The Firm-Specific Factors Influencing the Tax Haven Operations of Indian Multinationals

Abstract

The present study measures the role of firm-specific factors influencing the likelihood of establishing a subsidiary in tax haven countries. The panel data of Indian companies, which have business operations in foreign countries, are used for the study. The firm-level data for the period from 2007 to 2018 are analysed by using binary logistic regression model. The result shows that the intangible assets, long-term debt, number of subsidiaries and service sector dummy have significant and positive impact on tax haven operations of multinational companies, but the experience of the firm and return on equity are insignificant, and a firm’s size deters the likelihood of setting a tax haven subsidiary. The results also show that firms from high-technology manufacturing and knowledge-intensive sector have more influence on the likelihood of owning a tax haven subsidiary by Indian multinationals.

JEL Codes: F21, F23, H25, H26

Keywords

Tax haven countries Indian multinationals Intangible assets high-technology manufacturing firms

I. Introduction

There has been a notable change in Indian overseas direct investment destination in the past few decades. In the early period, overseas direct investment flowed to resource-rich countries like the UAE, Australia, etc., but, of late, overseas investment flows to low-tax countries such as Singapore, Mauritius, British Virgin Islands, Netherlands, etc. Tax haven countries are low-tax jurisdictions that provide the opportunities for tax avoidance (Desai et al., 2006). Multinational giants have shifted their profits to low-tax countries to reduce the tax burden on global income. According to Jones and Temouri (2016), Amazon, Google and Starbucks are the examples of multinational companies (MNCs) that have transferred their intangible assets to subsidiaries located in countries with low tax rate and high secrecy index. Desai et al. (2004) found that affiliates of US companies in tax haven countries were established as a part of international tax planning and affiliates in tax haven countries help US MNCs to shift their taxable income from high-tax jurisdictions to low-tax jurisdictions. The issue of overseas tax haven countries or territories has been a problem since many decades. But all national and international organisations failed to prevent the tax haven activities. According to the Organisation for Economic Co-operation and Development—OECD (2013) report, the current international tax standards might not have kept pace with changes in global business practices, particularly in the area of intangibles and the digital economy. Tax haven countries provide an opportunity to shift profits from countries that are high-tax jurisdictions to low-tax jurisdictions through transfer pricing. Aggressive pricing of intangibles and debt for intra-firm transfer has a major role in tax avoidance (The Institute of Cost Accountants of India [ICAI], 2018). According to Dunning (1988, 1995, 2000), the ownership advantage is an important category of advantages, which allows the firm to expand the business to foreign countries. The locational advantage of low tax in tax haven countries and the firm-specific advantages together facilitate the tax avoidance. There is a large flow of capital from India to tax haven countries, which leads to the negative impact on the Indian economy. A major portion of the investments in tax haven countries are not in real business activities, but it is in flow through nature. Wamser (2011) found that a higher statutory tax rate in a host country is related to a higher possibility to establish a conduit subsidiary in low-tax countries. The purpose and effect of overseas direct investment to tax haven countries are different from the overseas direct investment made in non-tax haven countries. According to Reserve Bank of India,¹ many tax haven countries occupy the top position in the list of Indian overseas direct investment receiving countries.

The core objective of this article is to analyse whether and to what extent the firm-level factors influence the likelihood of establishing subsidiaries in tax haven countries by Indian multinationals. We focus on mainly three variables that are considered as the most influencing characteristics of the firm on tax haven operations. The first emphasis is on the influence of intangible assets in choosing the tax havens as a location for the subsidiary company. The second is to focus on identifying whether the technological intensity of business operations are relevant for choosing the tax haven countries as a location for subsidiaries of Indian multinationals. The third emphasis is on the value of long-term debt. Other independent variables included in our model are number of subsidiaries, experience of the firm and firm’s size. Although a major part of the Indian overseas direct investments flow to tax haven countries, it does not get enough space in academic research and policymaking in the Indian context. Most of the earlier research on determinants of Indian overseas direct investment is based on the host-country-level factors, and there has been a lack of research about the use of a firm’s characteristics to avoid tax on global income of Indian multinationals. Research in the area of tax haven operations of multinationals has mainly focused on US firms. Studying the importance of firm-level characteristics influencing the tax haven operations of Indian firms will improve our knowledge of the tax sensitivity on attracting Indian firms to tax haven countries. Hence, the researcher made an attempt to explain the unexplored area of research.

For analysis of the objectives, we have developed six hypotheses based on firm-specific factors. We have analysed the data of 264 (2,835 firm years observation) Indian multinationals listed in Bombay Stock Exchange (BSE) 500 index and having at least one subsidiary in a foreign country. Period of study considered is from 2007 to 2018. The binary logit regression with white standard errors (White, 1982) has been used because our dependent variable is a categorical. The specified model includes intangible asset, technological intensity-based firm classification, experience of the firm, size of the firm and number of subsidiaries as independent variable.

The analysis focuses on the role of firm-specific factors that have potential to help the firms to reduce tax liability through tax haven-based tax avoidance strategies. The results show that intangible assets, long-term debt and dummy variable for firms from high-technology and knowledge-intensive industry have the most important role in tax haven operations of Indian multinationals. We also find that more the internationalisation (number of foreign subsidiaries), greater is the likelihood for tax haven operation. The experience of the firm has an insignificant influence on owning a tax haven subsidiary, at the same time, a firm’s size is significant but negatively related with the likelihood of tax haven operation of Indian firms.

The rest of this article is structured as follows: Section II includes theory background and review of literature of this study. Section III consist of theory on variables and hypothesis development. In Section IV, we explain the methodology, which includes the details on model specification, data source and measurements. Section V presents the analysis, results and discussion. Finally, in Section VI we conclude with implications of the study, scope of future research and limitations of this study.

II. Theory and Background

Generally, tax haven countries are the territories or jurisdictions, where low or nil tax or tax system favours a particular type of transactions. In terms of GDP, market size, resource availability, tax havens are not an attractive place for investment. Dharmapala and Hines (2009) found that 15% of countries in the world are tax havens, and these countries are small and affluent. Normally, the economy of tax havens depends on the financial service sectors, and they generate income through the fee charged from these multinationals for financial service. Mara (2015) found that low taxation and also the percentage of services in GDP are very important determinants of a tax haven country.

The MNCs have many ways to arrange its business to reduce tax on global income. Traditionally, the business has the practice of setting up subsidiary in low-tax countries. Subsequently, firm-specific characteristics are used for reducing the tax burden, which needs only internal ownership structure arrangements. Jaiswal (2017) found that jurisdictions, which are used for routing capital flows, have lower effective tax rates for investments in India as compared to the home country of the investors. Hong and Smart (2009) found that the income shifting to a tax haven may decrease revenues of high-tax countries, but the location of real investment is less reactive to tax rate difference since the substantial proportion of the foreign direct investment (FDI) were indirect. Investors channelise capital to the final destination through a tax haven or offshore financial centres or by using special purpose vehicle (SPV). The international profit shifting has an adverse effect on government revenues of high-tax countries and the competitiveness of domestic companies, which does not have access to tax haven profit shifting (Jog & Tang, 2001). According to Das and Banik (2015), Mauritius is a tax-free gateway to Africa for multinationals from India. They also found that there are multiple motives driving Indian firms to different overseas locations. The multinational enterprises (MNEs) transfer the intangible assets such as copy rights, patents, trademarks and licenses to subsidiaries established in low-tax jurisdictions and sell it to related subsidiaries in non-tax haven countries. It enables the firm to reduce the profit of subsidiary in high tax or non-tax haven country. Intangible assets, intra-firm debt, intra-firm transactions, etc., help the multinationals to shift profits to low-tax countries.

III. Theory and Hypothesis Development

Intangible Assets

The first objective of this article is to examine the influence of intangible assets on the likelihood of starting a tax haven subsidiary by Indian multinationals. As we have noted in the previous literature, the company will be able to shift profits to tax haven countries with the help of intangible assets. MNCs register the ownership of their patents and trademarks to tax haven subsidiaries (Jones & Temouri, 2016). The parent firm can easily transfer the ownership of intangible assets with subsidiaries in tax haven countries, so tax haven subsidiaries can receive a royalty. Grubert and Mutti (2009) found recently that the registration of intangible assets in foreign countries by US firms has increased. We see such a trend because of the MNC’s ability to route their investment through low-tax jurisdictions and make favourable cost-sharing agreements with affiliates in low-tax jurisdictions. According to Desai et al. (2006), companies with high research and development (R&D) expense to sales ratio are more likely to own a tax haven subsidiary. So we argue that:

H₁: MNCs with high intangible assets are more likely to own a subsidiary in tax haven country.

Technological and Knowledge Intensity

A technologically intensive firm has a great potential for growth and expansion at the international level. Their business is more dependent on internal R&D innovations and its outputs. Examples of such industries are robotics, telecommunications and semiconductor and computer industries. The high-technology firms have more valued intellectual properties than low-technology firms, and their major source of revenue is the sale of intellectual property–oriented products. So there is a high probability of high-technology manufacturing and service sector firms to establish a subsidiary in tax haven country. These firms can transfer intellectual property to tax haven countries and resell it to subsidiaries in non-tax haven countries. It enables MNCs to shift profits to low-tax countries through receiving of royalties by tax haven subsidiary. Jones and Temouri (2016) found technology-intensive manufacturing MNCs with significant level of intangible assets and MNCs in service industries have a higher likelihood of owning subsidiaries in tax haven countries. The industries are classified into six groups as per technological intensity of business operation to measure the impact of degree of technology in their business operation. Based on previous literature, we argue that:

H₂: Technologically intensive manufacturing firms and knowledge-intensive service firms have more likelihood to start tax haven subsidiary.

Long-term Debt

The multinationals enjoy the advantages relative to domestic firms in arranging internal debt instead of external debt for interest deduction (Desai et al., 2004). The interest expense on debt can be deducted from the profit of the company but return on equity cannot. When the loan is used for a profit-shifting purpose, the firm lends and borrows between parent and subsidiaries. Most of the literature suggests that a foreign subsidiary depends on more internal debt when a multinational has a subsidiary in a low-tax country (Graham, 2003; Mintz & Smart, 2004). MNCs establish conduit entities in tax haven countries and convert their debt into equity. The subsidiary in low-tax countries can provide debt to subsidiaries in high-tax countries. Since the interest is subject to tax in a low-tax country, the parent company can minimise their tax liability on global income. So the subsidiary in tax haven countries would be financed with equity and the parent and non-tax haven subsidiaries are allowed to borrow from tax haven subsidiaries. Huizinga et al. (2008) and Altshuler and Grubert (2003) found that host-country tax rates effect internal debt. So we argue that a higher value of long-term debt is positively related to the likelihood of starting a tax haven subsidiary by Indian firms.

H₃: Indian MNCs holding a high value of long-term debt is more likely to have a tax haven subsidiary.

Firms’ Experience

A firm’s experience is used as a control variable in this model. The age of the firm influences the firm’s performance and expansion decisions. Older firms can benefit from accumulated knowledge in better technology, well-developed supply channels, easier access to resources, well-established customer relations, better human capital and lower financing costs. Greiner (1972) found that the firm’s size is linearly related with the firm’s age. Gibrat (1931) found that younger firms are more likely to grow faster than older firms. Bhayani (2010) found that the age of the firm has a positive sign on the operating profit ratio, liquidity, interest rate and the firm’s profitability. Gaur (2010) found that age did not have significant influence on profit or net worth. Saika et al. (2020) found that prior experience and institutional advantage are important motives for expanding the business to overseas by Indian firms. After considering the aforementioned literature, we developed the following hypothesis.

H₄: Indian MNCs with more experience have greater likelihood on owning a tax haven subsidiary.

Number of Foreign Subsidiaries

To avail the benefits of host-country characteristics and firm-specific characteristics, the multinational firm would internalise its overseas operations. The number of foreign subsidiaries is considered to measure the extent to which the firm has internalised its overseas business operations. The firm will internalise their foreign business operations when the firm has technical know-how, marketing ability and consumer goodwill (Caves, 1971; Hymer, 1976). Internalising the overseas business operations gives more protection to the intellectual properties developed by firms. So, we argue that:

H₅: Indian MNCs with more number of foreign subsidiaries have greater likelihood on owning a subsidiary in tax haven country.

Firm’s Size

A firm’s size is defined as the annual revenue derived from sales turnover. Higher the sales, more the likelihood of starting a subsidiary in a foreign country. Local sales are the main motive for overseas direct investment by large firms (Kumar, 2008). Expansionary FDI is beneficial to domestic industries, since it increases the parent firm’s growth rate (Chen & Ku, 2000). Multinationals invest in foreign countries to access new markets for their products. Entering new markets is one of the ways to increase the sales in MNCs. Buckley et al. (2009) found that market seeking was a key motive for Chinese overseas direct investment. So we argue that:

H₆: Indian MNCs with higher sales positively relate to the likelihood of starting a subsidiary in a tax haven country.

Profitability

Return on equity indicates how well the firm has used the resources of owners (Pandey, 2015). The main agenda of opening a subsidiary in tax haven country is to shift profits to tax haven countries to reduce the tax burden. Firms can reduce their tax burden through locating their operations or shifting their profits to jurisdictions with low or no corporate taxation (Blouin & Krull, 2009). Firms with higher profitability also may have the capability to expand their business to foreign countries. So we argue that:

H₇: Indian MNCs with higher profitability have a greater likelihood of owning a subsidiary in a tax haven country.

Manufacturing or Service Sector Dummy

Based on sectoral division, highest investments are made in transport, storage and communication services, followed by manufacturing activities and agriculture and mining. (Care Ratings, 2014). Since the 1990s, firms have invested in almost all sectors, but increasingly in the services sector, led by the software industry. The services sector has turned out to be home to the largest number of outward-investing firms from India (Pradhan, 2017). Retail and services is the lead sector in investments overseas and the main activities identified within this sector are Information Technology (IT) services, logistical services and retail trade (Padilla-Perez & Nogueira, 2016). The universality of digital MNEs and their ability to sell goods and services in markets without having a physical presence enable them to shift profits easily. MNEs in service industries are, in general, also more likely to invest in tax havens. So we argue that:

H₈: Indian MNCs operating in the service sector have greater likelihood of owning a subsidiary in tax haven country.

IV. Research Methodology

Variables, Data Source and Measurements

For conducting this study, the data are collected from Bloomberg database and annual reports of the companies. Companies having at least one subsidiary in a foreign country are the criteria used for the selection of companies from BSE 500 index list. Then, companies without sufficient data and no foreign subsidiaries are excluded; finally, we listed 264 companies as the sample for the study. The study involved 2,834 observations, out of which 1,943 firms have subsidiaries in tax haven countries and 891 observations have no subsidiary in tax haven countries. Data are in the form of unbalanced panel of 2,834 firm-year observations. We use only the firm-level data of the parent firm. No financial data about the subsidiaries are used in this study.

Dependent Variable

Our dependent variable is a categorical variable whether or not the firm has a subsidiary in tax haven countries. The dependent variable is 1 if the firm has at least one subsidiary in tax haven country, otherwise 0. After identifying the firms with foreign subsidiary operation, we determined whether such a firm has subsidiary in tax haven country to create dependent variable, which is a categorical variable. This study considers 45 countries as tax haven countries. This list is based on the countries listed as tax havens by the International Monetary Fund—IMF (Vaidyanathan, 2017). The IMF list contains a total of 46 countries; from this list, we dropped the countries like Costa Rica and Malaysia based on the recent guidelines provided by Oxfam international, OCED and some researchers. Also, we made another change by including Netherlands in the tax haven countries list because many other researchers and organisations found that the Netherlands is a tax haven country. Netherlands is found to be one of the 15 worst corporate tax havens in the world by Oxfam international (Berkhout, 2016). Ireland, Luxembourg and Netherlands offer tax benefits to business firms with significant intangible assets and also offer a liberal tax system to domestic entities with overseas operations (Dorfmueller, 2003; Eicke, 2009). Our study uses the tax haven concept in a broad sense as explained by Dharmapala and Hines (2009), Diamond and Diamond (2002) and Hishikawa (2002). Names of all countries included in the tax haven country list are presented in Appendix 1.

Independent Variables

Firm-specific characteristics are used as independent variables in the specified model. According to Barney (1991), firm-specific resources include different assets, organisational processes, knowledge and information. The firm’s size, experience of the firm, number of subsidiaries, intangible assets, firm capabilities, long-term debt and technology-based industry classification dummies are included as independent variables in our binary logit regression model. The firm’s size and experience of the firm are used as control variables (Jorissen et al., 2005). To deal with 0 values for non-dummy variables, we added 0.001 to the value of the variable having 0 (Eichengreen & Irwin, 1995). The intangible asset is defined as assets without physical presence. It includes the value of patents, copy rights, trademarks, trade names, goodwill, organisation costs, capitalised development costs and franchises, licences, property rights and core deposits of intangibles in case of banks. The variable long-term debt is included in the model since the capital structure has an impact on tax haven decision. We defined long-term debt as all interest-bearing financial obligations that are not due within a year. It includes convertible, redeemable, retractable debentures, bonds, loans, mortgage debts, sinking funds and long-term bank overdrafts. It is the investment financed through long-term borrowings. The annual turnover of the firm is used as a proxy for measuring the firm’s size (Kumar, 2008). Sales is defined as the amount of sales generated by a company after the deduction of sales return, allowances, discounts and sales-based tax. To measure the experience of the firm, we use the variable age of the firm. A firm’s age is defined as the number of years that the firm has been operating since its incorporation (Ilaboya & Ohiokha, 2016). The company is an artificial legal person that comes into existence through incorporation (Gitzmann, 2008; Pickering, 1968). So we considered the year of incorporation as base year for calculating the age of the firm. To classify the firms based on technological intensity, we use the NACE two digit codes, which is the classification of economic activities introduced by Eurostat. The firms are classified into six groups as per the technological intensity of firms. According to this classification, we have created six dummy variables. Table 1 presents the industry classification details of firms. Our sector classification (Dummy variables) includes high-technology manufacturing firms (10.61%), medium/high-technology manufacturing firms (28.03%), medium/low-technology manufacturing firms (15.91%), low-technology manufacturing firms (12.12% of the sample firms), knowledge-intensive service firms (24.76%) and less knowledge-intensive service firms (7.58%). Low-technology manufacturing industry dummy is taken as the reference category dummy for comparing the significance of all other industrial dummies. The number of foreign subsidiaries is used to measure the extent to which a firm has internalised its overseas business activities. It is defined as the total number of subsidiaries owned by a parent firm in foreign countries. The domestic subsidiaries are not counted, and the data on the number of subsidiaries are obtained from the annual financial statements. The variable return on equity is defined as the ratio of net profit to capital employed. A dummy variable named as manufacturing or service dummy is used to identify firm from which sector is tend to start tax haven operations. We converted the values of intangible assets, firms’ size, long-term debt and return on equity to log form to reduce the problems due to high difference between minimum and maximum values. Intangible assets, firms’ size and long-term debt are measured in Indian rupee. The source of data is presented in Table 2.

Table 1.

Technology Intensity Classification of Firms.

Industry	Observations	Percentage of Observations	Number of Firms	Percentage of Firms
High-technology manufacturing	309	10.90	28	10.61
Medium- to high-technology manufacturing	799	28.18	74	28.03
Medium-technology manufacturing	447	15.77	42	15.91
Low-technology manufacturing	350	12.35	32	12.12
Knowledge-intensive service	706	24.90	68	25.76
Less knowledge-intensive service	224	7.90	20	7.58
Total	2,835	100	264	100

Source: Compiled by authors on the basis of data from Bloomberg and annual financial statements of companies.

Table 2.

Data Sources and Measurement.

Variable	Data source	Variable explanation
Tax haven firms or non-tax-haven firms (dependent variable)	Annual financial statement	An Indian firm is a tax haven firm if it has at least one subsidiary in any of the tax haven country, otherwise firm considered as non-tax-haven subsidiary
Firm size	Bloomberg	Amount of sales generated by a company after the deduction of sales return, allowances, discounts and sales-based tax
Intangible assets	Bloomberg	Total amount of intangible assets. This amount includes patents copy rights, trademarks, trade names, goodwill, organisation costs, capitalised development costs and franchises, licences, property rights and core deposits of intangibles in case of banks
Long-term debt	Bloomberg	All interest-bearing financial obligations that are not due within a year. It includes convertible, redeemable, retractable debentures, bonds, loans, mortgage debts, sinking funds and long-term bank overdrafts
Return on equity	Bloomberg	Ratio of net profit to capital employed
Firm experience	Bloomberg	Years since the company incorporated/registered
Number of foreign subsidiaries	Annual financial statement	Total number of overseas subsidiaries of a parent firm
Service sector dummy	Bloomberg	It is equal to 1 if the firm is operating in service sector, otherwise 0

Source: Compiled by authors on the basis of data from Bloomberg and annual financial statements of companies.

Analysis

The Specification of Estimated Model

The primary objective of this article is to examine whether the firm-specific characteristics have an influence on the likelihood of owning a subsidiary in tax haven countries with special emphasis on intangible assets and technological and knowledge intensity. Based on the review of literature, we developed seven hypotheses; the researcher proposes a binary logit regression model (Brajcich et al., 2016; Chari & Acikgoz, 2014; Mani et al., 2007).

\ln \frac{P}{1 - P} = α + β_{1} X_{1} + β_{2} X_{1} + β_{3} X_{3} + \dots + β_{n} X_{n}

P = \frac{b^{α + β_{1} X_{1} + β_{2} X_{2} + β_{3} X_{3} + \dots + β_{n} X_{n}}}{b^{α + β_{1} X_{1} + β_{2} X_{2} + β_{3} X_{3} + \dots + β_{n} X_{n} + 1}}

β = slope of the best-fitting equation.

X = independent variables.

α = intercept of best-fitting equation.

Ln = natural log of the odds.

p = probability that Y = 1.

b = the mathematical constant that is the base of natural logarithm.

Binary logistic regression is used to analyse the impact of the firm-level characteristics on tax haven country operations. To reduce the problem of heterogeneity, binary logit regression with white standard errors (White, 1982) is used. The basic empirical model is as follows:

\begin{array}{l} \ln \frac{P (THfirms/NTHfirms)}{1 - P (THfirms/NTHfirms)} \\ = β 0 + β \log {FIRM}_{SIZE} i t + β \log {INTANGIBLE}_{ASSETS} i t + β \log {LONG}_{TERM} DEBT i t + β {FIRM}_{AGES} i t \\ + β NO {.OF}_{FOREIGN} SUBSIDIARIES i t + β {PROFITABILITY}_{i t} + β SERVICE SECTOR DUMMY + β INDUSTRY DUMMY 1 \\ + β INDUSTRY DUMMY2 + β INDUSTRY DUMMY3 + β INDUSTRY DUMMY5 + β INDUSTRY DUMMY6 + ε_{i t} \end{array}

The dependent variable in our specifications is a dichotomous variable, indicating whether the firm has a subsidiary in the tax haven country, identified as per the tax list published by the international organisations such as OECD and IMF (see Appendix 1). A firm is classified as TH firm (tax haven firms), if the firm has at least one subsidiary in a tax haven country, otherwise non-TH firm (non-tax haven firms). In the regression model, i stand for the firm having a subsidiary in a foreign country, and t stands for time. So log INTANGIBLE ASSETS _it refers to log value of intangible assets of ith firm at time t. Log FIRM SIZE _it refers to the log value of sales turnover of ith firm at t time. Log LONG TERM DEBT _it refers to log of loans and other debts having a maturity period of more than 12 months by ith firm in t year. FIRM AGE _it refers to life of ith firm in tth year since the year the company was incorporated. NO. OF FOREIGN SUBSIDIARIES _it refers to the total number of overseas subsidiaries of ith firm in tth year. βPROFITABILITY _it refers to return on equity of ith firm at time t. βSERVICE SECTOR DUMMY is the categorical variable. It is equal to 1 if the firm is operating in service sector, otherwise 0. As mentioned in Table 1, we use the dummy variable for technology-based industrial classification of manufacturing and service sector. We classified the firms into six categories based on their technological intensity of business operation and created six dummy variables to compare which sector has an important influence on tax haven operation. INDUSTRY DUMMY1 is equal to 1 if firm i in high-technology manufacturing category, 0 otherwise. INDUSTRY DUMMY2 is equal to 1 if the firm is included in medium- to high-technology manufacturing category as per NACE classification, otherwise 0. INDUSTRY DUMMY3 is equal to 1 if the firm is included in medium technology manufacturing category as per NACE classification, otherwise 0. INDUSTRY DUMMY4 is equal to 1 if the firm is included in low-technology manufacturing category as per NACE classification, otherwise 0. INDUSTRY DUMMY5 is equal to 1 if the firm is included in knowledge-intensive service sector as per NACE classification, otherwise 0. INDUSTRY DUMMY6 is equal to 1 if the firm is included in less knowledge-intensive service category as per NACE classification, otherwise 0. Low-technology manufacturing category (INDUSTRY DUMMY4) is used as reference category dummy for comparing the significance of all other industry dummies. The definition of each variable in our model is highlighted in Table 2.

V. Empirical Results and Discussions

Descriptive Statistics

Table 3 reports the descriptive statistics of 264 firms for 2,834 firm-year observations. The table for descriptive statistics reports number of observations, mean, median, minimum, maximum, standard deviation and kurtosis. Mean of number of subsidiaries is 13.00, and the minimum is 1. Mean age of firms is 37.65 with the standard deviation of 23.13. The standard deviation for intangible assets, profitability, long-term debt and sales are 1.22, 0.476, 1.43 and 0.69, respectively.

Table 3.

Descriptive Statistics.

Variables	Mean	Median	Maximum	Minimum	Standard Deviation	Skewness	Kurtosis
Tax haven (DV)	0.68	1	0	1	0.46	–0.79	1.63
Age of the firm	37.65	30.00	100.00	1.00	23.13	0.92	2.98
Log intangible assets	2.85	2.93	6.10	–1.26	1.22	–0.18	2.80
Log long-term debt	3.35	3.55	6.57	–1.70	1.43	–0.67	3.35
Log sales	4.53	4.46	6.69	2.14	0.69	0.38	3.63
Number of foreign subsidiaries	13.00	6.00	60.00	1.00	15.75	1.72	5.10
Return on equity	1.139	1.253	2.280	–1.69	0.476	–2.050	5.48
Service sector dummy	0.327	0.000	1	0	0.469	0.73	1.53

Source: Authors’ calculation on the basis of data from Bloomberg and annual financial statements of companies.

Correlations

Table 4 provides correlation between each dependent variable in our specification. Multicollinearity is an assumption for logistic regression. Examining of correlation matrix indicates that all correlation values are less than 0.6. So there is no problem of multicollinearity. For further confirmation, we tested the variance inflation factor (VIF) to identify the correlation between independent variable. The collinearity statistics VIF and tolerance level are presented in Table 5. It shows that the data do not suffer from multicollinearity. Besides, none of the variable in Table 6 has a VIF more than 10, ensuring no serious problem of collinearity in our model.

Table 4.

Pearson Correlation.

Variable Name	Log Intangible assets	Log Long-term Debt	Return on Equity	Log Sales	No. of Foreign Subsidiaries	Service Sector Dummy
Log long-term debt	0.2010	1
Return on equity	–0.0603	–0.2161	1
Log Sales	0.26210	0.2893	–0.0054	1
No. of foreign subsidiaries	0.4561	0.2071	0.0203	0.2033	1
Service sector dummy	0.0747	–0.047	–0.190	–0.0345	0.1109
Age of the firm	–0.0834	0.1551	–0.0171	0.2947	–0.0422	–0.2124

Source: Authors’ calculation on the basis of data from Bloomberg and annual financial statements of companies.

Table 5.

Collinearity Statistics.

Co-variables lumn1	Tolerance	VIF
Log long-term debt	0.80	1.24
log intangible assets	0.74	1.36
log sales	0.77	1.30
log return on equity	0.88	1.14
High technology manufacturing dummy	0.58	1.73
Medium to-high technology manufacturing dummy	0.42	2.37
Medium technology manufacturing dummy	0.51	1.97
Knowledge-intensive service	0.42	2.39
Less knowledge-intensive service	0.63	1.59
Age	0.84	1.20
No. of foreign subsidiaries	0.74	1.36
manufacturing/service dummy	0.90	1.11

Source: Authors’ calculation on the basis of data from Bloomberg and annual financial statements of companies.

Table 6.

Logit Regression Results.

Variables/Specifications	Model 1			Model 2
Variables/Specifications	Coefficient	Odd Ratio (e^β)	p-Value	Coefficient	Odd Ratio (e^β)	p-Value
C	–0.43351	0.64822	0.234	–1.31117	0.2695	0.000
Log sales	–0.42335	0.65484	0.000***	–0.04274	0.9581	0.090*
Log long-term debt	0.24487	1.29745	0.000***	0.13466	1.1441	0.000***
Log intangible assets	0.30481	1.35637	0.000***	0.20728	1.2303	0.000***
Age of the firm	0.00339	1.00339	0.122	0.00157	1.0033	0.122
Number of foreign subsidiaries	0.13724	1.14711	0.000***	0.14114	1.1515	0.000***
Industry classification dummy
High-technology manufacturing	0.58949	1.80307	0.0075***	0.66578	1.9460	0.001***
Medium- to high-technology manufacturing	0.07579	1.07872	0.6165	0.12012	1.1276	0.420
Medium-technology manufacturing	0.41709	1.51754	0.0109**	0.37933	1.4613	0.017**
Knowledge-intensive service	0.63752	1.89179	0.0001***	0.70055	2.0148	0.000***
Less knowledge-intensive service	0.54692	1.72793	0.0201**	0.56456	1.7586	0.013**
Return on equity				–0.12861	0.8793	0.278
Manufacturing service				0.44793	1.5650	0.000***
Observations	2834			2834
Pseudo R²	23.64%			23.27%
Percentage of correct predictions	75.34%			75.72%

Source: Authors’ calculation on the basis of data from Bloomberg and annual financial statements of companies.

Note: ***, ** and * Statistically significant at 1%, 5% and 10%, respectively.

Logit Regression Results

The study estimates 2 logit regression models, and the results are presented in Table 6. Model (1) includes the variables such as sales, long-term debt, intangible assets, age of the firm, number of foreign subsidiaries and technology-based industry classification dummies. The specification 2 added two more variables, namely return on equity and service tax dummy. The sign and significance for all variables are more or less unchanged in both specifications. Further, the percentage of correct prediction increased marginally. This ensures the robustness of results. In this article, we discussed the results based on specification 2.

The results include the effect of each variable on the likelihood of establishing a tax haven subsidiary by the Indian multinational firms. We also report pseudo R² and percent of correct prediction as a measure of fit. Results support Hypotheses 1, 2, 3, 6 and 8. The variables—intangible assets, dummy variable for high-technology firms and knowledge-intensive firms, long-term debt, number of foreign subsidiaries and manufacturing/service dummy—are statistically significant at the 1% level of significance with a positive sign. The sales variables is significant but contrary to our Hypothesis 6. However, the age of the firm is statistically insignificant in our model. The sales variable is significant but negatively related.

The variable—intangible asset—shows significant and positive influence on the likelihood of establishing a tax haven subsidiary by the Indian multinational with p-value less than 1%. So our first hypothesis is supported. The results for intangible assets indicate that a 1 unit increase in intangible assets will result in an increase in the likelihood of establishing a subsidiary in tax haven countries by odds ratio of 1.2303 times. The result is consistent with the observation that MNEs with significant intangible assets are more likely to set up a subsidiary in tax haven country (Jones & Temouri, 2016). Treaty shopping and registration of intangible assets in tax havens enable the firms to reduce their tax burden. The result is also consistent with the observation that there is an increase in registration of intangible assets of US multinationals to foreign countries because the US firms can create hybrid firms in their affiliates abroad and form a cost-sharing agreement with affiliates in low-tax countries for reducing the tax (Grubert & Mutti, 2009). However, Brajcich et al. (2016) found that there is no association between the proportion of intangible assets held as intellectual property (IP), and the existence of resource shifting. Our results indicate that the firms with high-valued intangible assets have a higher probability to engage in more tax haven operations to reduce tax on their global income. Companies can register or transfer their intellectual properties to tax haven subsidiaries and then sell it to subsidiaries in high-tax countries and receive the income in the way of royalty. It leads to shifting of income to subsidiary located in the tax haven country.

To assess the effect of technology intensity of manufacturing and service firms, we introduced five technology-intensity-based industry dummies based on NACE two-digit classifications. The low-technology industry is used as the reference category in this model. The dummy variables—high-technology-intensive industry and knowledge industries—are strong and have greatest influence (p-value < 1%) among five dummies used as proxy for industry classification. The high-technology industry and knowledge industry dummy has significantly positive relation with the likelihood of establishing a tax haven subsidiary with the odds ratio of 1.9460 and 2.0148, respectively, for each category. Thus, our second hypothesis is also supported. The industry group consisting of medium-to-high technology is not significant in our model. But medium-technology and low-technology groups are significant at the 5% level of significance. Hence, we can concludes that the Indian multinationals operates in tax haven countries irrespective of technological intensity in their operations. But high-tech manufacturing and knowledge industry firms have more influence on tax haven operations of MNCs. The results suggest that firms from high-technology manufacturing and service sector attract tax haven countries because of more R&D activities and intellectual assets.

The variable ‘long-term debt’ is significantly (p-value = < 1%) and positively (odds ratio = 1.1441) related, suggesting that higher value of long-term debt is more likely to establish a tax haven subsidiary by Indian firms. This confirms hypothesis 3 that the value of long-term debt has a positive influence on the likelihood of establishing a tax haven subsidiary by an Indian multinational. It can be attributed to the benefit arising on account of charging interest as a tax deductible expense from business income. Results state that 1 unit increase in long-term debt will result in 1.298 times increase in the likelihood of owning a subsidiary in tax haven country by Indian multinationals. Buettner and Wamser (2013) found that the degree of tax effects was small for German firms, suggesting that internal debt is not important for German firms in shifting profits because of the controlled foreign corporation rules.

The number of foreign subsidiaries are significant (p-value = < 1%) and positively related with the likelihood of tax haven operation. The research suggests that more the internationalisation, greater is the likelihood of owning a subsidiary in tax haven countries. Hence, hypothesis 6 is confirmed by our results. Results show that 1 unit increase in the number of subsidiaries has resulted in 1.1515 odds of times increase the likelihood of owning a tax haven subsidiary by Indian multinationals.

The firm size and age of the firm are used as control variable in this study. The firm’s size is significant but negatively related in our model with the odds ratio of 0.9581. It is contrary to our sixth hypothesis that sale is positively related to the likelihood of owning a tax haven subsidiary by Indian multinational. The results suggest that the lesser the sales, more is the possibility for tax haven operation. Since it carries a significantly negative coefficient, our hypothesis is not supported by the results. Results suggest that increasing sales will lower the likelihood of having a tax haven subsidiary for Indian firms. But in literature, it is found that larger the sales, greater is the likelihood to own a subsidiary in tax haven countries. Our model explains that 1 unit increase in sales will result in odds of 0.9581 times decrease in the likelihood of owing a tax haven subsidiary.

The variable age of the firm is positively related but not significant (p = 0.4691) in our specifications. We do not find the effect of age of the firm on tax haven operations. So we rejected the hypothesis that age of the firm has a significant influence on the likelihood of establishing a tax haven subsidiary by an Indian company. Gibrat (1931) found that younger firms are more likely to grow faster than older firms. Pradhan (2004) found that age and size are important factors influencing Indian outward FDI. In the case of expansion to tax haven countries, the age of the firm does not have any significant influence on the likelihood of owning a tax haven subsidiary. The variable manufacturing/service dummy is significant at 1%. The results show that firms operating in the service sector prefer tax haven operations. The odds ratio is 1.565. The variable return on equity is not significant and negatively related with the likelihood of operating in tax haven subsidiary. Table 6 also contains pseudo R² value, which is the value indicating the explained variation. Pseudo R² in our model is 23.27%. It indicates that one unit change in all variables together will result in 23.27% changes in the likelihood of an Indian firm having a tax haven subsidiary. The overall model is fit since the LR statistic is significant at the 1% level of significance.

VI. Conclusion

The traditional FDI theories and empirical studies suggested that the firm’s characteristics were important in explaining overseas direct investment flow. Recently, India emerged as a source country for FDI and also a large amount of investment flowed to tax haven countries. We tested the influence of firm-specific factors on overseas direct operation of Indian firms in tax haven countries. This study has found that the firm-specific factors like intangible assets, long-term debt and more number of subsidiaries positively and significantly influence the likelihood of owning a subsidiary by the Indian multinationals in tax haven countries. The variable ‘firm size’ is found to discourage the likelihood of owning a tax haven subsidiary. The primary emphasis is given to whether and to what extent the intangible assets, long-term debt and technological and knowledge intensity influence the likelihood of owning a tax haven subsidiary. The findings have three dimensions. The first dimension shows that there is strong evidence that firms with high intangible assets, long-term debt and number of foreign subsidiaries positively affect the likelihood of tax haven operation. Second, the firms operating in high-technology and knowledge-intensive sectors are more likely to own a tax haven subsidiary in a tax haven country. Lastly, we find that there is weak evidence to support the experience (age of the firm). It is not significant in the specified model. In addition to that, the sales deter the likelihood of owning a tax haven subsidiary by Indian MNCs.

Implications, Limitations and Future Research Directions

The policy framers need to examine the problem of profit shifting to tax haven countries as an important issue to be addressed through tax reforms. The issue of overseas tax haven is a problem since many decades, but authorities fail to prevent tax haven activities. This research study helps to identify which firm-specific factors influence the use of tax haven countries in business operations. It may help to form the appropriate tax rules to prevent the tax haven countries to avoid corporate tax. Since the tax avoidance is a cause for revenue losses to the government, these findings have implication for tax policy.

One limitation of this article is that the study covers only firm-specific factors, but other factors like host-country-level advantages and internalisation advantages are ignored. Our sample is limited to companies listed in the BSE 500 index of Bombay Stock Exchange. Testing for a larger set of companies could give more robust results. The study uses only parent company data. To get deep evidence of intensity of tax profit shifting and internalisation advantages from tax haven operations, we need to examine the financial data of subsidiaries. Round tripping of overseas investment is another area of further research. Tax haven countries may encourage the firms to engage in more R&D activities to reduce the tax burden. It may cause an endogeneity problem and the need to conduct analysis with other statistical tools like generalized method of moments- instrumental variable (GMM-IV).

Appendix 1. Tax Haven Countries’ List

Macao, Andorra, Anguilla, Antigua and Barbuda, Aruba, Bahamas, Bermuda, Bahrain, British Virgin Islands, Barbados, Cayman Islands, Belize, Coock Islands, Cyprus, Dominica, Gibraltar, Guernsey, Sark and Alderney, Grenada, Isle of man, Jersey, Malta, Marshall Islands, Mauritius, Montserrat, Monaco, Netherlands, Nauru, Niue, Netherlands Antilles, Panama, Saint Kitts and Navis, Saint Vincent, Saint Lucia, Singapore, Hong Kong, Ireland, Switzerland, Luxembourg, Vanuatu, Turks and Caicos Islands, Samoa, Seychelles, Lebanon and Letchison.

Source: WWW.taxjustice.net/cms/upload/pdf/identifying_tax_havens_jul_pdf

Footnotes

Declaration of Conflicting Interests

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Vineeth P.

Note

References

Altshuler

, & Grubert

(2003). Repatriation taxes, repatriation strategies and multinational financial policy. Journal of Public Economics , 87(1), 73–107.

Barney

(1991). Firm resources and sustained competitive advantage. Journal of Management , 17(1), 99–120.

Berkhout

(2016). Tax battles: The dangerous global race to the bottom on corporate tax . Oxfam.

Bhayani

S. J.

(2010). Determinant of profitability in Indian cement industry: An economic analysis. South Asian Journal of Management , 17(4), 6–20.

Blouin

, & Krull

(2009). Bringing it home: A study of the incentives surrounding the repatriation of foreign earnings under the American jobs creation act of 2004. Journal of Accounting Research , 47(4), 1027–1059.

Brajcich

A. M.

, Friesner

D. L.

, & Schibik

T. J.

(2016). Do US pharmaceutical companies strategically shift income to international affiliates? Multinational Business Review , 24(1), 8–24.

Buckley

P. J.

, Clegg

L. J.

, Cross

A. R.

, Liu

, Voss

, & Zheng

(2009). The determinants of Chinese outward foreign direct investment. Journal of International Business Studies , 40(2), 353.

Buettner

, & Wamser

(2013). Internal debt and multinational profit shifting: Empirical evidence from firm-level panel data. National Tax Journal , 66(1), 63.

Care Ratings, 2014. (2014). Outward FDI investment by India . Economics Division of Credit Analysis & Research ltd. https://www.careratings.com

10.

Caves

R. E.

(1971). International corporations: The industrial economics of foreign investment. Economica , 38(149), 1–27.

11.

Chari

, & Acikgoz

(2016). What drives emerging economy firm acquisitions in tax havens? Journal of Business Research , 69(2), 664–671.

12.

Chen

T. J.

, & Ku

Y. H.

(2000). The effect of foreign direct investment on firm growth: The case of Taiwan’s manufacturers. Japan and the World Economy , 12(2), 153–172.

13.

Das

K. C.

, & Banik

(2015). What motivates Indian firms to invest abroad? International Journal of Commerce and Management , 14, 330–335.

14.

Desai

M. A.

, Foley

C. F.

, & Hines

J. R.

(2004). Economic effects of regional tax havens (Working Paper No. w10806). National Bureau of Economic Research.

15.

Desai

M. A.

, Foley

C. F.

, Hines

J. R.

Jr (2006). The demand for tax haven operations. Journal of Public Economics , 90(3), 513–531.

16.

Dharmapala

, Hines

J. R.

Jr (2009). Which countries become tax havens? Journal of Public Economics , 93(9–10), 1058–1068.

17.

Diamond

W. H.

, & Diamond

D. B.

2002. Tax havens of the world . Matthew Bender Books.

18.

Dorfmueller

(2003). Tax planning for US MNCs with EU holding companies: Goals, tools, barriers . Kluwer Law International.

19.

Dunning

J. H.

(1988). The theory of international production. The International Trade Journal , 3(1), 21–66.

20.

Dunning

J. H.

(1995). Reappraising the eclectic paradigm in an age of alliance capitalism. Journal of International Business Studies , 26(3), 461–491.

21.

Dunning

J. H.

(2000). The eclectic paradigm as an envelope for economic and business theories of MNE activity. International Business Review , 9(2), 163–190.

22.

Eichengreen

, & Irwin

D. A.

(1995). Trade blocs, currency blocs and the reorientation of world trade in the 1930s. Journal of International Economics , 38(1–2), 1–24.

23.

Eicke

(2009). Tax planning with holding companies-repatriation of US profits from Europe: Concepts, strategies, structures (Vol. 22). Kluwer Law International BV.

24.

Gaur

(2010). Financial performance measures of business group companies: A study of Indian non-metallic mineral products industries. IUP Journal of Business Strategy , 7(4), 45.

25.

Gibrat

(1931). Les InégalitésEconomiques . Librairie du Recueil Sire.

26.

Gitzmann

(2008). Legal personality of the corporate and international criminal law: Gobalisation, corporate human rights abuses and the Rorne statue. Queensland Law Student Review , 1(1), 36–54.

27.

Graham

J. R.

(2003). Taxes and corporate finance: A review. The Review of Financial Studies , 16(4), 1075–1129.

28.

Greiner

(1972). Evolution and revolution as organisations grow. Harvard Business Review , 76, 37–46.

29.

Grubert

, & Mutti

(2009). The effect of taxes on royalties and the migration of intangible assets abroad. In M. Reinsdorf, & M. Slaughter (Eds.), International trade in services and intangibles in the era of globalization (pp. 111–137). University of Chicago Press.

30.

Hishikawa

(2002). The death of tax havens. BC International and Comparative Law Review , 25, 389.

31.

Hong

, & Smart

(2009). In praise of tax havens: International tax planning and foreign direct investment. European Economic Review , 54(1), 82–95.

32.

Huizinga

, Laeven

, & Nicodeme

(2008). Capital structure and international debt shifting. Journal of Financial Economics , 88(1), 80–118.

33.

Hymer

(1976). The international operations of national firms . MIT Press.

34.

ICAI. (2018). International taxation and transfer pricing . The Institute of Cost Accountants of India.

35.

Ilaboya

O. J.

, & Ohiokha

I. F.

(2016). Firm age, size and profitability dynamics: A test of learning by doing and structural inertia hypotheses. Business and Management Research , 5(1), 29–39.

36.

Jaiswal

(2017). Foreign direct investment in India and role of tax havens . Centre for Budget and Governance Accountability. www.cbgaindia.org

37.

Jog

, & Tang

(2001). Tax reforms, debt shifting and tax revenues: Multinational corporations in Canada. International Tax and Public Finance , 8(1), 5–25.

38.

Jones

, & Temouri

(2016). The determinants of tax haven FDI. Journal of World Business , 51(2), 237–250.

39.

Jorissen

, Laveren

, Martens

, & Reheul

A. M.

(2005). Real versus sample-based differences in comparative family business research. Family Business Review , 18(3), 229–246.

40.

Kumar

(2008). Emerging MNCs: Trends, patterns, and determinants of outward FDI by Indian enterprises. In R. S. Rajan, R. Kumar & N. Virgill (Eds.), New dimensions of economic globalization: Surge of outward foreign direct investment from Asia (pp. 141–167). World Scientific Publishing.

41.

Mani

, Antia

K. D.

, & &Rindfleisch

(2007). Entry mode and equity level: A multilevel examination of foreign direct investment ownership structure. Strategic Management Journal , 28(8), 857–866.

42.

Mara

E. R.

(2015). Determinants of tax havens. Procedia Economics and Finance , 32(15), 1638–1646.

43.

Mintz

, & Smart

(2004). Income shifting, investment, and tax competition: Theory and evidence from provincial taxation in Canada. Journal of Public Economics , 88(6), 1149–1168.

44.

OECD. (2013). Addressing base erosion and profit shifting . OECD.

45.

Padilla-Perez

, & Nogueira

C. G.

(2016). Outward FDI from small developing economies: Firm level strategies and home-country effects. International Journal of Emerging Markets , 11(4), 693–714.

46.

Pandey

I. M.

(2015). Financial management (11th ed.). Vikas Publishing House Pvt Ltd.

47.

Pickering

M. A.

(1968). The company as a separate legal entity. The Modern Law Review , 31(5), 481–511.

48.

Pradhan

J. P.

(2004). The determinants of outward foreign direct investment: A firm-level analysis of Indian manufacturing. Oxford Development Studies, 32(4), 619–639.

49.

Pradhan

J. P.

(2017). Indian outward FDI: A review of recent developments. Transnational Corporations , 24(2), 43–70.

50.

Reserve Bank of India. (2019). Data on overseas investment . https://www.rbi.org.in/Scripts/BS_PressReleaseDisplay.aspx?prid=48391

51.

Saikia

, Das

K. C.

, & Borbora

(2020). What drives the boom in outward FDI from India? International Journal of Emerging Markets, 15(5), 899–922.

52.

Wamser

(2011). Foreign (in) direct investment and corporate taxation. Canadian Journal of Economics/Revue canadienned'économique , 44(4), 1497–1524.

53.

White

(1982). Maximum likelihood estimation of misspecified models. Econometrica: Journal of the Econometric Society , 1–25. https://doi.org/10.2307/1912526

54.

Vaidyanathan

(2017). Black money and tax havens (1st ed.). Westland Publications.