Student’s Name

Institutional Affiliation

Instructor’s Name

Course

Date
Descriptive Statistics

The average return of the stock is 6.24% with a standard deviation of 1.35%, indicating

that the stocks have shown a moderate variability in performance. The stock returns vary

between 4% and 8.5%, indicating fluctuation but generally positive performance. In contrast, the

market return is lower in average value of 5.76% and with a smaller standard deviation of 0.73%,

indicating greater stability, with returns fluctuating between 4.5% and 7%.

The beta coefficient, showing the volatility of the stock with respect to the market, has an

average value of 1.2, indicating that the stock is slightly more volatile than the market. Values of

beta range from a low of 0.8 to a high of 1.7, indicating variation in the stock's risk profile across

the observations. Changes in oil prices show high volatility, with an average of 1.03% and a high

standard deviation of 1.42%. Changes in oil prices range from -1% to +3%, reflecting

considerable fluctuation in the oil market conditions during the period.

Inflation is relatively stable, ranging from 2% to 2.6%, with a mean of 2.24%. Likewise,

GDP growth is steady at a mean of 3.1% with a standard deviation of 0.26%, reflecting stable
economic growth. Interest rates are moderately volatile, with a mean of 2.13% and a standard

deviation of 0.43%, reflecting monetary policy in reasonably stable but changing conditions.

The exchange rate is very stable, fluctuating little (mean of 1.11 and standard deviation of

0.03), while unemployment is also relatively steady, between 3.5% and 4.2%, with an average of

3.75%. Finally, the industry dummy variable, which is binary, has a mean of 0.6, indicating that

60% of the observations belong to a particular industry.

Multi-linear regression
 The regression output shows the perfect fit of the model with an R-square value of

1.0000, indicating that independent variables explain 100% of variation in stock returns. This

perfect fit is quite uncommon and may signal potential overfitting, especially for the limited

sample size of 10 observations. F-statistic and p-value associated cannot be computed because

residual variance is zero, again hinting at problems with this set or specification of data.

The exponential coefficients define the relationships between stock returns and the

independent variables. For example, the market return is significant with an exp (Coef.) of 3.01,

indicating that the return on stocks would be increased by a factor of 3.01 for every additional

unit in the market return. So, beta has an exp Coef of 5.56, which means that for each additional

unit in beta, returns on stocks would increase by a factor of 5.56. Oil price change also indicates

a positive effect on stock returns with an exp(Coef.) of 2.00. Inflation has a significant positive

effect on stock return, as indicated by an exp(Coef.) of 10.63, showing that high inflation results

in a huge rise in stock returns. Conversely, the GDP growth, having an exp(Coef.) of 0.0035,

which is very small, suggests minimal changes in the stock return. The exchange rate coefficient

is very high, with an exp(Coef.) of 611.39, which implies an unrealistically large positive

relationship with stock returns and may need further investigation. Unemployment has a small

negative effect on stock returns, with an exp(Coef.) of 0.26, while the industry dummy variable

shows a slight positive effect with an exp(Coef.) of 1.17.

However, there are significant problems with the statistical significance of the model, as

the standard errors, t-statistics, and p-values are missing or undefined. This suggests that the

regression model was unable to calculate these values due to problems such as multicollinearity

or an inadequate number of observations. The constant term (_cons) is also abnormally high at
69.32, indicating that when all independent variables are zero, stock returns would theoretically

be 69.32, which is not realistic.

Diagnostics tests

There is a perfect multi-collinearity thus the diagnostic tests will not apply as they will

automatically reflects an error.

Scatterplot
The scatterplot above indicates that there are no outliers in the dataset.

The histogram above indicates that there are no outliers in the dataset.
CAPM and APT

CAPM
0.09
0.08
0.1
0.07
0.085
0.115
0.085
0.075
0.095
0.105

APT
0.558
0.464
0.75
0.28
0.51
0.9775
0.5015
0.3375
0.646
0.756

The two fundamental models in finance that explain the relationship between risk and

return of investment portfolios are the Capital Asset Pricing Model and the Arbitrage Pricing

Theory (Fabozzi, 2021). While both models are used to estimate expected returns on assets, they

differ in their underlying assumptions, methodology, and factors considered. The CAPM

considers that the expected return of an asset is linearly related to its risk, measurable by the beta

coefficient. Beta indicates the sensitivity of the asset to systematic risk or to market-wide

movements and not to idiosyncratic or asset-specific risks. Thus, mathematically, the CAPM is

presented as follows:
is the expected market return. CAPM is predicated on simplifying assumptions: investors hold

diversified portfolios, markets are efficient, and one can borrow and lend risk-free. In the CAPM

data included, returns vary from 0.07 to 0.115, with variation reflecting minor deviations in the

risk-adjusted expected return across different scenarios or time periods. These relatively stable

values provide evidence that the relationship between the beta of an asset and market risk

premium is predictable. Nevertheless, CAPM does not take into consideration multiple risk

factors that limit its explanation of the deviations in returns arising due to macroeconomic or

firm-specific variables. In contrast, APT expands the scope by incorporating various risk factors

beyond the market index. Proposed by Stephen Ross, APT assumes that asset returns are

influenced by various macroeconomic and firm-specific factors, which could include inflation,

interest rates, GDP growth, and other idiosyncratic risks. The general form of the model is:

, which represent the factors which constitute the risk.

A factor model assumes that when mispriced securities deviate from their intrinsic value using

the factors described above. Arbitrage will restore or force the market price toward this intrinsic

value. Compared to CAPM, which gave values from 0.28 to 0.9775, APT data shows a wider

scatter of values. This reflects APT's responsiveness to a wider array of influences and captures

variations in risk that CAPM overlooks. For example, higher APT values may indicate a stronger

sensitivity to a particular factor, such as interest rate fluctuations or sector-specific risks. While
both models are important tools for investors, each has its own limitations. CAPM's reliance on a

single risk factor (beta) makes it overly simplistic in explaining returns in complex markets.

On the other hand, APT requires identifying and quantifying multiple risk factors,

which can be challenging in practice. Despite these differences, both models provide critical

insights into portfolio management. CAPM is often favored for its simplicity and ease of use,

especially in benchmarking performance against the market. APT, being flexible, is preferred for

analyzing portfolios influenced by diverse macroeconomic conditions. The comparative data

highlights some key contrasts. The narrower range of returns from CAPM corresponds to its

single-factor assumption, while the wider variability of APT underscores its multifactor

approach. Which of these two models to use depends on the investor's goals, the data available,

and the complexity of the portfolio being analyzed. These models together offer complementary

insights into the complex relationship between risk and return.

References

Fabozzi, F. J., Fabozzi, F. A., & Pamela Peterson Drake. (2021). Introduction to Finance:

Financial Management and Investment Management. World Scientific.