Title:
Can the effect of Covid-19 on firm performance be moderated by racial diversity?
Anna Grineva 11385200
Supervisor: Sibel Ozasir Kacar Date of submission: 25 june 2020
Table of contents
Abstract p 3 Introduction p 4 Literature review p 7 Market and Diversity side p 7 The relationship p 10 Methodology p 11 Data selection p 11 Data collection p 12 Conducted tests p 13 Results p 15 Testing assumptions p 15 Testing Hypothesis 1 p 19 Testing Hypothesis 2 p 21 Discussion p 25 Conclusion p 27 References p 28 Appendix p 31
Abstract:
The need for a more representative racial diversity composition amongst workforces is widely recognised. This paper evaluated whether racial diversity acts as a moderator for the relationship between the Covid-19 market shock and firm performance for a sample of US firms listed on the S&P 500. This study applied existing theories through multiple linear regression testing.
A significant relationship was established between the Covid-19 market shock and firm performance, measured in terms of share price growth. Racial diversity was not found to influence firm performance and there was no evidence of a moderating effect.
The recentness of the Covid-19 shock provides a possible explanation for the lack of
significant findings. There is the potential for future research that separates the shock into the initial market contraction, and the subsequent recovery period. Additional studies could also focus on the role played by the market sector of operation in predicting the effect of the market shock on a firm.
Introduction:
The global aspect of the modern economy means that competitors are not always
geographically close to one-another. As a result of this, firms will not always know what steps their competitors are taking to increase their competitiveness. In order to ensure that they are not overtaken, companies must always be searching and ready to implement strategies that will give them any slight edge over their competition.
A knock-on effect of the ever-increasing globalisation is that for many industries,
competition is so intense that even a percentage difference in performance can make or break firms. “Even small and medium-sized enterprises should keep abreast of the nature of
international events and how they impact on the business” (Goldman & Nieuwenhuizen, 2006, p25-26). The findings of diversity studies therefore have the potential to be of great significance to a variety of industries. The consensus between researchers is that higher levels of diversity in a workforce significantly increase the financial performance of firms across a variety of industries (Mckinsey Institute, 2015).
The equality act was implemented in the UK in 2010. This set a precedent that race, religious following, ethnicity and sexuality, amongst other factors are recognised as officially
protected characteristics (Hepple, 2010). Whilst this specific law relates only to the UK, it is evident that the issue of equality in the workplace is being prioritised on the agenda of governments. Research has demonstrated that the variety of “viewpoints, styles, and relationships” that result from diversity “maximise adaptability to changing environments” leading to increased financial performance (Linnehan & Konrad, 1999).
In their 2015 study, the Mckinsey Global institute evaluated the impact of diversity on firm performance, concluding that there exists a positive relationship between workforce diversity and firm performance (Mckinsey Institute, 2015). This study takes inspiration from these findings, however the Mckinsey institute measured diversity in terms of gender and race simultaneously. This study will exclusively analyse the racial aspect of diversity. Race is defined as “a social grouping of people who have similar physical or social characteristics that are generally considered by society as forming a distinct group” (Barnshaw, 2008) . In
line with this, diversity is defined as the number of races represented in a workforce and the proportion of the total workforce that belong to each present race.
In the modern globalised economy market conditions are constantly evolving. One major aspect of this is market shocks. Whilst market shocks affect most firms in the global market, the severity of impact varies drastically. As stated previously, a main priority for managers and industry analysts is determining what factors influence the impact of these shocks on firm performance. Presently, these groups rely on accounting ratios in order to judge a firm’s financial health and future potential (Kabera, 2009).
Many firms that perform well during normal market conditions find themselves not only suffering but in complete collapse when the market goes through periods of turbulence. Even with heavy government intervention, whilst rare, market shocks can last for months or even years. With the recent covid-19 global pandemic the unlikely outcome of an extended market shock is fast becoming a reality. After this crisis subsides, there will certainly be a vast quantity of new research into which firms were best suited to survive the crisis, with considerations for the factors that influenced performance during and following the shock.
Research gap and research question:
When considering the current extents of both diversity studies and research into firm performance, it is clear that current theoretical findings between these fields are moving towards the same conclusions from different directions; increasing the level of diversity in a workforce positively affects firm performance (Mckinsey Institute, 2015).
For research that focuses on firm-performance, there is a general understanding that the market goes through periods of growth and expansion. Rarely, there can also occur
unexpected market shocks. In just over a decade the world has experienced market crises that were both financial and, more recently, pandemic in nature. This variance in shock source reveals that firms cannot hope to effectively predict the sources of these market crises. They should better focus on insulating the firm from systematic risk on the whole.
This paper serves to offer a deeper connection between the two fields through the pursuit of an answer to the question: Does a diverse workforce act as a moderator for the impact of market shocks on the firm’s performance?
Responding to the question:
In order to evaluate the given hypothesis a quantitative research methodology will be employed using secondary data gathered from S&P 500, companies’ CSR reports and Yahoo!Finance.
Contributions to research:
The purpose of this research is to evaluate whether diversity in the workforce influences the relationship between market shocks and firm performance. Currently the fields of diversity studies and research into mitigating factors of market shocks on firm performance are separate.
The potential theoretical contributions of this research are firstly, the possibility to legitimise workforce diversity as a tool for protecting against market shocks. Secondly, taking a broader approach, these findings may disrupt the current non-financial based methods of firm
performance evaluation. If diversity is found to have a significant moderating role then this paves the way for future research into these non-financial indicators.
What will be discussed in sections:
In the literature review the existing literature will be presented and the relationships of the model will be further elaborated upon. The methodology section will discuss data collection and data analysis methods. In the results section the quantitative findings will be evaluated. The discussion will explore any weaknesses of the research and possible implications of findings, with respect to the existing literature and limitations of the study.
Literature review: Market and Diversity side:
In recent decades firms have become increasingly exposed to market shocks. A number of studies have attempted to explain this trend. One such study found that globalisation is a key factor that has increased the exposure of firms to market-wide shocks (Loecker & Goldberg, 2014).
Evidence from Slovenia indicates that a range of industry sectors are affected by market shocks; “for most industry sectors and financial indicators the effect of the crisis was evident/statistically significant” (Dolenc et al., 2012). An implication of this finding for this paper is that specific sectors need not be removed from the study outright. However, further findings of the research were that return on equity and employee value added were the
business aspects most affected by the economic crisis, as well as electricity, gas, water supply and construction industry sectors. Whilst entire sectors may not be removed from the dataset, control variables can be included for sectors of the economy to control for the potential inconsistent effect of the market shock between sectors.
Additionally, the authors noted that the significant relationship was consistent between multiple financial indicators. It follows logically from this that abnormal returns therefore, as a financial indicator, can be employed as the operationalisation of firm performance in this study.
Clarke et al. (2012) demonstrated that whilst managers of large firms reported demand contraction as their greatest concern, the greatest source of risk was diminished access to credit. The greater reliance of larger firms on access to credit as a component of their regular operations, in comparison to small firms explains how the loss of access to credit is a
significant risk. The uncertainty that arises in times of market shock limit firms’ ability to access this credit, as “financial constraints also become more binding. This was especially the case for larger firms” (Clarke et al., 2012). However, smaller firms find themselves in a worse position overall as they may never have had access to the required credit lines to continue operation during a crisis.
In order to maximise the validity of results, this paper should rely only on either small or large firms. The ease of access to large, publicly listed firm data dictates that for this paper, large firms data should be used.
In line with this, research confirmed that “firms with high financial flexibility suffer lower impact from the crisis” (Bancel & Mittoo, 2010). As demonstrated by Clarke et al., larger firms have greater access to credit, increasing their financial stability. When interpreting the findings of this research paper, it is necessary that any discovered relationship between the market shock and firm performance is considered as being more significant for small companies. This is due to the fact this paper will only include large firms’ data; the above studies consistently report greater effect of market shocks on smaller companies than large ones.
The market shock of interest for this research paper is the Covid-19 pandemic. The
uniqueness of this event means that generalising which sectors are most likely to be affected based on other recent shocks should be avoided where possible.
Despite the ongoing nature of the shock, there is evidence that some specific sectors were more significantly hit “especially those related to the service industries. Tourism and
hospitality are often among the first and hardest hit by turmoil” (Martin-Rios, 2020). This is considered with the inclusion of a control variable for the sector in the testing model.
Overall all firms will experience some degree of effect from the Covid-19 pandemic, it can be concluded therefore that any competitive advantage should be exploited where possible. Competitive advantage may come in the form of increased racial diversity. There is evidence that “members of racially or ethnically diverse groups hold different norms, assumptions, and preconceptions”. The researchers found that these aspects of racially diverse groups led to the consideration of “a wider range of perspectives”, allowing these groups to “generate more alternatives” as well as enhancing their “capacity to develop new competitive actions in competitive environments with high growth potential” (Andrevski et al., 2011, p.840). Similarly, a study in 2017 found a link between racial diversity and radical innovation in firms; “having greater workforce diversity ... is positively correlated to the share of a firm's turnover generated by radical innovation” (Mohammadi, Broström, & Franzoni, 2017).
These radical innovations enable firms to implement new competitive actions in response to the abnormal problems presented by market shocks.
As demonstrated previously, there is a significant increase in competition in times of market shock; it follows that the ability to realise high growth opportunities would be a significant advantage for firms experiencing market shocks. This implies that social diversity has the potential to moderate the relationship between the Covid-19 market shock and firm performance.
The Mckinsey institute’s ‘diversity matters’ paper indicated a significant relationship
between a more diverse leadership team and better financial performance. The companies in the top quartile for gender diversity were 15 percent more likely to have financial returns above their industry median, whilst companies in the top quartile of racial diversity were 35 percent more likely to have financial returns above their industry median (Mckinsey Institute, 2015).
The Mckinsey study encompassed both gender and racial diversity; their findings on gender diversity were independently confirmed by researchers evaluating Danish firms. The Danish researchers concluded that “the proportion of women in top management jobs tends to have positive effects on firm performance, even after controlling for numerous characteristics of the firm and direction of causality” (Smith, Smith, & Verner, 2006). This research paper will focus specifically on racial diversity, however the findings of the referenced studies remain highly applicable. The relationship between racial diversity and firm performance is in the same direction as other forms of diversity, the only potential difference is in the strength of this relationship.
The findings of a study into financial performance in the banking industry were also in support of the previous results that diversity did improve financial performance. “The results demonstrate that racial diversity does in fact add value and, within the proper context, contributes to firm competitive advantage” (Richard, 2000). Richard’s paper made a specific reference to market performance (the firm’s share price performance) as a measure of firm performance, amongst other measures. Richard further confirms previous conclusions that
racial diversity contributes to competitive advantage, which was found to influence firm performance during market shocks by studies referenced above.
The relationship:
This literature review has demonstrated, through evidence from existing research, the existence of a relationship between market shocks and firm performance, as well as a separate relationship between workforce racial diversity and firm performance.
This research paper aims to test whether workforce racial diversity acts as a moderator for the relationship between market shocks and firm performance.
A moderating relationship exists when the relationship between an independent variable (market shock) and dependent variable (firm performance) is conditional upon values of a separate independent variable, the moderator (workforce racial diversity).
Whilst a relationship can be present between the moderator and the dependent variable, separate from the moderation model, moderating figures do not include reference to this direct relationship and are instead always visually represented in the following way:
(Figure 1)
As visualised on the above figure, the relationship between market shock and firm performance is negative, meaning a market shock decreases firm performance.
The moderating effect, seen as the downward vertical arrow, is also negative. This is because workforce racial diversity is expected to decrease the severity of the market shock, meaning that the change to firm performance caused by a market shock should be lower when
Methodology: Data selection:
In order to evaluate the given hypothesis, a quantitative research methodology was used, employing secondary data on firm diversity gathered from the Corporate Social
Responsibility (CSR) reports of each company and data for corresponding firms’ financial performance from Yahoo!Finance. This research is interested in objective, rather than subjective results; there is an incentive for managers to misreport their opinions on firm performance. In juxtaposition to this, the stock market is not easily manipulated. Quantitative testing ensures consistency of reporting between subjects in the sample, allowing for
mathematical-certainty-based testing. Quantitative data is also easily standardized in order to reduce bias when collecting and analyzing the data. This ensures the objectivity and
generalisability in the findings.
In this study the relationship between the racial diversity of a company (diversity index) and its stock price performance during a market crisis was examined. The Simpson’s diversity index was used to measure the diversity of a company. The “Simpson's Diversity Index measures two critical elements of diversity, richness and evenness” (Prarolo et al., 2009, p7). It considers how many races are present within a company, as well as the quantity in each racial group.
The list of companies included on the S&P 500 was used to select firms. The S&P 500 “is largely considered an essential benchmark index for the U.S. stock market. Composed of 500 large-cap companies across a breadth of industry sectors, the index captures the pulse of the American corporate economy” (Beers, 2020). The ability to provide a clear representation of market performance as a whole made it a good baseline to judge firms’ relative performance. Yahoo!Finance was chosen as a source for share price data because “it provides financial data and a wide array of applications to help users obtain detailed and current market information” (Fuhrmann, 2020).
This study selected the recent and ongoing Covid-19 global pandemic as the market shock for the analysis. Whilst many qualitative aspects of company reporting have been impeded by
lockdown restrictions, this study only requires access to pre-existing firm composition data and stock price data, which has been updated throughout the pandemic despite quarantine. Consequently it can be concluded that the data gathering stage of this research was not noticeably affected by the Coronavirus pandemic.
Data collection: Diversity data:
Diversity data for firms included on the S&P 500 was sourced from the official company websites. Within S&P 500 companies there are reporting inconsistencies. For many firms workforce composition was broken down into simply ‘White’ and ‘Minorities’. This posed a significant problem for this study, due to requirements of the Simpson’s diversity index formula. For companies that provided data in this format, there was no alternative option but to exclude them from the analysis. To be included, firms must have provided data on all the following ethnic categories: white, asian, black/african american, hispanic/latino, other. As a result of this, 75 firms of the 500 remained that had provided sufficient data to allow for their inclusion in this study. S&P 500 companies are required to regularly produce a
significant quantity of data. To isolate the desired data firms’ 2018 and 2019 CSR reports were prioritised, with a primary focus on the diversity & inclusion sections.
There is a risk that the findings are less valid as this study only considers one form of diversity and doesn’t consider the fact that a firm has limited resources to allocate to improving diversity.
The Simpson’s diversity index takes absolute values per category, comparing this to the total number of employees. A difficulty that was faced during this data collection was that almost all firms gave their data in the form of percentages. In order to be consistent, when inputting the data into the Simpson’s index calculator the total employees value was given as 100, with the number of each minority group given as a numerical version of the percentage (e.g. 25% input as 25).
Data on ethnic diversity was the most readily available out of possible diversity indicators. Even then, only 75/500 firms from the S&P 500 provided sufficient data to allow for inclusion in this analysis. Branching out to firms outside of the S&P 500 would have
interfered with the validity of this study as individual firms’ performance was compared to the S&P 500 returns for the same period, allowing for clear analysis as to whether the firm performed better or worse than average.
Market shocks data:
This study uses the stock market as an indicator of firm performance. “The stock market measures the aggregate value of all publicly-traded companies.” This makes it “a valuable tool for gauging the health of the overall economy.” (Investopedia, 2020).
The stock prices over time form a better representation of firm performance than alternatives such as net profit, due to differing access to tax-mitigating strategies between firms, as well as other uncontrollable variances such as charitable donations. Data for this was gathered from Yahoo!Finance.
The time period selected for consideration in this event study was the 5 month period from the 1st of January 2020 to 31st of May 2020. During this period the covid-19 pandemic took hold around the world, leading to a market shock that increased in severity over time. The same time period a year prior was taken as a control seen as a non market shock period. This was chosen due to its relative proximity to the Covid-19 event, combined with the fact that market performance was generally stable throughout the period.
In order to obtain a single value for each firm’s performance per period, the daily stock prices were downloaded for both periods. The average daily change in stock price was then
calculated. The same was computed for the S&P 500 index. The S&P 500 index average daily return was subtracted from each firm’s average daily return for each year, providing two abnormal return values per firm. One for the control year, and one for the coronavirus market shock year. These were labelled as the ‘old excess return’ for the control year, and ‘crisis excess return’ for the crisis year.
Conducted tests:
To test hypothesis 1, a multiple linear regression was performed. Prior to running the regression it was necessary to remove outliers and test the linear regression assumptions (Olive, 2017). The multiple linear regression was separated into 3 models of increasing
complexity. All variables present in previous models are also included in subsequent ones, with the addition of an extra variable. This was done so that the relationship between predictor variables and the dependent variable could be confirmed for this sample, as predicted by the findings of the existing literature. All 3 stages of the regression testing included the same control variables.
The first model regressed the dummy variable ‘year (old/new)’ against the dependent variable ‘abnormal returns’. The outcome of model 1 therefore tested whether the coronavirus had a significant effect on firm performance. This was achieved through comparing the p-value of the variable ‘year (old/new)’ to the significance level. All tests were performed at the 5% significance level; a value lower than 0.05 indicates a significant relationship between variables. The second model added the variable ‘diversity_index’. The p-value of this
predictor variable demonstrated whether it could be considered to have influenced the firms’ performance, regardless of time period. The third model included the interaction variable ‘diversityXyear’, which combines the variables of ‘year (old/new)’ and ‘diversity_index’. The p-value of this variable indicated whether diversity acted as a moderator on the relationship between the coronavirus market shock and firm performance.
Testing hypothesis 2 followed the same procedure, however abnormal returns were
substituted for standard deviation. Standard deviation was included as a control variable in the regression testing hypothesis 1, this was removed for testing hypothesis 2 in order to prevent perfect multicollinearity.
Results: Testing assumptions:
A test for outliers, as well as the tests for linear regression assumptions were performed prior to running the regression.
Removing any outliers ensures that there are no significant deviations outside of the
interquartile range. The inclusion of outliers can impact the mean and median of results and skew results, affecting the accuracy and validity of findings.
(Table 1)
From the above table 1, comparing the values of ‘Mean’ to ‘5% Trimmed Mean’ it is clear that there is a significant difference when removing values outside of the 95% confidence interval; simulating the removal of any potential outliers. When the difference is significant,
as is the case for this sample, it means there are outliers in the data that should be removed before testing.
The above figure 2 visually demonstrates the presence of these outliers, denoted by the asterisk and circles. Data points 26 and 108 were significant outliers and so they were removed, along with the corresponding data points for the same firms in the different time period; 25 and 107 respectively.
Following the outlier removal the data was tested for independence of observations through the Durbin-Watson statistic.
(Table 2)
From the above table 2 the Durbin-Watson result can be observed for the total regression, seen as model 3. this. The observed result of 1.527 falls within the acceptable interval of 1.5 - 2.5 required for proceeding with linear regression testing. This corresponds to a slight but not significant positive autocorrelation; meaning the events of one time period have a slight effect upon the results of the next time period.
After this, homoscedasticity tests were performed to check that residuals are equally distributed. The below scatter diagram demonstrates an overall centering of data points, equally distributed above and below 0, confirming homoscedasticity.
(Figure 3)
(Table 3)
For the fourth step, a multicollinearity test was performed by running the complete regression and observing VIF values. From the above table 3 it can be observed that all VIF values, representing risks of multicollinearity, are below 10, apart from variables ‘year (old/new)’ and ‘diversityXyear’. This multicollinearity can be ignored because the p-value for the interaction variable ‘diversityXyear’ is not affected by the multicollinearity and therefore the
multicollinearity will not have an effect. In the regression, the market sector that the firms operate within were coded as dummy variables. In total there were 11 sectors, requiring the inclusion of 10 dummy variables. In the tables included in this results section, only the first 3 dummy variables are shown, however they are still present in the regression testing. The visual exclusion of these variables is in the interest of avoiding excessively large tables. None of the visually excluded variables had significant p-values, or VIF values above 10. A full table can be found in the appendix.
Lastly, the residuals (errors) were tested to ensure they are approximately normally distributed. For this purpose a P-P plot was used. It can be seen on figure 4 below that the residuals are equally distributed, following the trend line without significant deviation.
(Figure 4)
Testing variable relationships:
The moderating effect being tested in this paper relies upon two previously established relationships. These are the effect of a market shock on firm performance, and the effect of racial diversity in a workforce on firm performance. Based on these relationships taken from existing literature, the moderating effect is then tested, which can be seen below visually through the following diagram:
(Figure 1)
This model was used to test the following hypotheses:
Hypothesis 1:
H0: Workforce racial diversity does not have a moderating effect on the relationship between market shocks and firm performance.
H1: Workforce racial diversity acts as a positive moderator in the relationship between market shocks and firm performance.
Hypothesis 2:
H0: Workforce racial diversity does not influence the stability of firm performance.
H1: Firms with higher workforce diversity observe more stability in firm performance than firms with low workforce racial diversity.
Testing hypothesis 1:
A multiple linear regression was performed by taking the independent variables of market shock, in the form of a dummy variable, and a firm’s workforce diversity, given as a value from 0-1. These independent variables were regressed against the dependent variable of firm performance, given as the difference between the average daily change in share price of the firm, and the average daily change of the market index. The moderating effect was tested by creating an interaction variable between the variables of diversity and crisis year, given as ‘diversityXyear’. Testing the significance of this variable serves as a test for the moderating effect outlined in hypothesis 1.
The adjusted R-squared of the model, representing whether the included variables help to predict the dependent variable, was .144. This is considered to be quite low, however it is
enough to consider the model a weak representative of reality, meaning that conclusions can be tentatively drawn from findings. It should be noted that such a low R-squared value does indicate that there are omitted variables that would improve the accuracy of the findings; future research into this topic should therefore seek to implement more predictor variables.
Regression results:
(Table 4)
The above table 4 outlines the different models used to test the assumptions and hypotheses of this study. In each subsequent model different predictor variables are included in order to test whether they have an effect on the dependent variable of firm performance.
The models relate to the following assumptions/hypotheses:
Model 1: The effect of the Covid-19 market shock on firm performance.
Model 2: The effect of a firms’ workforce racial diversity on firm performance.
Model 3: The interaction effect of workforce racial diversity and Covid-19 market shock on firm performance, the moderating relationship being tested for hypothesis 1.
Predictor variable significance testing:
Model 1 introduces just the predictor variable ‘Year (old/new)’, this is a dummy variable representing whether the firm is in the normal year or the year in which the coronavirus occurred. For model 1 this variable is significant with a p-value of 0.005. This means that there is significant evidence that the coronavirus market shock had an impact on firm performance.
Model 2 brings the inclusion of the variable diversity_index. This is a value with a possible range from 0 to 1; for the given sample the lowest value was 0.30 and the highest was 0.75. The p-value of this variable was .497. This indicates that based on this sample firm
performance was not directly influenced by the firm’s degree of diversity. This is in contrast to the findings of existing research. This has implications for testing of hypothesis 1, as it has already been demonstrated through model 2 that one of the components of the interaction variable is not significant.
Model 3 allows for testing of the primary hypothesis that workforce racial diversity acts as a positive moderator in the relationship between market shocks and firm performance. This is achieved through testing the p-value of the interaction variable ‘diversityXyear’. This variable measures whether the diversity of a company combines with the market shock to impact the firms’ performance. The p-value of this interaction variable ‘diversityXyear’ was 0.803. This very high p-value indicates that rejecting the null hypothesis that workforce racial diversity does not have a moderating effect on the relationship between market shocks and firm performance, based on this sample, has an 80.3% chance of being incorrect. This provides sufficient evidence that for the given sample, at the 5% significance level, the null hypothesis that workforce racial diversity does not have a moderating effect on the
relationship between market shocks and firm performance cannot be rejected.
Testing hypothesis 2:
In addition to the main hypothesis of this research, it was determined that the volatility of firms throughout the periods would also be worthwhile evaluating. In order to do this the
following hypothesis was posed: Workforce diversity acts as a positive moderator in the relationship between market shocks and firm performance volatility.
The reason why this is an interesting hypothesis is that the average daily return only indicates whether a firm share price grew or shrank over the time period. The standard deviation shows the variation in daily share price change, providing a more accurate picture of the risk
involved in holding the stock.
In order to test hypothesis 2, the same predictor variables were used, with the exception of standard deviation, which was removed as a predictor variable and instead used as the dependent variable. Standard deviation refers to the standard deviation of average daily change in share price over the period. The interaction variable ‘diversityXyear’ was also still included.
The models relate to the following assumptions/hypotheses:
Model 1: The effect of the Covid-19 market shock on standard deviation of daily change in firm share price.
Model 2: The effect of a firms’ workforce racial diversity on standard deviation of daily change in firm share price.
Model 3: The interaction effect of workforce racial diversity and Covid-19 market shock on standard deviation of daily change in firm share price.
(Table 5)
The relationship tested in model 1 presents a significant result with a p-value of 0.000 for the variable ‘year (old/new)’. Therefore the coronavirus did impact the volatility in firms’ daily share price changes.
Model 2 tested whether racial diversity affected volatility in firms’ daily share price changes. The variable ‘diversity_index’ was not significant, with a p-value of 0.100. This means that diversity did not explain the volatility in firms’ daily share price changes.
Finally, Model 3 tests hypothesis 2, which states that workforce diversity acts as a positive moderator in the relationship between market shocks and firm performance volatility. The interaction variable ‘diversityXyear’ was not significant. The p-value of this interaction variable ‘diversityXyear’ was 0.516. The significance of the market shock dummy variable ‘year (old/new)’, demonstrating that the market shock did influence, combined with the insignificant p-value of the interaction variable ‘diversityXyear’ means that diversity
‘workforce racial diversity does not have a moderating effect on the relationship between market shocks and firm performance volatility’ cannot be rejected.
Possible reasons why it is not confirmed:
Several reasons can be given explaining why the hypotheses were not confirmed. First, the chosen time period only represents the initial period of the stock market crash in 2020. This assumes that firms are able to quickly implement solutions to overcome the shock. In reality this innovation could take a few months, explaining why the effect of diversity is not yet observed. This intuition suggests that the market shock could be split into 2 periods, the initial crash and the subsequent recovery.
Secondly, there are variables that presently cannot be controlled for in the model which have the potential to influence the results. The Covid-19 crisis is a unique situation, even amongst market shocks. Never before has the entire world shut down to this extent. The inability to interact face-to-face has created a need to switch to an entirely online-focused operation structure for many firms. This provides an uneven playing field for companies, as some are already predominantly online, whilst others cannot exist solely in an online capacity.
Covid-19 has seen the implementation of unprecedented restrictions on trade shipments. The virus’ origin in China, the production line of the world, meant that even before the virus had fully taken hold in countries, some businesses were experiencing shipment delays. This suggests that firms were not affected simultaneously or proportionately. These factors provide immeasurable influence on how Covid-19 has affected firm performance. It is clear from this analysis that a retrospective study once the pandemic is fully eliminated is optimal. Currently information on which sectors performed better is not available. A later study would allow for implementation of control variables that take into account which sectors were abnormally affected by Covid-19 and is the best approach to obtaining the most accurate explanation of variables’ effect.
Discussion:
The question that this paper aimed to answer was: Does a diverse workforce act as a moderator for the impact of market shocks on the firm’s performance?
The results of this study found that a diverse workforce does not act as a moderator for the impact of market shocks on firm performance.
There was significant evidence that the Covid-19 market shock had an effect on firm performance. However, the effect of racial diversity on firm performance, controlling for market shocks was not significant. The inclusion of the interaction variable measuring the moderating effect of racial diversity on the effect of the coronavirus on firm performance caused all independent variables to be insignificant. This demonstrates that racial diversity not only had no effect on firm performance, but it also did not influence the previously established relationship between Covid-19 and firm performance.
The implication of this finding is that firms should not rely on increasing their racial diversity solely for the purpose of improving their financial performance, be that for periods of
stability or market shock.
Although this paper does not present a significant relationship, it remains open to the possibility of testing different combinations of diversity types such as gender, age, tenure e.t.c. Previous studies considered diversity in terms of gender ratio or diversity in top
management teams only, while performance was calculated through return on assets (ROA), equity and firm size (Gompers & Kovvali, 2018; Rodríguez-Ruiz, Ó., 2016).
This provides another difference when compared to the approach of this study. For this research the average daily change in share price was used to represent firm performance. Previous studies referenced in the literature review used alternative operationalisations of firm performance, it is possible that this deviation from previous research is what caused a lack of significant findings.
Alternatively, it is possible that race alone gives less influence than when it is applied in combination with gender, age, tenure or other diversity type. In reality true diversity is more akin to a combination of "people with multiple backgrounds, mindsets and ways of thinking who work effectively together to perform to their highest potential in order to achieve
organizational objectives based on sound principles" (Pless & Maak, 2004, p. 130). A future study could analyse multiple types of diversity across different market shock types to explore whether there emerges a pattern between diversity type and response to different types of market shock.
There are a number of limitations faced by the current study that were experienced in similar previous studies. The main issue was the challenge of collecting diversity data where reports do not follow the same standards, or a total absence of diversity data when not required by law. These issues were confirmed by this research as indicated by the relatively small sample size. Data was collected for large US-based companies, thus, these results should be
interpreted with regard to other US firms. Future studies should consider companies with a more global scope and of different scale; this is a global shock and expanding the scope is a logical progression of this research.
Contributions:
The purpose of this research was to evaluate whether diversity in the workforce moderates the relationship between market shocks and firm performance, in the interests of legitimising workforce diversity as a tool for protecting against market shocks.
Despite the fact that this research paper did not find a moderating effect, or a positive effect of any kind on firm performance, it also did not have a negative effect. The absence of a significant relationship means this paper is unable to disrupt the current financial based approach to firm valuation as was intended. However, useful discussion is still possible regarding these results.
Evidence was provided in the introduction to this paper that governments are seeking to increase equality in the workplace. As a result of this managers are under growing pressure to increase the racial diversity of their workforces.
A contribution of the findings from this research is that managers that are increasing the racial diversity of their workforces due to this obligation, rather than our of choice, should not be concerned that this diversity could negatively affect their firm’s performance.
Conclusion:
Market shocks are recognised as abnormal periods or events that affect firm performance. The market shock of interest for this paper, the ongoing coronavirus pandemic, was found to support existing research, presenting a statistically significant impact on firm performance. This paper attempted to contribute to a better understanding of the complex relationship between diversity and firm performance, by analysing whether racial diversity acted as a moderator for this relationship between a market shock and firm performance.
Based on the sample of 73 US firms from the S&P 500 list, this research demonstrated that a higher degree of diversity does not have a significant effect on the average stock price of the company when considered independently. Building from this, it also did not act as a
significant moderator for the relationship between the market shock and firm performance.
The research was extended to see if the relationship between market shocks and the volatility of firm performance is affected by racial diversity. Similar to firm performance, volatility was found to be influenced by the Covid-19 market shock, however racial diversity did not influence volatility, nor was racial diversity a moderator for the effect of the Covid-19 shock on volatility of firm performance.
A possible reason for insignificant findings could be that more racially diverse firms
experience the same initial effects of the shock as less racially diverse firms; it is only when considering the recovery period that the benefits of racial diversity are observed.
At present these events have not fully occurred. Future research is therefore recommended once the long-term effects of the crisis have been realised, which may also implement the above strategy of splitting the shock period to compare the initial downward shock and the recovery period separately.
Overall, data for this research paper was gathered with an excess of caution, only including firm data that was acquired from high quality sources. Focusing on increasing the study’s validity by only including high quality sources of data presented a different threat to validity in the form of a low sample size. This quality to quantity trade off was unavoidable, however the decision to prioritise high quality data was more beneficial to the validity of conclusions than having a large but low quality sample would have been.
References:
Andrevski, G., Richard, O. C., Shaw, J. D., & Ferrier, W. J. (2011). Racial Diversity and Firm Performance: The Mediating Role of Competitive Intensity. Journal of Management, 40(3), 820–844. https://doi.org/10.1177/0149206311424318
Bancel, F., Mittoo, U. R. (2010). Financial Flexibility and the Impact of Global Financial Crisis: Evidence from France. SSRN Electronic Journal.
https://doi.org/10.2139/ssrn.1587302
Barnshaw, J. (2008). "Race". In Schaefer, Richard T. (ed.). Encyclopedia of Race, Ethnicity, and Society, Volume 1. SAGE Publications. pp. 1091–3. ISBN 978-1-45-226586-5.
Beers, B. (2020). Why do investors use the S&P 500 as a benchmark? Retrieved from https://www.investopedia.com/ask/answers/041315/what-are-pros-and-cons-using-sp-500-be nchmark.asp
Clarke, G. R. G., Cull, R., & Kisunko, G. (2012). External Finance and Firm Survival in the Aftermath of the Crisis: Evidence from Eastern Europe and Central Asia. Policy Research Working Papers. https://doi.org/10.1596/1813-9450-6050
Diversity Matters. (2015). Mckinsey Institute. Retrieved from
http://www.insurance.ca.gov/diversity/41-ISDGBD/GBDExternal/upload/McKinseyDivmatte rs-201501.pdf
Dolenc, P., Grum, A., & Laporsek, S. (2012). The effect of financial/economic crisis on firm performance in Slovenia-A micro level, difference-in-differences approach. Montenegrin journal of economics, 8(2), 207.
Fuhrmann, R. (2020). Yahoo! Finance vs. Google Finance: Which Is Better for Investors? https://www.investopedia.com/articles/investing/081314/yahoo-finance-vs-google-finance-w hich-should-you-use.asp.
Goldman, G., & Nieuwenhuizen, C. (2006). Strategy: sustaining competitive advantage in a globalised context. Cape Town: Juta.
Gompers, P., & Kovvali, S. (2018). The other diversity dividend. Harvard Business Review, 2018(JulyAugust).
Hepple, B. (2010). The new single equality act in Britain. The Equal Rights Review, 5, 11-24
Investopedia. (2020). How does the performance of the stock market affect individual
businesses?https://www.investopedia.com/ask/answers/042215/how-does-performance-stock
-market-affect-individual-businesses.asp.
Kabera, L. (2009) The use of accounting ratios in decision making. Retrieved from
https://www.memoireonline.com/12/09/3058/The-use-of-accounting-ratios-in-decision-makin g.html
Linnehan, F., & Konrad, A. (1999). Diluting Diversity: Implications for Intergroup Inequality in Organizations. Journal of Management Inquiry, 8(4), 399–414.
https://doi.org/10.1177/105649269984009
Loecker, J. D., & Goldberg, P. K. (2014). Firm Performance in a Global Market. Annual
Review of Economics, 6(1), 201–227. doi: 10.1146/annurev-economics-080113-104741
Martin-Rios, C. (2020, April 24). Unlocking Sustainable Strategic Renewal After COVID-19: Opportunities for Hospitality: By Carlos Martin-Rios. Retrieved from
https://www.hospitalitynet.org/opinion/4098309.html
Mohammadi, A., Broström, A., & Franzoni, C. (2017). Workforce Composition and Innovation: How Diversity in Employees’ Ethnic and Educational Backgrounds Facilitates
Firm-Level Innovativeness. Journal of Product Innovation Management, 34(4), 406–426. doi: 10.1111/jpim.12388
Olive, D. J. (2017). Multiple Linear Regression. Linear Regression, 17–83. https://doi.org/10.1007/978-3-319-55252-1_2
Pless, N., & Maak, T. (2004). Building an inclusive diversity culture: Principles, processes and practice. Journal of Business Ethics, 54, 129-147.
Prarolo, G., Bellini, E., Ottaviano, G. I., & Pinelli, D. (2009). Cultural Diversity and Economic Performance: Evidence from European Regions. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.1463199
Richard, O. C. (2000). Racial Diversity, Business Strategy, and Firm Performance: A Resource-Based View. Academy of Management Journal, 43(2), 164–177. doi: 10.5465/1556374
Rodríguez-Ruiz, Ó., Rodríguez-Duarte, A., & Gómez-Martínez, L. (2016). Does a balanced gender ratio improve performance? The case of Spanish banks (1999-2010). Personnel
Review, 45(1), 103–120. doi: 10.1108/pr-07-2014-0143
Smith, N., Smith, V., & Verner, M. (2006). Do women in top management affect firm
performance?A panel study of 2,500 Danish firms. International Journal of Productivity and
Appendix: Figure 1:
Table 1:
Table 2:
Table 3:
Table 4:
