AF7004 MSc International Finance Assignment Sample
Financial Econometrics and Forecasting I
On the capital markets, financial service providers are doing well. A high demand for loans has caused the performance to increase due to increased financial literacy among insurers. This trend is a result of the technological change. The stocks of these companies are more popular with investors than those of other sectors.
However, despite this, investors still need to diversify their investments. This process is called portfolio management. Due to the omission of certain variables like market size, momentum, and size in Sharpe’s CAPM model, two other models were introduced.
Carhart and Fama French are examples of these models. Basically, the goal of this report is to develop a stock portfolio by selecting several companies and then calculating the return on it. Also, we are also going to discuss the parameters and factors of the Fama French model and the dependent factors of the model as well as why they are critical.
Bloomberg.com is making use of 10 stocks from the United States in order to analyze them. These stocks include:
- S. Steel Corporation
- bioRad Lab Inc.
- The Qurate Retail Corporation.
- Lyondell Basel Industries, Inc.
- Nucor Corporation
- Metal Dynamics Inc.
- PLC Nielsen Holdings
- westlake Chemical Corporation
- Alaska Airlines, Inc.
- Xerox Holdings Corporation
Furthermore, we have examined the data for the last 10 years of each of these 10 listed US companies for the purpose of estimating the Price-to-Earnings ratio; Price-to-Book Value ratio and Dividend Yield ratio for each of the companies.
| P/E | |||||||||||
| Companies | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | |
| JPM US Equity
|
-41.91 |
-11.10 |
-3.10 |
32.40 |
-6.00 |
-4.90 |
51.10 |
3.70 |
4.10 |
-1.00 |
|
| BAC US Equity
|
7.14 | 17.80 | 11.52 | 11.43 | 8.80 | 4.47 | 2.40 | 4.44 | 0.00 | 7.81 | |
| GS US Equity
|
13.27 |
13.17 |
46.95 |
37.53 |
38.37 |
208.14 |
58.65 |
19.20 |
7.35 |
6.62 |
|
| MS US Equity
|
4.72 |
6.70 |
7.27 |
6.27 |
5.92 |
6.39 |
6.38 |
5.09 |
8.95 |
20.73 |
|
| WFC US Equity
|
12.01 | 23.05 | 24.90 | 18.01 | 134.54 | 22.94 | 12.88 | 7.41 | 13.86 | 23.14 | |
| C US Equity
|
8.38 | 15.04 | 20.38 | 22.84 | 0.01 | 21.22 | 13.58 | 4.16 | 10.61 | 13.94 | |
| PNC US Equity
|
100.38 | 32.91 | 18.88 | 36.46 | 25.27 | 25.80 | 26.85 | 0.00 | 0.00 | 0.00 | |
| BK US Equity
|
9.80 |
11.34 |
13.12 |
11.96 |
10.29 |
17.11 |
10.07 |
8.29 |
20.93 |
31.48 |
|
| STT US Equity
|
11.14 | 7.81 | 11.27 | 22.37 | 11.35 | 13.03 | 9.05 | 25.27 | 25.80 | 26.85 | |
| NTRS US Equity
|
5.39 | 5.94 | 17.03 | 19.19 | 14.47 | 9.40 | 34.67 | 13.60 | 7.76 | 27.42 | |
| P/BV | |||||||||||
| Companies | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | |
| JPM US Equity
|
1.87 |
0.91 |
2.42 |
0.98 |
0.43 |
2.42 |
3.14 |
1.85 |
1.48 |
1.08 |
|
| BAC US Equity
|
1.23 | 1.25 | 1.92 | 1.36 | 2.25 | 1.87 | 2.21 | 1.54 | 0.73 | 1.25 | |
| GS US Equity
|
1.59 |
1.54 |
2.63 |
2.60 |
3.66 |
1.99 |
2.49 |
1.52 |
2.15 |
1.94 |
|
| MS US Equity
|
1.20 |
2.63 |
2.70 |
3.12 |
6.33 |
1.54 |
2.63 |
2.60 |
3.66 |
1.54 |
|
| WFC US Equity
|
1.70 | 1.82 | 4.23 | 4.04 | 4.71 | 2.56 | 2.17 | 1.67 | 1.67 | 1.66 | |
| C US Equity
|
1.20 | 1.32 | 4.63 | 5.58 | 2.41 | 3.09 |
2.49 |
1.52 |
2.15 |
1.94 |
|
| PNC US Equity
|
29.24 | 30.16 | 35.89 | 15.29 | 27.34 | 41.95 |
1.54 |
2.63 |
2.60 |
3.66 |
|
| BK US Equity
|
20.38 |
19.83 |
29.15 |
21.73 |
25.66 |
57.66 |
105.51 |
66.17 |
69.0
1 |
79.86 |
|
| STT US Equity
|
18.37 | 11.55 | 46.69 | 29.76 | 10.51 | 88.73 |
1.54 |
2.63 |
2.60 |
3.66 |
|
| NTRS US Equity
|
20.17 | 17.97 | 12.06 | 56.52 | 18.01 | 23.00 | 29.15 | 19.76 | 36.8
7 |
23.19 | |
| Dividend Yield | |||||||||||
| Companies | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | |
| JPM US Equity
|
0.71 |
0.94 |
0.72 |
0.54 |
2.84 |
0.95 |
1.74 |
0.70 |
1.45 |
1.40 |
|
| BAC US Equity
|
0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | |
| GS US Equity
|
0.00 |
0.01 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
0.00 |
|
| MS US Equity
|
0.00 |
9.08 |
2.55 |
1.21 |
2.64 |
2.72 |
3.13 |
6.31 |
1.55 |
4.64 |
|
| WFC US Equity
|
3.62 | 3.38 | 2.92 | 2.93 | 3.65 | 2.45 | 2.35 | 2.92 | 2.83 | 3.01 | |
| C US Equity
|
3.03 | 2.95 | 2.21 | 2.41 | 3.07 | 1.54 | 1.46 | 2.53 | 2.80 | 2.71 | |
| PNC US Equity
|
0.00 | 0.00 | 0.00 | 2.27 | 2.28 | 2.82 | 3.67 | 5.48 | 5.43 | 1.48 | |
| BK US Equity
|
0.66 |
5.79 |
0.74 |
0.83 |
1.19 |
1.35 |
0.84 |
1.32 |
1.44 |
1.38 |
|
| STT US Equity
|
0.31 | 0.32 | 0.32 | 0.92 | 1.05 | 1.49 | 1.81 | 1.90 | 1.70 | 2.10 | |
| NTRS US Equity
|
2.03 | 2.34 | 1.96 | 1.64 | 2.58 | 3.40 | 3.32 | 5.01 | 2.68 | 4.29 | |
An EView using the Fama-French model and Carhart’s model:
Dependent Variable- RP Method: Least Squares
Date: 14/5/2022 Time: 1 pm Sample: 2012MDH1536
Included observations: 121
| Variable | Coefficient | Std. Error | t-Statistic | Prob. |
| C | -0.048325 | 0.006418 | -7.126787 | 0.0000 |
| MKT_RF | 1.008323 | 0.001676 | 597.7674 | 0.0000 |
| HML | 0.004833 | 0.002631 | 1.734293 | 0.0867 |
| SMB | 0.002678 | 0.006701 | 0. 676213 | 0.3670 |
| MOM | 0.002171 | 0.002144 | 1.012313 | 0.3135 |
| R Squared | 0.999757 | Mean Dependent variance | 1.103523 | |
| Adjusted R Squared | 0.996749 | S.D dependent variance | 4.135210 | |
| SE. of regression | 0.067526 | Akanke info criterion | -2.571523 | |
| Sum squared reside | 0.493767 | Schwartz criterion | -2.455818 | |
| Log likelihood | 159. 6778 | Hannah-Quinn criter. | -2.525296 | |
| F Statistic | 1183674.7 | Durbin-Watson stat | 0.112162 | |
| Prob (F-statistic) | 0.000000 | |||
Command Estimation:
=========================
LS Rp_ C MKT_RF hML SMB MOM
Equation Estimation:
=========================
RP_ = C(1) + C(2)*MKT_RF + C(3)*HML + C(4)*SMB + C(5)*MOM
Coefficients Substitution
=========================
RP_ = -0.048325 + 1.008323*MKT_RF + 0.004833*HML + 0.002678*SMB + 0.002171*MOM
Econometric issues for the Estimation Process:
In the above equation, it can be concluded that variables such as the Risk-free rate and the Headline Margin Level are significant in terms of their correlations, intercepts, as well as probability values as long as they fall below the 5% level. In the case of SMB, even though it has an intercept of below 5%, its probability is 35.10% which also not significant because it is over 5% and 10%. Additionally, the market return is not very significant from a probability standpoint because it is below 5%, but it is significant from a coefficient standpoint as it exceeds the 10% mark. As with the momentum factor, one can see that its coefficient is significantly little below 5% in this case. Therefore, the increase in momentum should not be taken seriously at this point. Although, it does not meet even the 10% significance level when measured in terms of probability, as it is at 31.35%. In this example, the independent variables do not seem to be significant. Furthermore, we can state from the data that model is appropriate and valid in that the Probability (F-Statistics) value is below the significance level of 5% and that this model is valid and appropriate.
Estimation Results:
Based on our Fama French & Carhart model, we have divided the data annually along with coefficients and significance levels as shown below:
| C | MKT_RF | HML | SMB | MOM | |
| 2011 | -0.11191 | -0.01652 | 0.08049 | 0.04580 | -0.07744 |
| 2012 | -0.02196 | -0.03529 | -0.05046 | 0.00804 | 0.07412 |
| 2013 | 0.04269 | 0.15205 | 0.05287 | 0.02045 | 0.07449 |
| 2014 | -0.14198 | -0.15214 | 0.01804 | 0.07803 | -0.07430 |
| 2015 | -0.02688 | -0.00999 | 0.02253 | 0.00815 | -0.07498 |
| 2016 | -0.00266 | -0.00399 | -0.00458 | -0.00802 | 0.00742 |
| 2017 | -0.02369 | -0.03899 | 0.02047 | 0.08005 | -0.00744 |
| 2018 | 0.19826 | -0.09699 | -0.18299 | -0.04391 | 0.18079 |
| 2020 | 0.16092 | 0.34309 | -0.10098 | -0.05766 | -0.05404 |
As can be seen from the table above, the coefficient fluctuates a lot on an annual basis in all the 10 years included in the study. Observing the market return, we can observe that the significance level is not reached for a period of four times, and for small and medium businesses it is reached merely once. It is evident from these calculations that both the SMB and the market return of the investment portfolio are significant. There is no significant momentum for 1 of the 10 times, more so for HML where it is not significant for 3 of the 10 times. Thus, it is evident that the significance level is maintained most of the time out of every ten years. Thus, it is evident that the significance level is maintained most of the time out of every ten years. The constant (alpha) displayed in the picture indicates a security that has a better performance than the market. F-F models are designed to incorporate independent variables such as Rm- Rf, SML, and HML. Based on our analysis of the following review, we can conclude that when an investor invests in small-sized companies, his/her portfolio return will increase, and when his/her book-to-market ratios are increasing, his/her portfolio return will decrease. Based on the F-F-model and the Carhart functions applied to the entire data, we can conclude that the alpha value for the stock is positive, meaning that it is performing better than the market. Hence, the betas of the curve determine the coefficients of the regression line. According to the subsamples, we can observe an increase of beta coefficients in the first year compared to later in the study. In other words, between 2009 and 2011, the beta coefficients have increased. Carhart’s data analysis shows that the values of the security are changing but the security is performing well, as we can see from the data derived from the model. Considering the results of the study, I believe that small size companies tend to operate more effectively than large companies. There are two reasons for this: first, small businesses can grow faster than large companies since they have a high growth rate. Furthermore, firms with a high ratio between book value and market value will be able to improve their profitability better than companies with low ratios. Normally, companies with a high book-to-market ratio will have a high investment ratio and a high Return on investment as a result of this.
Section B: Intervalling Effect
Calculation of Beta using CAPM in EViews:
|
Companies |
Beta Monthly | The daily beta |
Relevance |
Differences |
| Boeing inc | 6.72 | 1.21 | 0.00 | 2.50 |
| Chevron Group | 1.86 | 3.29 | 0.00 | 1.57 |
| Salesforce | 3.21 | 1.27 | 0.00 | 1.94 |
| Apple inc | 1.89 | 2.68 | 0.00 | 1.22 |
| Dow | 1.78 | 2.34 | 0.00 | -0.57 |
| Procter & Gamble | -1.86 | -3.16 | 0.00 | -0.79 |
| Intel Corporation | 5.70 | 2.27 | 0.00 | -0.52 |
| Visa | 3.41 | 2.14 | 0.00 | 1.24 |
| Boots Alliance Inc. | 1.60 | 2.12 | 0.00 | -0.53 |
| Goldman Sachs group Inc. | 2.88 | 1.97 | 0.00 | 2.94 |
| IBM Inc. | -3.16 | -2.60 | 0.00 | -0.52 |
| Johnson & Johnson | -2.84 | -2.06 | 0.00 | -0.74 |
| McDonald’s Corpo. | -2.29 | -1.87 | 0.00 | -1.43 |
| Merck & Co Inc. | -1.70 | -1.85 | 0.00 | 0.12 |
| Honeywell Inc. | 1.85 | 2.34 | 0.00 | 1.52 |
During this part of the analysis, it was noted that a typical volume for the Boeing Company was 9,759,187, while its market cap was 126.955 billion. As for the Beta, for the 5 years that the company has been in business, it is 1.51. This means that the ESG score risk of our company for the five years has been calculated at 36.1, which has the significance of a high risk for our company. The Visa Company, in contrast, has a stock volume of 575,482 and a market capitalization of 425.021B. Additionally, this company has a Beta value of 0.97 over the past five years. As opposed to Johnson & Johnson Company whose stock has a volume of 7,961,033 shares and a market cap of 450.833 billion dollars, the Johnson & Johnson Company has a stock volume of 7,364033 shares and a market cap of 42.833 billion dollars.
Econometric Issues
Autocorrelation – Regression models that perform well won’t show autocorrelation. We assume that residual terms are auto correlative, otherwise, residual variables are not auto correlative. We will assess the assumptions of both models by applying the Durbin- Watson statistic. Eviews determines the significance of the statistic.
- Autocorrelation:
I took the Durbin-Watson statistic Test to check the autocorrelation for daily and monthly frequency and these were the results
| Durbin-Watson statistics- Test for Autocorrelation-CAPM Model | |||
| Market cap size | Ticker name | Daily Frequency | Monthly Frequency |
| Large Cap | AAPL US Equity | 1.81782 | 2.346444 |
| FB US Equity | 1.977565 | 2.065962 | |
| JPM US Equity | 1.987926 | 1.643169 | |
| PG US Equity | 1.950438 | 1.576191 | |
| UNH US Equity | 2.043524 | 2.265119 | |
| Medium Cap | COP US Equity | 1.978905 | 2.134323 |
| CZR US Equity | 1.65796 | 2.287933 | |
| PXD US Equity | 2.052081 | 2.245921 | |
| IFF US Equity | 2.037543 | 2.207603 | |
| NRG US Equity | 1.886221 | 1.671999 | |
| Small Cap | MKTX US Equity | 1.986005 | 2.312385 |
| NLSN US Equity | 1.95357 | 2.155309 | |
| IPGP US Equity | 2.051118 | 2.38618 | |
| FFIV US Equity | 2.004866 | 1.733443 | |
| JNPR US Equity | 2.187305 | 2.219928 | |
The value 0 indicates the positive correlation and the value 4 indicates the negative correlation but The values of all the companies were lies between 1.5 to 2.5 in both daily and monthly frequency. Hence there are no signs of autocorrelation in these stocks
Heteroscedasticity- When the predicted variable in a standard deviation as compared to a value of an independent variable, or with respect to the standard deviations of a previous time period, are not constant, then this is a skewed distribution in the prediction
To test the heteroskedasticity I perform whites tests in all the 15 stocks both monthly and daily and these were the results.
| White test – Test for Heteroskedasticity-CAPM Modle | ||||||
| Market cap size | Ticker Name | Daily Frequency | Monthly Frequency | |||
| Obs*R-Squared | Prob Chi-square | Obs*R-Squared | Prob Chi-square | |||
| Large Cap | AAPL US Equity | 5.083131 | 0.079 | 0.287226 | 0.866 | |
| FB US Equity | 0.057809 | 0.972 | 3.710243 | 0.156 | ||
| JPM US Equity | 37.67357 | 0 | 4.713201 | 0.095 | ||
| PG US Equity | 78.50388 | 0 | 0.721154 | 0.697 | ||
| UNH US Equity | 78.41739 | 0 | 0.01012 | 0.995 | ||
| Medium Cap | COP US Equity | 105.6406 | 0 | 11.51283 | 0.003 | |
| CZR US Equity | 202.301 | 0 | 25.83419 | 0 | ||
| PXD US Equity | 105.7084 | 0 | 14.23327 | 0.001 | ||
| IFF US Equity | 3.270429 | 0.195 | 0.013264 | 0.993 | ||
| NRG US Equity | 4.06002 | 0.131 | 0.320205 | 0.852 | ||
| Small Cap | MKTX US Equity | 13.28614 | 0.001 | 4.335036 | 0.115 | |
| NLSN US Equity | 13.52086 | 0.001 | 0.141234 | 0.932 | ||
| IPGP US Equity | 6.423702 | 0.040 | 0.182082 | 0.913 | ||
| FFIV US Equity | 3.191065 | 0.203 | 3.847757 | 0.146 | ||
| JNPR US Equity | 2.9996864 | 0.224 | 0.771334 | 0.724 | ||
|
||||||
| CAPM Model -Beta estimation with S&P 500 | ||||||
| Market cap size | Ticker name | Beta Daily | P-Value | Monthly | P-Value | |
| Large Cap | AAPL US Equity | 0.070175 | 0.0478 | 1.488473 | 0.1027 | |
| FB US Equity | 0.024568 | 0.5914 | 0.159078 | 0.8401 | ||
| JPM US Equity | -0.00624 | 0.8536 | -0.09782 | 0.8738 | ||
| PG US Equity | 0.014637 | 0.6102 | 0.456843 | 0.3731 | ||
| UNH US Equity | 0.035928 | 0.3154 | 1.080463 | 0.1106 | ||
| Medium Cap | COP US Equity | -0.08215 | 0.1728 | -2.03813 | 0.0691 | |
| CZR US Equity | 0.059165 | 0.5564 | -0.74762 | 0.7421 | ||
| PXD US Equity | -0.07899 | 0.247 | -2.52777 | 0.0245 | ||
| IFF US Equity | -0.04941 | 0.2293 | -1.21754 | 0.2098 | ||
| NRG US Equity | 0.04143 | 0.4892 | 1.053643 | 0.4189 | ||
| Small Cap | MKTX US Equity | 0.096029 | 0.0539 | 2.293223 | 0.0566 | |
| NLSN US Equity | -0.11423 | 0.043 | -3.07833 | 0.0032 | ||
| IPGP US Equity | 0.019257 | 0.7694 | 0.073456 | 0.9566 | ||
| FFIV US Equity | 0.00377 | 0.9298 | -0.15139 | 0.86 | ||
| JNPR US Equity | -0.05866 | 0.16 | -1.02262 | 0.1829 | ||
|
||||||
As seen in the table the beta estimation of parameters was not significant in most areas hence deriving the direct relation between the index and stocks would be not meaningful. But there were 3 parts in daily beta value that show that negative relationship between the stock and index with a 10% significance level.
In the large-cap estimates, the average sum of daily beta is higher than the average sum of monthly betas indicating the frequency of the trading is higher when coming to the daily beta and the relation with index is also closer compared to the monthly data.
As same as the large-cap, the mid-cap stocks the average sum of daily beta is higher than the average sum of monthly betas but the values were positive here stating the direct relation between the index movement and the stock movement
In the small-cap, the average sum of beta value is daily higher than the monthly. And also, the difference is huge between the average sum of daily beta and monthly beta indicating the market moves more forward or vice versa in the monthly period of time not on daily basis
Multicorrilation- Assuming heterogeneity in independent variables is an assumption of linear regression. Endogenous variables will not result in linear regression. In equations in which the independent variable is defined, multicollinearity occurs. Multicorrelation will be tested using a variance factor test in our analysis
- The beta coefficient measures how volatile a particular stock is in relation to the market-wide systematic risk. Simply put, betas tend to tilt lines back to the right direction. All of the information in finance is aimed at making a profit on stock in comparison to others in the market. When it comes to volatility (Lin, 2018), the beta model accurately captures the profit margin movement. By changing the market interest rate over a predetermined period of time, we can determine the collateral beta by dividing the covariance effect of both collateral profit and market gain.
Assuming the stock has a beta of 1, it indicates strong correlation between its price and the market. Such stocks are accurate risks. A beta statistic, however, cannot measure random risk. Portfolio risk does not increase with the addition of stock to beta 1.0, but it does not increase its chances of providing additional benefits. In order to buy a protection contract with a beta less than 1.0, the investor has to assume that the protection is not as mental as the market. If we remember this portfolio stock, we are far safer than if we had the same portfolio without it. The beta of service stocks is typically low since they move freely than market capitalization stocks. Security costs are not substantially higher than the market because beta is more prominent than 1.0. Stocks that are new or small have a higher beta than market benchmarks. Accordingly, adding stocks to a portfolio will increase the risk of the portfolio, while increasing its normal return. However, there are some stocks with a negative beta. Stocks with negative beta equal negative market benchmarks. With complete representation of marking patterns, this stock can be considered counterintuitive. A counter ETF provides protection against bad beta. Negative beta is also common in several industrial circles, like gold mines.
Section C: Determinants of Capital Structure
In the following case, we will use the Pooled Regression Method according to the Redundant Fixed Effect to evaluate whether the Method is appropriate for the following data.
| Tests of Redundant Fixed Effects Equation: Untitled
Test fixed effects cross-section |
|||
| Test Effects | Statistic | d.f. | Prob. |
| Cross-section F | 42.327953 | (63,469) | 00.0000 |
| Cross-section Chi-square | 892.880204 | 62 | 00.0000 |
test equation of Cross-section fixed effects: Dependent Variable of LEV
Method: Least Squares of Panel Date: 15/5/22
Time: 10pm
Sample: 53545354
Periods included: 11
Cross-sections included: 65
Unbalanced
Total panel observations: 539
| Variable | Co- efficient | Standard. Error | T Statistic | Prob |
| C | -0.246352 | 0.081175 | -2.757558 | 0.0751 |
| Beta | -0.025295 | 0.017571 | -2.117544 | 0.0751 |
| Dividend | 0.052774 | 0.027568 | 1.375156 | 0.1590 |
| nwc | -0.184352 | 0.047566 | -3.927531 | 0.0003 |
| roa | -0.355209 | 0.077521 | -4.687568 | 0.0004 |
|
|
SIZE 0.035021 0.005437 6.440899 0.0000
| R Squared | 0.224658 | Mean dependent var | 0.224641 |
| Adjusted R Squared | 0.214615 | S.D. dependent var | 0.146460 |
| SE. of Regression | 0.124680 | Akaike info criterion | -1.211757 |
| Sum Squared reside | 8.974683 | Schwarz criterion | -1.183179 |
| Log likelihood | 336.7468 | Hannan-Quinn criter. | -1.211124 |
| F Statistic | 29.94683 | Durbin-Watson stat | 0.111703 |
| Prob F-statistic | 0.000000 |
With the Likelihood Ratio and the redundancy fixed effects, we see that the probability value is 0%, which is less than 5%. The pooled regression model is perfect according to the Null Hypothesis model. In addition, the fixed effect model is appropriate according to the alternative hypothesis. Thus, the Fixed Effect Model can be demonstrated to be a perfect model according to the Redundant Fixed Effect Model.
Regression Equation of Random Effects:
Using the EViews panel data, we ran random effect regression and obtained the following results:
Variable Dependent – LEV
Method: Cross-section random effects- Panel EGLS
Date: 14/5/22
Time: 2.00
Sample: 382473847
Periods included: 13
Cross-sections included: 35
Total unbalanced panel observations: 524
Arora and Swamy estimator of component variances
Variable Co- Efficient Stand. Error T Statistic Prob. C -0.335807 0.105665 -2.798634 0.3554
Beta -0.023539 0.010136 -2.796374 0.0055
Divodend -0.032356 0.016686 -2.165630 0. 5508
Nwc -0.135747 0.036685 -2.863090 0.0025
Roa -0.193565 0.042664 -4.519635 0.0020
Size 0.043580 0.006629 5.828635 0.0000
Specification Effects
S.D. Rho.
random Cross-section 0.113658 0.7942
|
|||||||||||||||||||||||||||||||||
Idiosyncratic random 0.063562 0.2354 Statistics Weighted
Test for Normality:
EViews can be used to calculate the normality histogram as follows:
The Histogram – Normality Test can be viewed under Residual Diagnostics: View > Residual Diagnostics > Histogram Normality Test
The panel data analysis has been conducted using three models.
- Regression models by pooled datasets
- Hausman Model or Fixed Effect Model
- Random Effect Model
Step 1 In order to obtain and maintain the appropriateness of the outcomes, we need to determine first whether the panel data should be analyzed with the Fixed Effects Model or the Random Effects Model.
Step2: Selection of the Panel Options > Under Period, select ‘Random’.
Step 3: Click on OK
Step 4: Correlated Random Effects (also known as the Hausman Test) can be found under View > Fixed/Random Effects Testing
In EViews, the Hausman test is carried out as follows:
You may load the panel data in EViews, select Equation Estimation from the drop-down menu, then Specifications, and invert each variable’s order of input from the first variable to the last variable.
Null Hypothesis of the Hausman Test:
It can be seen that this test is likely to be appropriate for the random effects model, considering its result of the null hypothesis test of Hausman Test.
Hypotheses alternative to the Hausman Test: In addition, the alternative hypothesis of Hausman test can also be applied to the Fixed effect model.
Result of Test:
In terms of the Hausman Test, we are able to see that the probability value is at 18.09% and this is substantially higher than the standard 5%. In other words, this indicates that the results are both acceptable and significant. When such a case occurs, it is obvious that the null hypothesis will be retained and that the Random Effect model is most appropriate. It can also be stated that the Random Effect Model has also been run after confirmation of the results obtained from the Hausman Test and its effects have also been taken into consideration.
Model of Random Effects:
Following are the steps that have been followed in order to run the Random Effect Model in EViews:
Step 1: In EViews, click on Panel Data, then click Equation Estimation, then click Specifications, then arrange the variables according to dependent first, then independent.
Step 2: In step 1, select the Panel Option, then Select the ‘Random’ under Cross-section.
Step 3: click on ok
Output Analysis:
The following table is derived from the Random Effects Model. The extracts from the data panel indicate that the independent variable is quite statistically significant, and that can be seen clearly from the data. Furthermore, based on this model, the probability value is below 5%
| Variable | Co- efficient | Standard. Error | TStatistic | Prob. |
| C | -0.300207 | 0.107465 | -2.765334 | 0.0474 |
| BE-TA | -0.028209 | 0.016536 | -2.796574 | 0.0473 |
| DIVI-DEND | -0.020016 | 0.016586 | -2.166540 | 0.4708 |
| Nwc | -0.120747 | 0.036585 | -2.816590 | 0.0475 |
| Roa | -0.192065 | 0.048654 | -4.519755 | 0.0000 |
| Size | 0.043200 | 0.007659 | 5.828655 | 0.0000 |
Also, we can find out from the F-statistics calculations that the probability of the Random Effects Model being significant is zero, which indicates that the model is quite significant in terms of accuracy. With respect to the capital structure of the company, all the independent variables have negative impacts. Therefore, the size alone of the company has a positive effect on its capital structure. In essence, this means that if Beta increases, dividend payments rise and the net working capital increases, and return on capital assets increases then the company capital structure will be negatively affected.
Conclusion
By using the EViews software, the financial data can be evaluated precisely. As well as that, EViews provided econometric tools. Additionally, it can be noted that there are some limitations in analyzing the panel data with EViews. A pivotal aspect of this analysis involves the selection of five large, five medium, and five small cap stocks. It is likely that the stock analysis and outcomes would differ if different stocks were chosen, as the model for selection was the Fama- French and Carhart Model. Second, we selected the stocks of the company using the same process. Furthermore, if the stocks chosen had been different, the beta would have been different. In addition, S&P500 was used to calculate beta in the US. The result of the beta analysis would have been different had we chosen a different stock exchange.
References
Anon. 2022. [online] Available at: <https://www.investopedia.com/ask/answers/010915/what-considered-good-price-book- ratio.asp> [Accessed 14 May 2022].
Equitymaster.com. 2022. [online] Available at: <https://www.equitymaster.com/glossary/price-book-value-ratio/> [Accessed 14 May 2022].
Investopedia. 2022. Dividend Yield; Formula and Calculation. [online] Available at: <https://www.investopedia.com/terms/d/dividendyield.asp> [Accessed 14 May 2022].
Investopedia. 2022. Price-to-Earnings (P/E) Ratio. [online] Available at: <https://www.investopedia.com/terms/p/price-earningsratio.asp> [Accessed 14 May 2022].
Investopedia. 2022. What Is the Fama and French Three Factor Model?. [online] Available at: <https://www.investopedia.com/terms/f/famaandfrenchthreefactormodel.asp> [Accessed 14 May 2022].
