1
Faculty of Economics and Business Bachelor year: 3
Amsterdam Business School Course: 2015/2016
Finance Group Semester: 2
Plantage Muidergracht 12 1018 TV Amsterdam
Thesis Seminar Finance and Organization (6013B0326)
The effect of oil prices on the value of firms. Evidence based
on Dutch listed firms.
Cas Reparon 10552480
Coordinator:
Dr. Philippe Versijp (Finance)
Supervisor:
2
Statement of Originality
This document is written by Student Cas Reparon who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Abstract
This thesis is an original study and opens a new debate on the potential effect of recent oil price changes on the firm value of Dutch listed firms. Using a sample of 8 firms, all participating in different sectors in 2012-2016, we compute the economic importance of oil price changes on the firm’s Tobin’s q and relate the significant oil price parameter to existing economic theory. Given the economic importance of oil prices, these results add important financial interest for policymakers, business leaders, researchers and many more in order to consider oil prices in both financial and cost management.
4
1. Introduction
What is the relationship between changing oil prices and firm returns? What is the effect of oil price uncertainty on firm-level investments? What is the impact of changes in oil price volatility on stock returns?
These research questions are commonly discussed in the finance theory, starting with Naravan and Sharma (2011) and continuing with Yoon and Ratti (2011) and Sadorsky (1999). Yet the effect of oil prices on the actual value of firms is still unexplored. Prior literature focuses primarily on the effect of macro-economic variables such as the inflation rate, the exchange rate and the gross domestic product. In contrast, there has been done relatively little research at the micro-economic level and more specifically at the firm level.
The oil price reached record highs in 2008 due to increased oil demands in Asia paired with decreased production in the Middle East, set by the Organization of Petroleum Exporting Countries (OPEC). Shortly after, the global recession caused oil prices to drop again but due to economic recovery the oil prices stabilized around $120 a barrel until 2014. At the beginning of 2014 there was a steep drop in oil prices, due to numerous factors, such as the decreased demand from China, India, Brazil and Russia and the non-cooperating actions from Saudi Arabia (Baffes et al., 2015). Since the outcome of a binding referendum in the United Kingdom resulted in a victory for Leave, as a result oil prices are one again decreasing, the effect of oil prices on the value of firms is stillan area of ambiguous worth investigating.
Jones and Kaul (1996) incorporated the effect of oil prices into the stock market. Given the sensitivity of stock prices, there is a negative relation between oil prices and the stock market returns. Given that the stock market shocks are fully based on the actual cash flows, the stock market becomes riskier and more uncertain when oil prices fluctuate. This affects investment decisions. On the other hand, Sadorsky (1999) showed that changes in oil price and oil price volatility have a significant negative impact only on stock returns. Both the stock prices and returns are important for calculating the firm value in the model.
Aye et al. (2014) investigated the changing oil prices and the resulting uncertainty on large manufacturing companies of South Africa. Based on the assumption that investments are influenced by estimated returns, the oil price volatility has a negative impact on the investments because uncertainty affects the potential returns. The effect on investments, including research and development and capital expenditures, is therefore important because these investments contribute to a long-term sustainability.
5 The effect of changing oil prices on firm values is different by sector of activity. Huang and Masulis (1996) stated that the effect of oil prices depends on whether the industry is a producer or consumer of oil. Mohanty et al. (2014) supported this research by providing that an increasing oil price is positively related if the firm uses oil as an input, consumer role, but negatively related if the firm uses oil as an output, producer role. In addition to this, a study from Elvasiani et al. (2011) concluded that oil prices always have a significant negative effect in the travel and leisure sector. This study is based on the cost pressures derived from increasing oil prices. Literature shows that the impact of oil prices differs across sectors, but which sectors are significantly affected is still ambiguous.
Based on all of the previous studies, the uncertain risk of changing oil prices is sector specific and affects investments, stock returns, cash flows and therefore the firm value. This thesis continues and completes the current studies. Based on all of these studies, the goal is to study the effect of changing oil prices on sector specific firm values. Naturally, the follow-up research question is: what is the effect of oil prices on the value of firms? Evidence from Dutch listed firms should provide the guideline for further research and goes further than previous literature.
Section 2 presents the empirical approach, summarizes the model selection and exemplifies the regression variables. Section 3 describes the data and time set, the empirical results are presented in section 4. Section 5 concludes the thesis and finally section 6 and 7 details the references and appendix.
6
2. Empirical approach
This section describes the economic meaning of the variables used in the model. Specifically, the regression analysis will be explained as well as related hypotheses.
2.1 Constructing a firm value benchmark
The value of a firm is the present value of the cash flows generated by the firm’s assets. This consists of both growth opportunities and assets in place (Habib and Ljungqvist, 2003). There are multiple possibilities to calculate an estimation of the firm value. The first one is the market capitalization method. This method is relatively easy to obtain, because both the stock prices and quantities are easily available. However, this estimator only reflects the simple results in the quality of the regression and is therefore unsuitable. The second one is the enterprise value. Regardless the fact that this estimator could be reliable, it is impossible to make a regression without making some adjustments to the published data, these include often subjective estimations of values. The last one is the Tobin’s q, which is a ratio between the market value of the firm and its replacement costs of the firm’s assets. This last estimator will be used because the regression determinants of Tobin’s q have been modelled extensively in prior studies and the empirical results are the foundation of our research analysis.
The Tobin’s q is the ratio between the total market value of a firm divided by the total asset value (Berk and DeMarzo, 2014). If the effect of changing on the firm value is significant in a certain sector, that does not necessarily mean it will be in a different sector as well, therefore the Tobin’s q should be stochastic. This will allow errors in the estimation and to prevent outliers that drives the regression analysis outcome too much.
2.2 Hypothesis development
In order to analyse the potential impact of oil prices on the firm value it is evident to control for differences among the various firms and sectors. The effect of changing oil prices may be well observed in some sectors, but less so in others.
Dayanandan and Donker (2011) predicted a negative effect of oil price effects on firms working in the industry sector, however research on other firms is ambiguous. The choice is motivated by the fact that firms from this industry sector are assumed to use relatively more oil as an input. H1: Changes in the oil prices affects the firm’s value.
7 Research from Huang et al. (1996) showed the empirical results that the changing oil price is affected by the sector in which the firm operates. They claimed that the relationship between the firm and sector size plays an important role in estimating the oil price effect.
H2: Changes in the oil prices affects every sector differently.
2.3 Empirical model
2.3.1 Model selection
To get an answer to the research question “What is the effect of oil prices on the value of firms?” the model should be structured explicitly. The determinants of the Tobin’s q have been modelled extensively and prior research from Habib and Ljungqvist (2003) and Zaabouti et al. (2016) based their results on a stochastic frontier model. However, this paper only considers the location of the frontier, with some minor changes, and ignores the shortfalls from the frontier, because the shortfalls focuses primarily on errors around the frontier. Since we already account for random influences on variables in the error term the shortfalls are left out. The model estimated in this paper is based on the following formula: the following form:
Qit = β0 + β1 OILPRICEit + β2 ln(SALESit) + β3 ln(SALESit) 2 + β 4 "&$ % + β5 &$' % + β6 (&)*+ % + β7 % ,&-*, + β8 ( % ,&-*,)0 + β9 1 ,&-*, + β10 FFRit + Eit
2.3.2 Descriptive sample statistics
A summary of the variable definitions is listed in Table 1, furthermore the appendix provides additional information.
Variable Definition
Tobin’s q Value of the firm. Ratio between the market value of the firm and the book value of assets
Oil price Price of a barrel oil from the Crude Oil WTI
Sales Quarterly revenues at the end of period t
R&D /K Ratio of research and development expenses to the fixed assets
ADV/ K Ratio of advertising expenses to the fixed assets
CAPEX / K Ratio of capital expenditures to the fixed assets
K / Sales Ratio of tangible long-term assets to revenues
Y / Sales Ratio of operating income to revenues
FFR Fama-French estimate of the risk premium
8
2.3.3 Economic meaning the variable set
Here we will focus on the basic economic theory behind the set of variables. The precise statistical definitions are shown in the data appendix.
The oil price is the main explanatory variable of the regression model and prior research concluded a negative relation between the changes in the oil price and the stock prices, cash flows and investments.
For simplicity reasons we assume the higher the revenues, the higher the firm size and consequently the firm value and therefore take into account the diminishing returns. The log sales captures this implied inverse relation, but to capture potential nonlinearities we also include log sales squared. Another measure of profitability is the operating margin, or the division of the operating income (Y) and revenues.
The proxy for growth opportunities is divided in both soft spending and hard spending. Both research and development costs (R&D) and advertising costs (ADV) are part of the soft spending category, but they also proxy for intangible assets and therefore control for upward bias. The capital expenditures (CAPEX) is part of the hard spending category and is expected to correlate positively with the Tobin’s q.
According to Himmelberg et al. (1999) the division of the fixed assets (K) and revenues and its square, control for the importance of tangible capital in the firm. Using the Tobin’s q as a measure of the firm value might cause an understatement of the replacement costs of intangibles. Derived from this is a possible negative relation between the Tobin’s q and the firm’s tangible capital intensity, or the mentioned division, and is therefore an important variable to take into account.
Given that the Tobin’s q numerator is the value of the firm, and is consisting of the ratio between the market value the firm and the total assets of the firm, future cash flows are discounted. These cash flows are discounted at the firm’s costs of capital. The industry risk premium estimated by Fama and French (1997) measures this.
However, not all variables from prior research have been taken into account. This paper disregards the use of leverage, because of the data unavailability of the industry specific tax shields. For simplicity reasons we therefore assume that in a Modigliani-Miller world leverage should not impact the firm value. Also, the net effect of regulations, the reasoning that regulations may constrain a firm’s ability to create value, is ambiguous. Assuming all sectors in the Netherlands are treated the same, we ignore this variable.
9
3. Data
This section gives insight in the dataset used in order to answer the hypotheses. Both the sector specific characteristics and the used timespan are discussed.
3.1 Data and sources
The sample consists of 8 Dutch listed firms for the period between 2012 and 2016. The dataset is derived from the financial quarterly reports via the website of each firm and the stock prices from the yahoo finance historical price chart. The data on oil prices is from the NASDAQ price chart on West Texas Intermediate (WTI) crude oil prices. Literature shows that the impact of oil prices differs across sectors. In order to test the hypotheses and account for the sector specific differences across the firms, we must divide the Dutch economy in to 12 sectors.
The electronics company Philips represents the consumer goods sector and Royal Dutch Shell represents the oil and gas sector. We used the historical prices of PHG and RDSA.AS respectively to determine the stock prices and to calculate the market value. All the required data was available in the quarterly results, but for the following firms the data was adjusted in the following way. The results on the chemical firm AkzoNobel estimates the basic materials sector. Unfortunately, the research and development datasets from the first quarter of 2014 until the first quarter of 2016 were not reported separately from the selling, general and administrative expenses. Thus, we used a weighted average based on the data available in 2013 and 2014 instead and estimated the research and development costs accordingly. The AKZA.AS stocks are used to determine the market value. The company DSM represents the healthcare sector. However, the quarterly reports did not indicate the research and development expenses at all, so we used their “innovation center” expenses as an estimate for this. Further, the data on the operating income and thus margin was not reported, instead we used the earnings before interest and taxes (EBIT) as a reliable estimator. The historical stock prices are obtained from the RDSMY. The industrials sector is based on the technology firm ASML. Here, we also used EBIT as an estimator of the operating income. Another minor adjustment was the use of the “purchase of power, plant and equipment” expense to estimate the capital expenditures, because it is equal to the company’s spending on physical assets. The ASML historical prices are used. TomTom represents the technology sector. The purchase of power, plant and equipment is also used here to estimate the capital expenditures for TomTom. One minor adjustment is the use of the operating result to calculate the operating income and margin. The
10 TOM2.AS displays the stock prices. Unfortunately, the research and development data for both Ahold, an international retailer active in the retail sector, and Air France-KLM, a flag carrier airline working in the travel and leisure sector, are not available through the quarterly reports or press releases. Also, there was no supporting literature available to presume a fixed percentage in the expenses. The historical stock prices are obtained from the AF.AS. The capital expenditures of Air France-KLM are estimated by a combination of the purchase of property, plant and equipment and intangible assets. For Ahold the purchase of non-current assets is used to estimate the capital expenditures. The AH.AS is used to determine the market value.
Unfortunately, not all sectors could be used in this research. Financial institutions, i.e. Rabobank, part of the financial sector, report their financial statements in a different way. The purpose is to give firms insight into the bank’s solvability, capital and securities, but this results in an unavailable dataset for this current model. The Dutch utilities sector with listed companies like Nuon was also not suitable. Nuon merged with the Swedish company Vattenfall in 2009 and combines its quarterly results with Vattenfall. Eneco was also unavailable for study since it only reports on annual basis. Both Ziggo, active in the media sector, and KPN, a company in the telecommunication sector, report their assets, liabilities and expenses differently. Also, because Ziggo is deemed a subsidiary of the parent company Virgin the reports on outstanding shares, required to calculate the market value and consequently the Tobin’s q, are not reported disjointed. It is therefore necessary to leave these four sectors out of the regression analysis.
According to Kokemuller (2015) businesses typically allocate between 2 to 12 percent of the total sales towards advertising. Scientific research from Garber (1995) verifies this by auditing the ratio between advertising costs and the average revenues. Advertising costs are normally not reported separately on quarterly financial statements, so for the sake of simplicity, the model assumes that 10 percent of the quarterly sales are allocated towards advertising.
3.2 Data timespan
Oil prices dropped radically during the year 2014 and continued to do so afterwards. In perspective this meant a decrease of $100 per barrel in 2012 to $40 per barrel in 2016. In order to investigate this negative change in oil prices on the value of firms the timespan ranges from two years before this drop to two years after. The regression analysis is performed on quarterly results starting at the first quarter of 2012 up to and including the first quarter of 2016. The analysis covers 17 observations per firm.
11
4. Empirical results
The discussion of the empirical results is structured as follows. First we estimate the variables and showcase them in Table 2. We then review the estimated variable of interest, the oil price. Finally, we summarize and discuss the results for any improvements.
4.1 Estimation results
Variables AkzoNobel Philips DSM ASML Shell Ahold TomTom KLM
Constant -253,165 (1361,318) 548,104 (225,792) 111,739 (103,501) 576,672 (54,522) 54,542 (199,457) 177,184 (170,313) -102,863 (181,754) -8,367 (44,740) Oil price -0,008 (0,0044)** 0,009 (0,003)*** 0,001 (0,001)* 0,016 (0,004)*** -0,001 (0,008) -0,006 (0,002)** * -0,007 (0,009) 0,001 (0,0002)*** Ln(SALES) 98,688 (288,306) -122,400 (54,420)** -29,607 (26,992) -159,496 (14,700)*** -8,309 (33,958) -37,946 (29,807)* 33,648 (61,639) 3,251 (9,575) Ln(SALES)2 -6,180 (17,629) 6,958 (3,126)** 1,920 (1,755) 11,542 (1,025)*** 0,338 (1,483) 2,009 (1,625) -2,903 (5,683) -0,226 (0,551) R&D/K 2,987 (3,889) 0,346 (1,093) 6,883 (7,128) -12,704 (4,904)*** -58,776 (121,518) 0,918 (0,571)** ADV/K -538,622 (1253,131) -12,687 (33,800) 7,952 (13,788) -94,666 (12,850)*** -22,471 (44,595) 6,458 (100,813) 0,312 (10,451) -12,680 (37,632) CAPEX/K -2,607 (1,759)* -2,565 (2,124) -0,073 (1,224) 11,078 (2,948)** 3,804 (5,573) -4,372 (2,421)** 1,741 (1,311)* -0,201 (0,444) K/SALES -116,173 (347,027) -19,928 (58,163) 2,046 (1,369)* -10,638 (2,073)*** -1,656 (2,917) 7,057 (63,432) 61,927 (194,152) -1,676 (3,483) (K/SALES)2 31,053 (111,190) 10,109 (45,043) -0,523 (0,379)* 1,100 (0,280)*** 0,191 (0,309) -5,746 (29,823) -161,666 (449,171) 0,296 (0,637) Y/SALES 0,162 (1,565) 0,528 (1,461) -0,372 (0,395) -19,950 (3,771)*** 2,001 (2,377) 3,554 (6,221) 1,676 (2,957) 0,374 (0,237)* FFR -0,530 (1,006) -0,003 (0,891) 0,187 (0,141)* -1,597 (1,401) 0,182 (0,338) -0,016 (0,512) 1,027 (0,703)* -0,078 (0,110) R2 0,7912 0,8071 0,7022 0,9689 0,2505 0,8699 0,9067 0,8964 t -1,77** 3,63*** 1,50* 4,45*** -0,15 -2,49*** -0,78 3,69***
Table 2. Estimation results. *** denotes significance at 1 percent, ** at 5 percent, and * at 10 percent. Robust standard errors are calculated.
12
4.2 Estimating the oil price effect
The results obtained from the regression analysis in Table 2 highlights the sector specific effects. If firms are independent of oil price changes the two-sided t score, listed at the bottom of the table, will not be significant. The estimation results in Table 2 reports a predominantly significant effect of this null hypothesis, which we comfortably reject in most of the sectors (t>1,28). Thus, in our sample, most firms and sectors are dependent on oil price changes. The most important finding is that the changes in oil price are significant in almost every sector.
The estimated results in the consumer goods sector by Philips have a great effect (t=3,63). This result is significant at the 1% level. The sign of the coefficient is notable, this tells us that the dependent variable firm’s value is positively related to the oil price. This can be justified by the fact that it particularly uses oil as an input to produce consumer electronics, lighting and medical equipment. A low oil price would produce a decrease in the price of goods and services in the consumer goods sector, and likely stimulate competition. Also, the industry sector estimated by ASML is both significant and positively correlated (t=4,45). ASML manufactures machines for the production of integrated circuits. DSM reflects the healthcare sector and is also positively correlated and is significant at the 10% level (t=1,50). These oil price changes could affect the production efficiency of the firm and consequently stimulate opportunity growth and decrease the costs of corporate investments, which in return will enhance the value of the firm (Zaabouti et al., 2016).
The result on the basic materials sector estimated by AkzoNobel is significant at the 5% level (t=−1,77). The results of the technology sector represented by TomTom shows indeed a negative relation between the oil price and the firm’s value, however this result is not significant (t=−0,78). Although TomTom is a manufacturer that produces navigation and mapping products, the impact of using oil as an input variable in the production process is limited. Ahold, estimator of the retail sector, also has a significant negative correlation between the firm’s value (t=−2,49) and oil prices. This negative relationship between the oil prices and the firm’s value can be justified by the negative impact of oil prices on the firm’s stock prices. Consequently, this affects the market value of the stocks and thus the Tobin’s q. This could also mean that the corporate investment strategy is negatively influenced and the firm could experience funding problems in the distant future. According to Pawlina and Renneboog (2005) if a firm experiences distortion in the corporate investments strategy it is inhibited from reaching the optimal firm’s value.
13 Surprisingly, Royal Dutch Shell, which represents the oil and gas sector, is not significant negatively related (t=−0,15). There is clear evidence of significant correlation across other sectors, but this result is inconsistent with the persistence in performance. However, the lower coefficient of determination (R2
) is worth noting. The same ordinary least squares regression is used to either test the hypotheses or predict the future results. Another interesting result is the positive significant result of 1% for KLM (t=3,69). This result in the travel and leisure sector is similar to that in the consumer goods, healthcare and industry sector.
4.3 Estimating other variables
The coefficient estimated for the ratio of tangible long-term assets to revenues and its square, predominantly has the opposite signs to those estimated for the operating income. The operating income is positively correlated with the value of a firm. This effect is verified by Habib and Ljungqvist (2003).
The proxy for growth opportunities divided in research and development, advertising and capital expenditures are also estimated. The ratio of research and development expenses to fixed assets is positively related with the Tobin’s q and proves a significant result of 5% at TomTom. Most of the other firms and sectors are mostly positively related as well; however, these results do not prove to be significant. Economic intuition states research and development stimulates adopting a system of saving energy (oil). However, ratio of capital expenditures to fixed assets is negatively related with the firm’s value and proves a significant result of 10% and 5% at AkzoNobel and Ahold respectively.
The firm’s tangible capital intensity, denoted with the ratio of fixed assets to sales, is negatively related with the firm’s value. There is also a significant result of 1% at ASML. Economic theory enforces this result, because the Tobin’s q usually tends to understate the intangibles’ replacement costs. The inversed relation holds for the square of the equation.
The Fama-French risk premium was expected to negatively correlate with the Q, because the higher the costs of capital, the lower Q. However, the estimation results stated a significant positive correlation of 10% at DSM and TomTom.
14
4.4 Summary and discussion
The dependent variable in the regression is the firm’s value, defined by the Tobin’s q which is calculated by dividing the market value of the firm by the total assets. All standard errors are robust to heteroskedasticity. The results obtained from the regression analysis are sector and firm specific effects. The most important finding is that the changes in oil price are significant in almost every sector.
The data analysis reports significant positive correlations between the variable of interest, the oil price, and the value of firms in the consumer goods, healthcare and industry sectors. These positive correlations can be explained by a decrease in the price of goods and services for these companies, which could stimulate competition. Another possible explanation is the fact that these changes in oil prices can affect the production efficiency, which stimulates opportunity growth. The data also reports significant negative correlation between the oil price and the value of firms in the basic materials, technology, and the retail sectors. These negative correlations can be justified by the negative impact of oil prices on the firms’ stock prices. Consequently, this affects the market value of the stocks and thus the Tobin’s q. This could also mean that the corporate investment strategy is negatively influenced and the firm could experience funding problems in the distant future. A priori, the effect is ambiguous: changing oil prices may increase or decrease the firm’s value.
The reports on the oil and gas sector and the travel and leisure sector are intriguing. The value of firms in the oil and gas sector is not significantly negatively related to the oil price. Further, the firms participating in the travel and leisure sector are reported to have a significant positive effect by oil prices on the firm’s value. These results are inconsistent with our predictions.
The research had some limitations. Unfortunately, there were no data sets available on the research and development expenses for the retail, or travel and leisure sectors, estimated by Ahold and Air France-KLM respectively. This resulted in a regression for these two firms without the variable: R&D/K. This may have influenced the outcome of the regression and could also explain why the travel and leisure sector is positively correlated with a significance level of 1%. Perhaps a dummy for missing variables could solve this problem. In the regression analysis we made use of an estimation of the advertising costs, but the real numbers could provide a better insight in the actual results.
15 It was virtually impossible to calculate the four remaining sectors because the correct data was not available. We had to leave out the financial sector, media sector, telecommunication sector, and the utilities sector. But studying the effect of oil prices on the utility sector estimated by an energy company could provide a valuable insight in the actual effect. Adding more variables to the regression could also isolate the significance of the oil prices more. In this paper we assume that firms participate in a Modigliani-Miller world and leverage will therefore not affect firm value. However, if tax shields are valuable, they should increase the Q.
Since we consider the latest oil price drop of 2014 as a benchmark to estimate the general result of changing oil prices on the value of firms, combined with the use of quarterly reports, the number of observations is limited. The estimated coefficients and standard deviations could differ if we increased the number of observations. Doing so would also produce more precise results. To prevent the problem of availability bias, where recent events tend to be more magnified, future research should compare recent oil price changes with changes in the past, i.e. the oil crisis in 1973. This could also generate more observations and valuable insight.
To investigate the effect of oil prices on the firm’s value in further research, a range of robustness checks provides evident information for the production of statistical methods that are not fulsome affected by scale, location and regression parameters. Tests for possible endogeneity bias, size affects, classification and outliers could provide tested robust results. Another method of economic modelling is stochastic frontier analysis (SFA). Habib and Ljungqvist (2003) and Zaabouti et al. (2016) based their results on a stochastic frontier model. The explanatory power could differ between these models, because unlike the ordinary least squared regression this paper uses, the stochastic frontier analysis also estimates the shortfalls for the Tobin’s q.
16
5. Conclusion
The main goal of this paper is to establish the effect of changing oil prices on the value of firms, whereas the value of the firm is a topic that is globally ignored in the finance literature. This paper examines for the first time the impact of oil price changes on the value of firms for 8 different sectors, where evidence is based on Dutch listed firms.
Existing related literature focuses primarily on the effect of oil prices on macro-economic variables, but rarely on micro-macro-economic and firm variables. These studies investigated the effect of oil prices on investments, stock returns and cash flows. However, our expectations indicate that the effect of oil prices is different across sectors and can be both positively and negatively related. The results presented here validate new important discoveries that contribute to the understanding of the impact of oil prices on the value of firms.
The data analysis reports significant positive correlations between the variable of interest, the oil price, and the value of firms in the consumer goods, healthcare, and industry sectors. The data also reports significant negative correlation between the oil price and the value of firms in the basic materials, technology, and the retail sectors. These positive correlations can be explained by an increase in competition and production efficiency. These negative correlations can be justified by the negative impact on the firm’s stock prices and consequently the corporate investment strategy. A priori, the effect is ambiguous: changing oil prices may increase or decrease the firm’s value. Notable are the reports on the oil and gas sector and the travel and leisure sector. The regression states that the value of firms in the oil and gas sector is not significant negatively related to the oil price. Also, the firms participating in the travel and leisure sector are reported to have a significant positive effect of oil prices on the firm’s value. These results are inconsistent with our predictions and further research is evident.
Future research should focus on the limitations of this paper and further validate the effect of oil prices on the firm value of different sectors, including the omitted sectors. Besides increasing the number of observations, the correct data of both the advertising expenses and the missing research and development expenses should be included to improve the estimation results. Future research should also check for robustness and compare the results with a stochastic frontier analysis to validate the effect of oil prices.
Given the economic importance of oil prices, the results add important implications for both financial and cost management. Specifically, because the oil price fluctuates frequently and dramatically. It is important for the entire economy, particular in sectors where there are enormous energy consumers, to acknowledge this.
17
6. References
Aye G C, Dadam V, Gupta R, Mamba B. Oil price uncertainty and manufacturing production. Energy Economics, 2014, 43: 41-47
Berk, J. & DeMarzo, P. (2014). Corporate Finance (third edition). Pearson.
Dayanandan A, Donker H. Oil prices and accounting profits of oil and gas companies. International Review of Financial Analysis, 2011, 20(5): 252-257
Elyasiani E, Mansur I, Odusami B. Oil price shocks and industry stock returns. Energy Economics, 2011, 33(5): 966-974
Habib M A, Ljungqvist A. Firm value and managerial incentives: a stochastic frontier approach, 2003, 9-20
Himmelberg, C P, Hubbard, R G, Palia, D., 1999. Understanding the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics 53, 353-384
Huang R D, Masulis R W, Stoll H R. Energy shocks and financial markets. Journal of Futures Markets, 1996, 16(1): 1-27
Jin Y, Jorion P. Firm value and hedging: evidence from U.S. oil and gas producers. Journal of Finance, 2006, 61(2): 893-919
Jones C M, Kaul G. Oil and the stock markets. Journal of Finance, 1996, 51(2): 463-491 Kokemuller N. (2015), How much does it cost to advertise a business? Retrieved from http://smallbusiness.chron.com/much-cost-advertise-business-66152.html Mohanty S, Nandha M, Habis E, Juhabi E. Oil price risk exposure: the case of the U.S. Travel and Leisure Industry. Energy Economics, 2014, 41: 117-124
Narayan P K, Sharma S S. New evidence on oil price and firm returns. Journal of Banking and Finance, 2011, 35(12): 3253-3262
Pawlina G, Renneboog L. Is investment-cash flow sensitivity caused by the agency costs or asymmetric information? Evidence from the UK. European Financial Management, 2005, 11(4): 483-513
18 Sadorsky P. Oil price shocks and stock market activity. Energy Economics, 1999, 21(5):
449-469
Yoon K H, Ratti R A. Energy price uncertainty, energy intensity and firm investment. Energy Economics, 2011, 33(1): 67-78
7. Appendix
Table 1. Variable definition
Variable Definition
Tobin’s q Value of the firm. Ratio between the market value of the firm and the book value of assets.
Oil price Price of a barrel oil from the Crude Oil WTI
Sales Quarterly revenues at the end of period t. Net revenues as reported in the quarterly reports expressed in €m. Used to measure firm size and is both logged and logged squared.
R&D /K Ratio of research and development expenses to the fixed assets. Fixed assets are equal to the stock of property, plant and equipment (K). Estimates the role of research and development capital relative to other non-fixed assets, soft spending.
ADV/ K Ratio of advertising expenses to the fixed assets. Estimates to role of advertising capital relative to other non-fixed assets, soft spending.
CAPEX / K Ratio of capital expenditures to the fixed assets, hard spending. K / Sales Ratio of tangible long-term assets to revenues.
Y / Sales Ratio of operating income to revenues. Also called the operating margin, the ratio of operating income before depreciation to sales.
FFR Fama-French estimate of the risk premium
Table 2. Estimation results
Columns (1), (2), (3), (4), (5), (6), (7) and (8) present the models estimated results for AkzoNobel, Philips, DSM, ASML, Shell, Ahold, TomTom and KLM respectively using the ordinary least-squares. The dependent variable in all models is Tobin’s q. The revenue is the natural log of one and we also include the natural log squared. The models also include the square of the K/SALES division.
19
Table 2. Estimation results (Continued)
Variables AkzoNobel Philips DSM ASML Shell Ahold TomTom KLM
Constant -253,165 (1361,318) 548,104 (225,792) 111,739 (103,501) 576,672 (54,522) 54,542 (199,457) 177,184 (170,313) -102,863 (181,754) -8,367 (44,740) Oil price -0,008 (0,0044)** 0,009 (0,003)*** 0,001 (0,001)* 0,016 (0,004)*** -0,001 (0,008) -0,006 (0,002)** * -0,007 (0,009) 0,001 (0,0002)*** Ln(SALES) 98,688 (288,306) -122,400 (54,420)** -29,607 (26,992) -159,496 (14,700)*** -8,309 (33,958) -37,946 (29,807)* 33,648 (61,639) 3,251 (9,575) Ln(SALES)2 -6,180 (17,629) 6,958 (3,126)** 1,920 (1,755) 11,542 (1,025)*** 0,338 (1,483) 2,009 (1,625) -2,903 (5,683) -0,226 (0,551) R&D/K 2,987 (3,889) 0,346 (1,093) 6,883 (7,128) -12,704 (4,904)*** -58,776 (121,518) 0,918 (0,571)** ADV/K -538,622 (1253,131) -12,687 (33,800) 7,952 (13,788) -94,666 (12,850)*** -22,471 (44,595) 6,458 (100,813) 0,312 (10,451) -12,680 (37,632) CAPEX/K -2,607 (1,759)* -2,565 (2,124) -0,073 (1,224) 11,078 (2,948)** 3,804 (5,573) -4,372 (2,421)** 1,741 (1,311)* -0,201 (0,444) K/SALES -116,173 (347,027) -19,928 (58,163) 2,046 (1,369)* -10,638 (2,073)*** -1,656 (2,917) 7,057 (63,432) 61,927 (194,152) -1,676 (3,483) (K/SALES)2 31,053 (111,190) 10,109 (45,043) -0,523 (0,379)* 1,100 (0,280)*** 0,191 (0,309) -5,746 (29,823) -161,666 (449,171) 0,296 (0,637) Y/SALES 0,162 (1,565) 0,528 (1,461) -0,372 (0,395) -19,950 (3,771)*** 2,001 (2,377) 3,554 (6,221) 1,676 (2,957) 0,374 (0,237)* FFR -0,530 (1,006) -0,003 (0,891) 0,187 (0,141)* -1,597 (1,401) 0,182 (0,338) -0,016 (0,512) 1,027 (0,703)* -0,078 (0,110) R2 0,7912 0,8071 0,7022 0,9689 0,2505 0,8699 0,9067 0,8964 t -1,77** 3,63*** 1,50* 4,45*** -0,15 -2,49*** -0,78 3,69***
20
Table 3. Experimental input
Table 4. Experimental instructions
After importing the data in StataSE we convert the specified string variables to numeric variables and set the time-analysis to Q (quarterly). We then regress the dependent variable (TOBINSQ) to all of the independent variables (OILPRICE LSALES LSALES2 RDKK ADVKK CAPEXKK KKSALES KKSALES2 YSALES R), but also account for possible heteroskedasticity (vce(robust)). This process is repeated for all the firms and sectors.
