The Empirical Study on Dividends of Cross-listing Oil and Gas Firms in the U.S. Capital Market

(1)

Faculty of Economics and Business International Financial Management

Master Thesis Proposal

The Empirical Study on Dividends of Cross-listing

Oil and Gas Firms in the U.S. Capital Market

Xiaofei Zhang

Student number: 1939068 Msc of IFM

University of Groningen

Faculty of Economics and Business Course Code: EBM010A05

Supervisors: Dr. Henk von Eije Dr. Wim Westerman

(2)

2

Abstract

This empirical research employs Tobit analysis to compare the dividend levels during the period from 2006 to 2010 among the cross-listing oil and gas firms, the listed U.S. based firms, and listed firms outside the U.S. The results show that the cross-listing oil and gas firms pay more dividends than their home county counterparts. Moreover cross-listing oil and gas firm tend to pay the higher levels of dividend than the U.S. based firms. Higher levels of dividends paid help the cross-listing firms to decrease the negative impact of low recognition, and also to build a reputation for later equity issues. Finally, even though the civil-law-country firms do pay more cash dividends, there is no such corporate governance effect on the dividend payments of cross-listing firms.

(3)

3

Table

1. Introduction ... 4

2. Literature review ... 6

2.1 Dividends and agency theory ... 6

2.2 Cross listing ... 7

2.3 Signaling ... 8

2.4 Legislation environment ... 9

3. Hypotheses ... 11

4. Data and Methodology ... 13

4.1 Data and descriptive statistics ... 13

4.2 Control Variables ... 14

4.3 The applied model ... 15

5. Results and Analysis ... 17

6. Robustness check ... 20

6.1 Five years robustness check on the winsorized data ... 20

6.2 One year robustness check on the winsorized data ... 22

7. Conclusion ... 24

8. References ... 26

9. Appendix ... 29

(4)

4

1. Introduction

Over the last twenty years, more and more foreign firms listed their securities on the U.S. capital. The dramatic increase of the number of listed firms moving to the U.S. capital market even further enhanced its status as the financial center of the world. Because of these development researchers pay a lot of attention on the gains and costs of dual listing, and to the firms` intention of doing so. Miller (1999) finds that firms listing shares outside the home market gain positive abnormal returns around the announcement date. Miller also supports that more liquidity and the larger shareholder base increase shareholder wealth. Others also find that cross-listing on U.S. exchanges are associated with significantly positive stock market reactions. (Foerster and Karolyi, 1999, Doukas and Swizer, 2000)

However, the literature so far pays less attention to specific industries. Without the understanding the impact of dual listings for specific industries, the picture is not complete. This research only focuses on the oil and gas industry, since oil and gas firms represent different characteristics as firms from other industries. Most oil and gas firms, because of the limitation on geographic restriction and national legislation restriction, are mainly actively operating in a certain area. Normally oil and gas firms are relatively larger than the firms from other industries, thus some projects require vast amounts of money, like exploration, extraction, transportation, and distribution. Thus oil and gas firms are assumed to be thirsty for capital, which may be one of the main reasons why non-U.S. oil and gas firms choose to list in U.S. capital market which provides outstanding liquidity.

(5)

5

In a frictionless world, dividend policy has no relationship to the shareholder wealth. (Miller and Modigliani, 1961) However, the real world is not frictionless and paying dividends may signal that the company is strong. Such a signal increases the probability that investors are willing to provide additional cash for investment. Thus the research question is,

Do cross-listing oil and gas firms pay more dividends when they list on the U.S. capital market?

There are 41 non-U.S. oil and gas firms on the NYSE and 2 on the NASDAQ, while there are 102 U.S. oil and gas firms on the NYSE and 41 on the NASDAQ. The 43 foreign firms have distinct features because the home countries they come from represent systematic different investor protection systems and different financial market developments (La Porta et al., 2000, 2002). In order to give a clear picture of the cross-listing firms, this paper classifies non-U.S. firm in to two categories, firms from well-developed financial markets with good investor protection and firms from developing financial markets with poor investor protection system.

(6)

6

2. Literature review

2.1 Dividends and agency theory

Miller and Modigliani (1961) present irrelevance propositions which state that a firm`s dividend policy is irrelevant to its current market valuation with the firm`s investment policy already given in a perfect capital market. In the real world, cash inside and outside the firm is not equivalent. (Blanchard et al, 1994) Black (1976) pointed about the “dividend puzzle” that in a frictionless world firms with zero-dividends payments will be more attractive, since the dividend taxes are too heavy for most of investors. And therefore, many firms will be tempted to eliminate dividends payment. Firms, however, also face an expensive agency problem. This agency problem comes from the information asymmetry when the desires of the principal and agent conflict and when it is difficult or expensive for the principal to verify what the agent is actually doing. (Eisenhardt, 1989) According to agency theory, the profit should be partly paid out to the shareholders as dividends; otherwise, profits will be under the control of insiders for personal benefits or to undertake projects which generate unfavorable return. (Easterbrook 1984, Jensen 1986) To prevent managers from building unnecessary empires, shareholders want to see an increase in dividends that makes cash less available to managers. (Lang and Litzenberger, 1989)

(7)

7

capital. Later, however, internal funds will exceed the investment opportunities and the firms ends up with paying out the excess funds in order to mitigate the possibility that free cash flow will be wasted.

2.2 Cross listing

The major stock exchanges in United States, such as New York Stock Exchange and the NASDAQ, are attractive to the foreign firms, since they provide an easy access to a very high volume capital market which favors the needs of both shareholders and managers. Therefore firms either from developed economies or from developing economies were seeking to list in these capital markets in recent years. During the 1990s, the number of firms listed in major capital markets outside their home country dramatically increased to more than 4700 (Karolyi, 2006).

Firms have to consider barriers like regulation restrictions, additional cross-building costs and information asymmetry problems before they decide to execute a dual-listing. According to Karolyi (2006), cross-listing could offer new investors, access to a more efficient capital market, provide a competitive corporate governance environment and set up a high reputation for firms. 28% of respondents of the managers in overseas companies confirm that their primary purpose of listing in the U.S. is to increase the trading liquidity. (Mittoo, 1992) Also, Doidge (2007) indicates that more than 31% of foreign lists in the world are concentrated in the U.S. capital market.

(8)

8

employed weekly abnormal returns for two years around the listing date for 183 ordinary and ADR listings, and they find a significant positive 1 percent on average abnormal return. Foerster and Karoly indicate that the share price changes of cross-listing firms that are not only due to the international investment barriers, but also to the investor recognition and to liquidity factors.

Besides the benefits from dual-listing, there are also draw backs. The Financial Times (2005) indicates that a number of European firms listed in the U.S. capital market are considering quitting the U.S. market, since the overall cost outweighs the benefits. Some firms even claim that there are no benefits at all because cross-listing firms currently suffer large cost burden of having to reconcile financial statements with international standards, of paying a large amount of dividend and of being exposed to legal liabilities, taxes and various trading frictions.

2.3 Signaling

In no way will cross-listing firms have the same operating environment as to U.S. firms, and investors in U.S. are more familiar to the U.S. based firms instead of cross-listing firms. Because of the low recognition, cross-listing firms have to stand on a higher level of agency cost caused by the information asymmetry. Dividends, which function as signal, are able to smooth the tensions in the relationship between agent and principal. (Spence, 2002)

Dividend is one of the means to signal information to investors. Cantale (1996) shows the firms are trying to communicate the private information regarding their quality to outside investors. According to Lintner, (1956) the firms increase dividends when managers predict that the profits will increase, thus a dividend increase implies a positive signal to outsider shareholders about the firm`s profitability. Dividends send firm`s information to outsiders. It is assumed that in practice, the share price goes up when dividends are increased or initiated, and the share price goes down when dividends are cut.

(9)

9

flows (dividend signal hypothesis). Miller and Rock (1985) indicate, in line with Bhattacharya, that a lower level of investment maybe accompanied by a higher level of dividend. Since the founders of the firms might not accept wasting profitable investment opportunities, they try to eliminate the asymmetries. Especially for the cross-listing firms which list on U.S. capital market, the information they convey to the investors is not as full as the U.S. firms, and there are always information asymmetries between investors in the U.S. and the non-U.S. cross-listing firms. Hence investors would like to have dividends as a signal to convey information about the future profitability.

Leland and Pyle (1977) show that a manager holds large amount of shares of the firm to signal to outside shareholders that the firm has a high market value. Shlomo et al. (1997) also illustrate that dividend-increasing firms are less likely to have subsequent earning decreases. La Porta et al. (2000) stated in “substitute model” that firms pay dividends to establish a good reputation for decent treatment of outside shareholders in order to issue more equity in the future.

However, empirical research based on time-series regression analysis does not fully testify the hypothesis that dividends convey information to investors. Brickley (1983) finds an insignificant earnings increase in the year and the following year right after the dividend increase, which indicate an ambiguity relationship between dividends and future earnings. Penman (1983) examines 2217 management forecasts of annual earnings per share and dividend announcements from 1968 to 1973, and concludes that the changes in dividends do not convey much information about the future earnings. Further, Chihwa and Chunchi (1994) indicate that potential information content of dividends is very small.

2.4 Legislation environment

(10)

10

contractual framework of Jensen and Mecking, the rights of investors are protected by the legal system. When investors feel their rights, for instance the voting rights, are well protected by the legal system, they are willing to finance firms. Otherwise, the firms will find themselves in a difficult external finance situation because the investors hold back their investment.

In most countries, laws and regulations are enforced in part by market regulators, in part by court, and in part by market participants themselves. (La Porta et al., 2000) Due to the difference business environments in each country, the legal systems are not the same. La Porta et al. concluded that common law countries have a tradition of stronger protection of outsider shareholders, while civil law countries have a relatively weaker protection.

(11)

11

3. Hypotheses

Non-U.S. oil and gas firms are standing in a different position as the U.S. counterparts do. They come from other business environments with distinctive characteristics and backgrounds. The problem of information asymmetry rises because the outside investors are not quite familiar with the business environment the firms operate in, the business condition in the home country and, the quality of the management team of the cross-listing firms. And those types of information cannot be obtained from the annual report. Because worries like this, investors would hesitate to put their money into cross-listing firms, to the cost of forgoing some profitable opportunities.

To gain the confidence of U.S. investors, cross-listing firms are assumed to use their dividends as a signal to investors, which conveys the information about the future cash flow, and the quality of the firms` management and operation (Bhattacharya, 1979). Firms choose to payout dividends, according to La Porta et al. (2000) in the “substitute model”, to establish a good reputation for decent treatment of outside shareholders in order to be able to issue more equity in the future. Reputation can be set up if firms keep a constant high payout ratio, which makes investors to believe the firm possesses abundant cash flows.

In line with the “substitute model” (La Porta et al., 2002), the oil and gas firms pay dividends in order to please the investors to get higher equity issues in the future. But cross-listing firms are not so familiar to U.S. investors; they need additional stimuli to get money from investors, and so I assume that they even pay more dividends.

H1: the cross-listing oil and gas firms have a higher level of dividends payouts than U.S. firms in the U.S. capital market.

(12)

12

investment opportunities are not there. Moreover, investors in well protection environments have more trust in the firms where they put their money in. They are even willing to get little dividends or zero dividends, if they believe firms would make a better use of their money and return it in future. (La Porta et al., 2000) The expropriation in firms of civil law countries is not as difficult as that in firms of common law countries. Therefore the corporate governance of firms in common law countries are assumed to be better.

Oil and gas firms are assumed to be well managed in common law countries. On the other hand, oil and gas firms from civil law countries, would have incentives to signal investors, for they are not seen having a better corporate governance as firms from common law countries do. Therefore dividends could function as a signal. I assume cross-listing oil and gas firm from civil law countries have more incentives to do so.

H2: Cross-listing oil and gas firms from civil law countries have higher dividend levels than cross-listing firms from common law countries.

The decision of a dual-listing surely provides more access to capital, and therefore they may have better investment opportunities and better corporate governance in most cases. They are not only complying with the local legislation, but also the law in U.S. capital market as well. This change will significantly influence the capital structure and even corporate structure. Thus firms will present different characteristics as to the firms who did not list abroad. Because of better corporate governance and larger capital access, dual-listing firms should have a higher dividend level in comparison with firms that did not list abroad.

(13)

13

4. Data and Methodology

4.1 Data and descriptive statistics

I start from the complete sample of companies in NYSE and NASDAQ. Firm whose data are missing in the whole period are excluded in the sample. There are 174 oil and gas firms (131 oil and gas firms at the NYSE and 43 firms at the NASDAQ). The U.S. based firms are 102 from NYSE and 41 from NASDAQ. This means that cross-listing oil and gas firms are 29 from NYSE and 2 from NASDAQ. Since the oil and gas industry is highly related to the economic developments and fluctuate over time, I collect data in the last 5 years (2006-2010). All the data are obtained from Datastream. I put the data into two samples; the first one is the sample of all the oil and gas firms in NYSE and NASDAQ, which is divided into two sub-samples with U.S. firms and cross-listing firms. The second one is the oil and gas firms from non-U.S. countries.

The 43 cross-listing firms are from 10 countries separately, which according to the classification of La Porta, et al. (2000), are divided into a common law country sample (3 countries) and a civil law country sample (7 countries). Then I collect the list of oil and gas firms in those countries, for the period from 2006 to 2010, as well. There are 299 oil and gas firms form similar countries (49 firms from civil law counties, and 250 firms from common law countries). (Table 1) Together with the 174 oil and gas firms in the U.S. capital market, I have a panel data set with 473 firms for 5 years.

Table 1, numbers of observed oil and gas firms

Country Number Law system

Argentina 7 civil

Brazil 7 civil

China 6 civil

France 4 civil

Netherlands 6 civil

South Africa 2 civil

Canada 113 common

Hong Kong 12 common

Norway 17 civil

UK 125 common

(14)

14

The dependent variable in this research is the dividend level. In order to avoid any negative numbers from the dividends payout ratio, I use the ratio total dividends divided by total assets. There are 3 dummy variables as independent variables in my research which are used to test my hypotheses. The first independent variable is a dummy (cross-listing firms) which defines the cross-listing oil and gas firms as 1, and other firms as 0. The second independent variable is dummy (cross-listing civil-law-country firms) which defines all the cross-listing firms listed at the NYSE and NASDAQ which come from civil-law-country as 1, and other firms as 0. The third is independent variable is a dummy (all civil-law-country firms)) which defines the firms from civil-law-country as 1, and other firms as 0.

4.2 Control Variables

The proxy of leverage (LA) is total debt to total assets, which measures the impact from debt holders and measures the agency cost. Debt holders also play an important role in a firm, especially when managers decide on the level of cash dividends to pay. Debt holders do not have such motivation to have cash payout in form of dividends as shareholders do. In most of the cases they do not even like firms to pay out dividends. The larger amount of dividends paid to shareholders, the less operating cash flow are left inside the firm. Therefore, firms with high leverage may reduce dividends as agency costs are then less likely dread. Thus the relationship between LA and dividends level is assumed to be negative.

(15)

15

The proxy of market to book ratio (MB) values the investment opportunities and is shown as market value of asset divided by total book value of assets. It also values the extent of information asymmetry. Fama French (1999) conclude that firms with more investment opportunities have less incentive to pay dividends. In line with Fama French, I assume the dividends level of oil and gas firms has a negative relationship with the market to book value. Normally firms will not pay high dividends if the investment opportunities are larger, and shareholders will accept fewer dividends for higher future gains.

The proxy of relative change in assets (RCA) values the firms` assets growth rate. The higher assets growth rate means that there is large volume of investment opportunity. Firms normally intend to reduce the dividends to keep higher cash inside the firms when there are sound investment opportunities. Therefore I assume a negative relationship between dividends level and relative change in assets. Descriptive statistics of the dependent variable and the control variable are presented in appendix I.

4.3 The applied model

I use the Tobit regression to compare the dividends level of cross-listing firms and other kinds of firms. The choice of Tobit regression is motivated by the reason that some firms choose not to pay any dividends. Dependent variables therefore may be zero, and Tobit analysis in this case is suitable (Casu and Molineux, 2002).

(16)

16

The overall model I use is:

Where the dependent variable is dividend per asset unit at firm i in time t. Independent variables are the dummy variables with coefficient. measures the control variable leverage,

measures controls of the net income, measures the control of market-to-book ratio of each firm and measures the relative changes in assets. I will use lagged levels of the control variables in order to avoid endogeneity problem. is the intercept, are the coefficients of dummy variables. If not applicable,, a dummy is excluded from the equation.

(17)

17

5. Results and Analysis

5.1 The comparison of dividends levels within the U.S. capital market

Table 2 reports the Tobit results that tests for dividends level over 5 years (2006-2010). Column 2 shows the comparison of dividend level of U.S. based firms and cross-listing firms listed in NYSE and NASDAQ. Cross-listing oil and gas firms pay a significant higher level of dividends than U.S. based firms from 2006 to 2010, with a significant coefficient of 0.316 of the dummy (cross-listing firms). Thus, hypothesis 1 is strongly supported, which in line with Miller (1999) that the cross-listing firms having higher level of dividends, because they intend to decrease the negative effect of low recognition.

In the sample of the first column, there is a positive but insignificant additional effect of the cross-listing civil-law-country firms. This means that in this sample hypothesis 2 has to be rejected.

(18)

18 Table 2 The Comparison of Dividend Levels from 2006 to 2010

The dependent variable is dividends to assets. The table provides three Tobit regressions for 3 different groups of firms. The first column refers to the group of the U.S firms and cross-listing firms and tests the first hypothesis (H1). The second column refers to the group of cross-listing firms and tests the second hypothesis (H2). The third column refers to the groups of cross-listing firms and listed firms in foreign countries and tests the third hypothesis (H3).

Column 1 Column 2 Column 3 Column 4

Variables Sample Cross-listing firms and

U.S. based firms Cross-list firms

Cross-listing firms and foreign listed firms Dummy (cross-listing firms)

0.316 (H1) 0.047 (H3) (0.007) (0.000) Dummy (cross-listing civil-law-country firms) 0.604 0.015 (H2) (0.112) (0.192) Dummy (all civil-law-country firms)

0.033 (0.000) Leverage Level (-1) -0.655 -0.146 0.035 (0.160) (0.003) (0.024) Return on Assets (-1) -0.423 -0.025 0.044 (0.000) (0.077) (0.000)

Market to Book ratio (-1)

0.142 -0.009 -0.003

(0.000) (0.221) (0.142)

Relative Changes in Assets (-1)

-0.032 -0.014 -0.023 0.771 (0.253) (0.000) Constant -0.501 0.088 -0.029 (0.003) (0.000) (0.004) Left censored 225 17 712 Uncensored 259 94 393 Total firms-years 584 111 1105

(19)

19

(Miller and Rock, 1985; Fama French, 1999) is this not testified in this data set of oil and gas firms.

5.2 The comparison of dividends levels between cross-listing firms from common-law- countriy and from civil-law-country

Column 3 in table 2 shows the Tobit regression result of comparing dividend levels between cross-listing firms from common-law-country firms and from civil-law-country firms. As the results indicate, the coefficients of dummy (cross-listing civil-law-country firms) are 0.015, with the probability is 0.192. The result proves to be insignificant implying that civil-law-country cross-listing firms have no systematic higher level of dividends payment to shareholders than that of common-law country firms. The second hypothesis is again rejected, and there is not strong evidence to prove that cross-listing civil-law-country firms pay more dividends than cross-listing common-law-country firms.

The lagged leverage level, with the coefficient is -0.146, moves into the opposite direction with dividends level, indicating that the previous year leverage level does positively influence the dividends of cross-listing firms. The lagged return on assets proves to be insignificant negatively related to the dividends level (coefficient-0.025). The lagged market-to-book ratio is insignificantly negatively related to the dividends level (coefficient=-0.009). Firms keep high level cash inside the firms when agency cost is going up. Finally, the lagged relative change in assets is insignificantly negatively related to dividends level (coefficient=-0.014).

5.3 The comparison of dividend levels of cross-listing firms and firms from their home countries

(20)

20

gas firms do pay in general higher dividends than common-law-country firms. Therefore, while the hypothesis 2 is not confirmed for cross-listing firms in the U.S. (column 2 and 3), it is confirmed that outside the U.S., civil-law-country oil and gas firms pay more dividends. The lagged return on assets and lagged leverage are significantly negative related to the dividends level, and lagged relative change in assets positively related to the dividends level. The lagged market-to-book ratio is negative, but insignificant, related to dividends level.

6. Robustness check

6.1 Five years robustness check on the winsorized data

In order to decrease the possibility of extreme value of data to drive the result, I recreate a data set with the only highest 5 dividends to assets winsorized, and with the highest 5 and lowest 5 of each control variable winsorized. Then I perform the same test on the winsorized data, with the results shown in the table 3.

(21)

21 Table 3 The robustness check with winsorized data from 2006 to 2010

Variables Sample Cross-listing firms and

U.S. based firms Cross-list firms

Cross-listing firms and foreign listed

firms Dummy (cross-listing firms)

0.154(H1) 0.046(H3) (0.011) (0.000) Dummy (cross-listing civil-law-country firms) 0.091 0.015(H2) (0.502) (0.192) Dummy (all civil-law-country firms)

0.032 (0.000) Leverage Level (-1) -0.383 -0.146 0.046 (0.023 (0.002) (0.007) Return on Assets (-1) 0.093 -0.025 0.037 (0.002) (0.077) (0.000)

Market to Book ratio (-1)

0.049 -0.009 -0.003

(0.000) (0.221) (0.007)

Relative Changes in Assets (-1)

(22)

22

6.2 One year robustness check on the winsorized data

The duration of the previous tests are five years, from 2006 to 2010. Some of the cross-listing firms may not list in the U.S. capital market for whole period. Therefore I reduced the period of testing to only 2010 in which there are not missing data problems with cross-listing firm, and re-perform the Tobit analysis. The results are shown in table 4.

Table 4 The robustness check with winsorized data on 2010

Variables Sample Cross-listing firms

and U.S. based firms Cross-list firms

Cross-listing firms and foreign listed firms Dummy (cross-listing firms)

0.079(H1) 0.043(H3) (0.388) (0.004) Dummy (cross-listing civil-law-country firms) 0.125 0.014(H2) (0.322) (0.424) Dummy

(all civil-law-country firms)

0.034 (0.124) Leverage Level (-1) 0.016 -0.112 0.063 (0.910) (0.109) (0.043) Return on Assets (-1) -0.026 -0.008 0.023 (0.533) (0.732) (0.005)

Market to Book ratio (-1) 0.074 -0.008 -0.001

(0.000) (0.346) (0.803)

Relative Changes in Assets -0.027 -0.036 -0.003

(23)

23

The second column of table 4 shows that the dividends level of cross-listing firms in the U.S. capital market is higher than the U.S. based firms. However, now the coefficient of dummy (cross-listing firms) is 0.079 but the probability is 0.388. Thus the first hypothesis has to be rejected when I only investigate 2010 with winsorized variables. The lagged leverage level and lagged market-to-book ratio are all positively related to dividends level, but only the coefficient of the market-to-book ratio remains significant. The relations of leverage level and return on asset to dividends level contradict the 5 years test. This may also be caused by the possibility that the financial crisis in 2007 and 2008 biased the data.

The third column shows that cross-listing oil and gas firms from civil-law-country pay higher dividends in comparison with those firms from common-law-country, with an insignificant coefficient of dummy (cross-listing civil-law-country firms) that equals to 0.014. Hypothesis 2 is again rejected.

The last column of table 4 indicates that the cross-listing oil and gas firms have a significant larger amount of dividends payments than the firms listed only in the home market, where the coefficient of dummy (cross-listing firms) is 0.043 and the probability is 0.034. The third hypothesis is supported for the third time.

(24)

24

However, some limitations are relevant. First, the research is only focusing on the oil and gas industry firms which list in the U.S. and in a limited number of other countries. The sample size is thus not equally balanced between cross-listing firms and U.S. firms, neither between the common-law-country firms and civil-law-country firms. Secondly, the time period in my research is 5 years but the impact of financial crisis is not taken into consideration.

7. Conclusion

I collect dividend data for 473 oil and gas firms from NYSE, NASDAQ, and from the concomitant foreign stock exchanges, from 2006 to 2010. The empirical results of this paper suggest that the cross-listing oil and gas firms tend to pay more dividends than U.S. oil and gas firms. The dividends functions as a signal which conveys to the investors the positive information about the firms, which supports the “substitute dividend model” showed by La Porta et al. (2002) Besides, the dividends may reduce the agency cost (Jensen, 1986), moreover foreign firms may pay dividends to overcome the barrier of low investor recognition (Miller, 1999)

Further, I examine the dividend levels between cross-listing firms from civil-law-country and cross-listing firms from common-law-country. I cannot empirically conclude that civil-law-country firms have a higher level of dividends than firms from common-law-county. It is likely that the effect of cross-listing dominates so much, that the effects of governance in the common-law or civil-common-law country become materially irrelevant.

(25)

25

(26)

26

8. References

Bhattacharya, S., 1979. An Exploration of Non-dissipative Dividend-Signaling Structures.

Journal of Financial and Quantitative Analysis, Vol. 14, pp 667-668

Black, F., 1979, the Dividend Puzzle, Journal of Portfolio Management, Vol. 2, pp 5-8

Blanchard, O. J., Lopez-de-Silanes, F. and Shleifer, A., 1994. What do firms do with cash windfalls? Journal of Economics, Vol. 36, pp 337-360

Brickley, J., 1983, Shareholder Wealth, Information Signaling and the Specially Designated Dividend, Journal of Financial Economics, Vol. 12, pp 187-209

Cantale, S., 1996, the Choice of a Foreign Market as a Signal, Working Paper, Tulane University, Casu, B., Molyneux, P., 2002, A comparative Study of Efficiency in European Banking, Working Paper, University of Wales

Chihwa, K., Chunchi, W., 1994, Rational Expectations, Information Signaling and Dividend Adjustment to Permanent Earnings, Review of Economics and Statistics, pp 490-502

DeAngelo, H., DeAngelo, L., 2006, The Irrelevance of the MM Dividend Irrelevance Theorem,

Journal of Financial Economics, Vol. 79, pp 293-315

Doidge, C., Karolyi, G. A. and Stulz, R.M., 2007, Has new York become less competitive in global markets? Evaluating foreign listing choices over time, NBER Working Paper No. 13079 Doukas, J., Switzer, L. N., 2000, Common Stock Returns and International Listing Announcements: Conditional Tests of the Mid Segmentation Hypothesis, Journal of Banking and

Finance, Vol. 24, pp 471-501

Easterbook, F. H., 1984, Two Agency-Cost Explanations of Dividends, the American Economic

Review, Vol. 74 (4), pp 650-659

Eisenhardt, K. M., 1989, Agency Theory: An Assessment and Review, the Academy of

(27)

27

Fama, E. F., French, K., 1999, the Corporate Cost of Capital and the Return on Corporate Investment, the Journal of Finance, Vol. 54 (6), pp 1939-1967

Foerster, S.R. and Karolyi, G. A., 1999, the effects of market segmentation and investor recognition on asset prices: evidence from foreign stock listing in the United States, journal of

finace 54, 981-1013

Fuerst, O., 2008, A Theoretical Analysis of the Investor Protection Regulations Argument for Global Listing of Stocks, Working paper, Yale School of Finance

Karolyi, G. A., 2006, the world of cross-listing and cross-listings of the world: Challenging conventional wisdom, Review of Finance, Vol. 10, 99-152.

La Porta, R. F., Lopez-de-Silanes, A. Shleifer, and Vishny, R. W., 1997, Legal Determinants of External Finance, Journal of Finance, Vol. 54, pp 471-517

La Porta, R. F., Lopez-de-Silanes, A. Shleifer, and Vishny, R. W., 2000, Agency Problem and Dividend Policies around the World, Journal of Finance, Vol. 55, (1) pp. 1-33

La Porta, R. F., Lopez-de-Silanes, A. Shleifer, and Vishny, R. W., 2002, Investor Protection and Corporate Valuation, Journal of Finance, Vol. 57, (3) pp. 1147-1170

Lang, L., Litzenberger, R., 1989. Dividend Announcements: Cash Flow Signaling vs. Free Cash Flow Hypothesis, Journal of Financial Economics, Vol. 24, pp 181-191

Leland, H., Pyle, H., 1977, Information Asymmetries, Financial Structure, and Financial Intermediation, Journal of Finance, Vol. 32 (2), pp 371-387

Lintner, J. 1956. Distribution of incomes of corporations among dividends, retained earnings, and taxes. American Economic Review, Vol. 61(2), pp 97-113

Jensen, M. C., 1986, Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers, The

American Economic Review, Vol. 76 (2), pp 323-329

(28)

28

Miller, D. P., 1999, The Market Reaction To International Cross-Listing: Evidence From Depositary Receipts, Journal of Financial Economics, Vol. 51, pp 103-123

Miller, M. H., and Modigliani, F. 1961. Dividend Policy, Growth and the Valuation of Shares.

Journal of Business, Vol. 34 (4), pp 411-33

Miller, M. H., and Rock, K. 1985. Dividend policy under asymmetric information. Journal of Finance 40 (September): 1031-51

Mittoo, U. 1992, Managerial Perception of the Net Benefits of Foreign Listing: Canadian Evidence, Journal of International Financial Management and Accounting, Vol. 4, pp 40-62 Penman, S. H., 1983, the Predictive Content of Earnings Forecasts and Dividends, the Journal of

Finance, Vol. 38 (4), pp 1181-1199

Shlomo, B., Michaely, R., Thaler, R., 1997, Do Changes in Dividends Signal the Future or the Past? Journal of Finance, Vol. 52, pp 1007-1034

Spence, M., 2002, Signaling in Retrospect and the Informational Structure of Markets, American

(29)

29

9. Appendix

Appendix I. Descriptive statistics of variables

the U.S. capital market the foreign capital market

U.S. based oil and gas firms

cross-listing oil and gas firms from

common-law-country

cross-listing oil and gas firms from

civil-law-country

common-law-country oil and gas

firms

civil-law-country

oil and gas firms total firms Number of obersvation 143 firms 17 firms 14 firms 250 firms 49 firms 473 firms

Period of obersvation 5 year, fromb2006 to 2010

Dependent vatiable

Dividends to Assets Mean 0.030 0.037 0.046 0.014 0.041 0.022

St devi. 0.092 0.047 0.037 0.041 0.045 1.197

Control variables

Leverage Mean 0.252 0.245 0.158 0.164 0.184 0.195

St devi. 0.223 0.131 0.080 0,278 0.155 0.250

Return on Assets Mean 1.266 0.462 0.936 0.273 1.103 0.655

St devi. 7.533 0.345 0.366 0.783 0.810 4.202

Market to Book Mean 9.729 1.099 0.898 1.722 1.251 4.076

St devi. 108.119 0.917 0.676 3.432 1.071 59.759

Mean 0.516 0.237 0.167 1.086 0.182 0.823

Relative change in