The Boeing Company (BA) — NYSE Single Factor and Multi Factor Stock Analysis

7 min readFeb 21, 2022

Purpose of Report

Purpose of this report is to estimate single index and multi-index variables and build models for different number of variables for the Boeing Company stock traded in NYSE. Main goal was to see how the explanatory power of different models changed and to monitor the change in beta values. Lastly, to see the differences in the single index and two different multi-index models and to examine the effects of significant and non-significant variables on the models.

Methods of Analyzing the Problem

Yahoo Finance stock dataset and provided independent variable data has been used for the analysis. The investigated interval was between June-2016 and June-2021. There were 60 different observations since the stock prices monitored monthly. Risk analyzes are calculated on excel with ANOVA tables.

Findings and Conclusions

After analyzing the Boeing Company with single index model, it’s found that 35.11% of the total risk is coming from the systematic part and there is 64.89% unsystematic firm specific risk. Also, as different models are used an increase in the explanatory power is monitored from single index to F&F 3-factor model, but explanatory power decreased with new insignificant terms in F&F 5-factor model, so RMW and CMA are not good estimator variables for the company. Also, in any model beta value of market portfolio is greater than 1 which means the company overreacts any change in the market. Overall, with the effect of COVID-19, we can say that Boeing company stock is underperformed than the benchmark since alpha value is negative and significant.

Recommendations

· Effects of Covid-19 changed the way Airline companies work and as a result it affected aircraft manufacturer. Precautions for any further COVID-19 related industry risk might eliminate some part of this unsystematic risk, and from the investors side diversifying portfolio may eliminate that risk.

· Since alpha values are negative, benchmark and manager leading stock might be investigated and reconsidered.

· High market β values might possess greater risk in crisis scenarios, it should be decreased.

Boeing Company

The company was founded in 1916 in Washington, and headquarters are located in Chicago. The Boeing Company is a big player in the aircraft manufacturing industry, which produces both commercial jets and military aircrafts. Also, they work in the defense industry and have missile projects. In the Commercial Airplanes segment, it offers commercial jet aircraft as well as fleet support services for passenger and cargo needs. The defense segment focuses more on unmanned and manned fighter aircrafts and missiles. They compete with aviation, aerospace and defense industry, and main competitors are Airbus, Lockheed Martin, Bombardier, and General Electric. In the industry, competition is based on taking the shares of commercial planes in airlines. They have %38 share on the commercial airplanes industry against to Airbus according to Leeham News (Figure 1).

Single Factor Model

Results for single index are, adjusted r-squared value is 0.34. Our intercept value (α) is -9.71 and its p-value is 3.44E-07 (so close to zero), t-stat is -5.75. Our independent variable Mkt-RF has an intercept of 1.98 and p-value of 6.11E-07 with t-stat 5.60.

In the single index model, firstly we have an adjusted r-squared around 0.34, which explains how much of the change in our dependent variable is explained by the regression. Rm-Rf has a β coefficient of 1.98 which means if market goes up or down around 1% our security will move same direction 1.98%, and since p-value is almost zero (6.11163E-07) and absolute t-stat is greater than 1.96 it is significantly different than zero. Also, we have intercept coefficient of -9.71, which is alpha value, and it is surprisingly significant since t-stat is -5.75. That means our security underperformed 9.71% less than the benchmark index. (Exhibit 2)

ANOVA table can be used to decide on the risks (variance) and decompose them. Firstly, to find total risk, SS (sum of error squares) needs to be divided to df (degree of freedom). Total risk is 234.64. Then for the systematic part of the risk SSregression should be divided to total df to find its ratio, which is 82.38 and stands for 35.11% of the total risk. Unsystematic firm specific risk part is 1-systematic, but also it can be found from SSresidual/df which is 152.25 and stands for 64.89% of the total risk of Boeing. (Exhibit 2)

Fama-French 3 Factor Model

Results from Fama-French 3 factor model has adjusted r-squared 0.40. It means, explanatory power of our regression has increased with new added variables. Also, α has coefficient of -8.82 with t-stat value of -5.37, which means it is significant and results show that our security has underperformed than benchmark. Mkt-RF has coefficient of 1.64 with t-stat 4.38, it is also significant. SMB (small minus big) has a coefficient of 0.60 but since t-stat value is less than 1.96, it is not significantly different than zero and has no significant effect on the returns. Lastly, HML β coefficient is 0.90 and the t-stat value is 1.99 which is just above the 1.96 and it can be considered significant. That positive coefficient means the excess return is due to high book-to-market ratio of company and it is positively related with value firms. (Exhibit 3)

Fama-French 5 Factor Model

When the results from Fama-French 5 Factor model is checked, adjusted r-squared value has decreased to the 0.378, that means added new two independent variable has improved explanatory power less than expected. α still has coefficient of -8.82 with t-stat value of -5.26, which is significant and shows lower performance than benchmark. Mkt-RF has coefficient 1.58 with t-stat value of 3.64, means it is significant and company is positively correlated with market index. T-stat value of SMB is 0.89, so it is not significant and does not affect company’s return. Same valid for HML as well, not needed to be checked since t value is 1.56 (<1.96), means not significantly different than 0. New independent variables RMW and CMA are also insignificant since their t-stat values are 0.17 and -0.24 respectively. That also explains why adjusted r-squared value has decreased and why model’s explanatory power declined. Since new variables are insignificant, even though the r-squared increasing adjusted r-squared decreased. (Exhibit 4)

Explanatory Power

Explanatory power (adjusted r-squared) increased when we go from single index model to Fama-French 3 factor model, but then from 3-factor to 5-factor model it has decreased, and explanatory power decreased. It was not expected, because generally as you increase number of variables, adjusted r-squared increases and explanatory power increases. In our model, since the two new variables on 5-factor model is not significant the explanatory power of model has decreased.

On the other hand, beta coefficient of market portfolio has decreased as we go from single index to 3-factor Fama-French and also 5-factor Fama-French. This was expected because multi-factor model variables are more useful to explain movements in the security more than just movements in market index. So, in Fama-French models, other variables also explain some part of the variation and it makes market portfolio beta lower.

Observations about Validity and Usefulness

Since there is high level of errors in both single and multi-index models, validity of the tests are controversial. Yet, FF-style models consist of different variables to explain variance better, so we can say they are more useful and valid than single factor model. In single index model, since sometimes variation may be emerged from different unidentified risk factors results may be misleading. We see that 3-factor explains best for our security since RMW and CMA is not the source of variation.