some of the key requirements include: knowing what information exists and what it is about, extracting a portion of information for a particular purpose, managing data, including history, for life. Meeting all the requirements of an information system can be difficult and expensive. One other major problem that faces development of financial models is the reality of the computer-based information. A number of problems occur as a result of the way information systems hold data. Arbitrary or inappropriate restrictions can be placed due to data structures and constraints. False data can be introduced to overcome the imposed restrictions and uncontrolled redundancy arises from the same data occurring and being updated in multiple systems. These and more computer related problems become quite a challenge in financial modeling (Ho and Yi, 2004). In doing so, therefore, the development of a financial model heavily relies on assumptions. A model cannot be developed without a model. The formula errors and computer related problems can be easily identified. The ultimate solution to the model problems would be to reduce the number of assumptions as much as possible. This can be explained by Occam’s razor which is a logical principle attributed to William of Occam The principle states that “one should not make more assumptions than the minimum needed”. It is a very important guide that helps us to shave away the variables that are not necessary in explaining the phenomenon (Martineau, 2000). It makes the models simpler with the most befitting solution Appendix 1A: Model Output for CAPM Regression-output for Shop-all years Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: ShpRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .777a .604 .603 3.88178 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 8226.228 1 8226.228 545.932 .000b Residual 5394.427 358 15.068 Total 13620.655 359 a. Dependent Variable: ShpRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .083 .206 .403 .687 Mkt-RF 1.042 .045 .777 23.365 .000 a. Dependent Variable: ShpRF Regression-output for technology for all the years Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: TechRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .759a .576 .575 5.77599 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 16253.182 1 16253.182 487.175 .000b Residual 11943.635 358 33.362 Total 28196.817 359 a. Dependent Variable: TechRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .148 .307 .484 .629 Mkt-RF 1.465 .066 .759 22.072 .000 a. Dependent Variable: TechRF Regression output for the 10 year periods Regression-1st 10 year shop Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: ShpRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .855a .731 .728 2.93946 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 2765.458 1 2765.458 320.061 .000b Residual 1019.567 118 8.640 Total 3785.025 119 a. Dependent Variable: ShpRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -.079 .271 -.293 .770 Mkt-RF .988 .055 .855 17.890 .000 a. Dependent Variable: ShpRF Regression 1st 10-year technology Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: TechRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .842a .709 .707 4.01256 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 4630.531 1 4630.531 287.600 .000b Residual 1899.871 118 16.101 Total 6530.402 119 a. Dependent Variable: TechRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -.574 .370 -1.551 .124 Mkt-RF 1.278 .075 .842 16.959 .000 a. Dependent Variable: TechRF Regression-2nd 10 year shop Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: ShpRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .706a .498 .494 3.77258 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 1666.046 1 1666.046 117.060 .000b Residual 1679.420 118 14.232 Total 3345.465 119 a. Dependent Variable: ShpRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -.484 .357 -1.358 .177 Mkt-RF .937 .087 .706 10.819 .000 a. Dependent Variable: ShpRF Regression-2nd 10 year tech Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: TechRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .680a .463 .458 5.55859 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 3143.522 1 3143.522 101.739 .000b Residual 3645.956 118 30.898 Total 6789.478 119 a. Dependent Variable: TechRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .556 .525 1.059 .292 Mkt-RF 1.288 .128 .680 10.087 .000 a. Dependent Variable: TechRF Regression-3rd 10-year shop Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: ShpRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .790a .625 .622 4.54073 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 4051.805 1 4051.805 196.516 .000b Residual 2432.948 118 20.618 Total 6484.753 119 a. Dependent Variable: ShpRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .987 .415 2.381 .019 Mkt-RF 1.213 .087 .790 14.018 .000 a. Dependent Variable: ShpRF Regression-3rd 10 year tech Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 Mkt-RFb . Enter a. Dependent Variable: TechRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .778a .606 .603 7.00535 a. Predictors: (Constant), Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 8908.616 1 8908.616 181.531 .000b Residual 5790.849 118 49.075 Total 14699.466 119 a. Dependent Variable: TechRF b. Predictors: (Constant), Mkt-RF Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .833 .640 1.301 .196 Mkt-RF 1.799 .134 .778 13.473 .000 a. Dependent Variable: TechRF Appendix 1B. Fama French Model Regression-Fama French model shop Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 HML, SMB, Mkt-RFb . Enter a. Dependent Variable: ShpRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .883a .780 .779 2.89846 a. Predictors: (Constant), HML, SMB, Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 10629.864 3 3543.288 421.765 .000b Residual 2990.791 356 8.401 Total 13620.655 359 a. Dependent Variable: ShpRF b. Predictors: (Constant), HML, SMB, Mkt-RF Table 1B4. Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -.173 .156 -1.109 .268 Mkt-RF 1.025 .036 .764 28.348 .000 SMB .850 .051 .440 16.589 .000 HML .422 .054 .219 7.827 .000 a. Dependent Variable: ShpRF Regression fama french tech Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 HML, SMB, Mkt-RFb . Enter a. Dependent Variable: TechRF b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .910a .828 .827 3.68732 a. Predictors: (Constant), HML, SMB, Mkt-RF ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 23356.533 3 7785.511 572.620 .000b Residual 4840.284 356 13.596 Total 28196.817 359 a. Dependent Variable: TechRF b. Predictors: (Constant), HML, SMB, Mkt-RF Table 1B 8. Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) .306 .199 1.539 .125 Mkt-RF 1.161 .046 .602 25.245 .000 SMB 1.257 .065 .452 19.279 .000 HML -.438 .069 -.158 -6.379 .000 a. Dependent Variable: TechRF Appendix 2: Factors affecting wage for American workers Descriptives Descriptive Statistics N Minimum Maximum Mean Std. Deviation WAGE 534 2 67 13.58 7.717 OBS. 534 1 534 267.50 154.297 EDUCATION 534 2 18 13.02 2.615 EXPERIENCE 534 0 55 17.82 12.380 GENDER 534 0 1 .46 .499 AGE 534 18 64 36.83 11.727 MARRIED 534 0 1 .66 .476 UNION 534 0 1 .18 .384 Valid N (listwise) 534 Regression-wages vs education Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 EDUCATIONb . Enter a. Dependent Variable: WAGE b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .381a .145 .143 7.143 a. Predictors: (Constant), EDUCATION ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 4602.214 1 4602.214 90.207 .000b Residual 27141.823 532 51.018 Total 31744.037 533 a. Dependent Variable: WAGE b. Predictors: (Constant), EDUCATION Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -1.046 1.571 -.666 .506 EDUCATION 1.124 .118 .381 9.498 .000 a. Dependent Variable: WAGE GGraph-Wages vs Education Regression-education and experience Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 EXPERIENCE, EDUCATIONb . Enter a. Dependent Variable: WAGE b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .447a .200 .197 6.916 a. Predictors: (Constant), EXPERIENCE, EDUCATION ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 6346.984 2 3173.492 66.351 .000b Residual 25397.053 531 47.829 Total 31744.037 533 a. Dependent Variable: WAGE b. Predictors: (Constant), EXPERIENCE, EDUCATION Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -7.224 1.833 -3.942 .000 EDUCATION 1.384 .122 .469 11.309 .000 EXPERIENCE .156 .026 .251 6.040 .000 a. Dependent Variable: WAGE Regression-education, gender and experience Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 GENDER, EDUCATION, EXPERIENCEb . Enter a. Dependent Variable: WAGE b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .502a .252 .248 6.692 a. Predictors: (Constant), GENDER, EDUCATION, EXPERIENCE ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 8008.738 3 2669.579 59.611 .000b Residual 23735.300 530 44.784 Total 31744.037 533 a. Dependent Variable: WAGE b. Predictors: (Constant), GENDER, EDUCATION, EXPERIENCE Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -6.104 1.783 -3.423 .001 EDUCATION 1.406 .118 .477 11.868 .000 EXPERIENCE .169 .025 .270 6.716 .000 GENDER -3.552 .583 -.230 -6.091 .000 a. Dependent Variable: WAGE Correlations Correlations WAGE EDUCATION EXPERIENCE GENDER AGE WAGE Pearson Correlation 1 .381** .085* -.208** .175** Sig. (2-tailed) .000 .049 .000 .000 N 534 534 534 534 534 EDUCATION Pearson Correlation .381** 1 -.353** .002 -.150** Sig. (2-tailed) .000 .000 .963 .001 N 534 534 534 534 534 EXPERIENCE Pearson Correlation .085* -.353** 1 .075 .978** Sig. (2-tailed) .049 .000 .082 .000 N 534 534 534 534 534 GENDER Pearson Correlation -.208** .002 .075 1 .079 Sig. (2-tailed) .000 .963 .082 .068 N 534 534 534 534 534 AGE Pearson Correlation .175** -.150** .978** .079 1 Sig. (2-tailed) .000 .001 .000 .068 N 534 534 534 534 534 MARRIED Pearson Correlation .104* -.036 .271** .011 .279** Sig. (2-tailed) .017 .413 .000 .796 .000 N 534 534 534 534 534 UNION Pearson Correlation .163** -.024 .118** -.157** .119** Sig. (2-tailed) .000 .582 .006 .000 .006 N 534 534 534 534 534 Correlations MARRIED UNION WAGE Pearson Correlation .104 .163** Sig. (2-tailed) .017 .000 N 534 534 EDUCATION Pearson Correlation -.036** -.024 Sig. (2-tailed) .413 .582 N 534 534 EXPERIENCE Pearson Correlation .271* .118** Sig. (2-tailed) .000 .006 N 534 534 GENDER Pearson Correlation .011** -.157 Sig. (2-tailed) .796 .000 N 534 534 AGE Pearson Correlation .279** .119** Sig. (2-tailed) .000 .006 N 534 534 MARRIED Pearson Correlation 1* .093 Sig. (2-tailed) .031 N 534 534 UNION Pearson Correlation .093** 1 Sig. (2-tailed) .031 N 534 534 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Regression Variables Entered/Removeda Model Variables Entered Variables Removed Method 1 UNION, EDUCATION, MARRIED, GENDER, AGE, EXPERIENCEb . Enter a. Dependent Variable: WAGE b. All requested variables entered. Model Summary Model R R Square Adjusted R Square Std. Error of the Estimate 1 .516a .266 .258 6.648 a. Predictors: (Constant), UNION, EDUCATION, MARRIED, GENDER, AGE, EXPERIENCE ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 8454.488 6 1409.081 31.885 .000b Residual 23289.549 527 44.193 Total 31744.037 533 a. Dependent Variable: WAGE b. Predictors: (Constant), UNION, EDUCATION, MARRIED, GENDER, AGE, EXPERIENCE Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -2.593 10.197 -.254 .799 EDUCATION 2.050 1.672 .695 1.226 .221 EXPERIENCE .813 1.673 1.305 .486 .627 GENDER -3.275 .588 -.212 -5.567 .000 AGE -.663 1.671 -1.007 -.396 .692 MARRIED .779 .632 .048 1.232 .219 UNION 2.169 .767 .108 2.828 .005 a. Dependent Variable: WAGE References Doing business in 2011. (2010). Washington, USA: World Bank. Ho, T. and Yi, S. (2004). The Oxford guide to financial modeling. Oxford: Oxford University Press. Lintner, J. (1965). The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets. The Review of Economics and Statistics, 47(1), p.13. Lohrmann, C. (n.d.). Comparison of the CAPM, the Fama-French Three Factor Model and Modifications. Martineau, M. (2000). Occam's razor. Gurnee, IL: Nightengale Press. Schmidt, M. (2008). Taking Shots At CAPM | Investopedia. [online] Investopedia. Available at: http://www.investopedia.com/articles/financial-theory/09/capm-error-problem.asp [Accessed 17 May 2016]. Sharpe, W. (1964). Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk. The Journal of Finance, 19(3), p.425. Synthesispartnership.com. (2016). Critical Issues 06 Financial Modeling. [online] Available at: http://www.synthesispartnership.com/critical06/ [Accessed 17 May 2016]. Read More