Incorporation in Offshore Financial Centers:

Naughty or Nice?

Warren Bailey and Edith X. Liu *

Cornell University

15th April 2014

Abstract

We examine how the choice of legal and regulatory environment affects a firm’s market

value and attractiveness to investors. Incorporation in an offshore financial center is

associated with lower Tobin’s q and initial public offering returns, and is more likely to

be selected by poorly-governed or low growth firms. However, offshore incorporated

firms are not shunned by US institutional investors. Moreover, firms from a few

countries appear to benefit from offshore incorporation, after controlling for firm-level

governance. Thus, the firm’s choice of legal domicile is not necessarily

value-destroying but must be assessed jointly with other firm and home country

characteristics.

Keywords: Offshore financial center, tax haven, law and finance, corporate governance,

regulatory arbitrage

JEL classifications: F21, F38, G15, G34

* Johnson Graduate School of Management, 387 Sage Hall, Ithaca NY 14850 U.S.A., (607)

255-4627, wbb1@cornell.edu; Dyson School of Applied Economics and Management, 466

Warren Hall, Ithaca, NY 14853 U.S.A., (607) 354-8308, edith.liu@cornell.edu. We thank

Andrew Karolyi, Denis Sosyura, Ying Wu, and seminar participants at Shanghai Advanced

Institute of Finance, University of International Business and Economics, 2013 McGill Global

Asset Management Conference, 2013 European FMA meetings, Queen Mary University

London, Chinese Academy of Sciences Institute of Systems Science, and Tsinghua University

PBCSF for helpful discussions.

1

1. Introduction

There is much evidence that the legal, regulatory, and disclosure environment affects firm

value and quality. For example, Daines (2001) finds that US firms incorporated in Delaware

have higher value and are more likely to experience takeover attempts than other firms.1

Doidge, Karolyi, and Stulz (2004) find that cross-listing of a non-US firm in the US is

associated with a substantially higher Tobin’s q relative to non-US firms that do not list in the

US.2

These findings suggest that the environment a firm selects can have substantial

consequences.3 A firm can choose an environment that enhances its value or signals its

intention to improve. Alternatively, a firm can choose a lax environment that is less expensive

and intrusive, but allows managers to expropriate ordinary shareholders (Fernandes, Lel, and

Miller, 2010; Doidge, Karolyi, and Stulz, 2010). Some jurisdictions attract incorporations with

a high quality value-enhancing legal regime while others offer weak law, regulation, and

disclosure that may allow management to benefit at the expense of other owners. Furthermore,

whether or not the incorporation decision enhances firm value depends on the legal

environment and characteristics of both the legal host country and the home country.

1 Bebchuk and Cohen (2003) and Subramanian (2004) find this is sensitive to the time period. 2 The choice of venue to incorporate can affect other corporate decisions and, thus, value. See, for example, the findings of Wald and Long (2007) that leverage decisions of US firms depend on which US state a manufacturer chooses to incorporate in. 3 See, for example, poison pill studies like Malatesta and Walking (1988) and Comment and Schwert (1995).

2

This paper evaluates the consequences of selecting an alternative legal and regulatory

system by incorporating in an offshore financial center (OFC). Jurisdictions ranging from

city-states to tiny island dependencies structure their legal and regulatory regimes to encourage

many financial activities including incorporation of foreign firms (Lane and Milesi-Ferretti,

2010). Places like Bermuda, the British Virgin Islands, and the Cayman Islands compete with

larger capital markets and can offer low-cost, efficient legal conditions that can enhance firm

value and pressure other jurisdictions to improve (Morriss, 2010). At the same time, OFCs

have been accused of enabling mismanagement and expropriation by self-serving corporate

managers or controlling owners, in addition to tax evasion, accounting opacity, and other

questionable or illegal behavior (Suss, Williams, and Mendis, 2002; Zoromé, 2007).4 While

the opposing views of OFCs have largely been discussed qualitatively, we offer new data and

empirical evidence with which to evaluate these arguments.

We study the factors that govern the choice to incorporate in an OFC and the impact of

OFC incorporation on firm value. For additional insight into the consequences of incorporation,

we measure investment decisions taken by professional investors concerning OFC-domiciled

firms. We focus on OFCs because they represent an opportunity to study an extreme form of

jurisdiction-shopping or regulatory arbitrage that has become increasingly controversial.5

4 See “At Orient Express, the Board holds All the Cards”, New York Times Dealbook, 29th November 2012,

(http://dealbook.nytimes.com/2012/11/29/at-orient-express-the-board-holds-all-the-cards/) for an example of

how managers can control a firm and their own employment conditions under Bermuda law. 5 For an example, see Houston, Lin, and Ma (2012) who report patterns in cross border banking flows that are

3

We find that incorporation in an OFC is associated with lower stock market value as

measured by Tobin’s q and returns on initial public offerings. Furthermore, OFC incorporation

is more likely to be chosen by firms with weak corporate governance as measured by the

fraction of insider ownership. This is consistent with the hypothesis that OFC incorporation

enables managers and insiders to exploit other shareholders, rather than enhancing value with

an efficient, low cost legal environment that offers advantages over incorporation in the US,

UK, or other large capital markets. We also find that the stock market penalizes firms with

relatively high capital expenditure if they are incorporated in an OFC, suggesting that the

potential expropriation of shareholders is particularly costly for growing firms. How offshore

incorporation affects shareholder value depends on the firm’s growth opportunities and the

extent to which it is controlled by insiders.

On the other hand, there is evidence that, after controlling for firm-level governance, OFC

incorporation by companies from a few countries enhances value. Furthermore, we find no

consistent evidence that OFC incorporated firms are shunned by US institutional investors. Our

evidence indicates that a variety of outcomes from competition and innovation in financial

market institutions is possible. Therefore, the impact of electing to incorporate in an OFC

depends on the characteristics of the individual firm and the environment in its home country.

consistent with regulatory arbitrage.

4

The balance of this paper is organized as follows. Section 2 provides details on offshore

financial centers and what it means to incorporate there. Section 3 presents a model, testable

hypotheses, and an overview of data and empirical methodology. Section 4 presents empirical

results while Section 5 summarizes and concludes.

2. Offshore Financial Centers

What exactly is an offshore financial center? Zoromé (2007) cites definitions from IMF

publications. For example:

“…’the banking system, acting as financial entrepôt, acquires substantial external

accounts beyond those associated with economic activity in the country concerned,’ or

countries where the ratio of deposit banks’ external assets to exports of goods and

services is significantly higher than the world average.”

“…a jurisdiction in which its international investment position assets, including as

resident all entities that have legal domicile in that jurisdiction, are close to or more

than 50 percent of GDP and in absolute terms more than $1 billion.”

The scale of activity in OFCs is difficult to measure as statistics are often sparse, but the

amount of financial activity in these places seems large. In Bermuda in 2003, for example

5

(Zoromé, 2007; Table 8), net export of financial services was 42% of GDP and portfolio assets

were over 100 times GDP. Jurisdictions with the 2003 ratio of net export of financial services to

GDP more than two standard deviations above the mean include Bahrain, Barbados, Bermuda,

Cayman Islands, Guernsey, Hong Kong, Isle of Man, Jersey, Latvia, Luxembourg, Mauritius,

Netherlands Antilles, Panama, Singapore, Switzerland, Uruguay, and Vanuatu. Offshore banks

hold about one-eighth of global bank assets. OFC activities can contribute significantly to

government revenue and employment in a small economy,6 diversifying away from traditional

sectors such as agriculture and tourism.

There are several major differences in corporate law and regulation across the three most

popular incorporation jurisdictions selected as legal domicile by listed companies.7 Bermuda

requires new companies to apply for approval, disclose beneficial ownership, register a

prospectus for securities issues, and make the registry of owners publicly available. There are

no such requirements in the British Virgin Islands or Cayman Islands. British Virgin Islands

companies must make their articles of association available publicly but Bermuda and Cayman

Islands companies need not. The British Virgin Islands also features minimal numbers of

6 In 2001 for the British Virgin Islands, for example, Suss, Williams, and Mendis (2002) report there were over

300,000 “international business corporations”, and 15% of local employment and 55% of government revenue

related to OFC functions. 7 See Conyers, Dill, and Pearman’s “Comparison of Laws in Bermuda, the Cayman Islands, and the British

Virgin Islands Relating to Offshore Companies” http://www.conyersdill.com/ and Bickley (2004).

6

directors and shareholders, management anonymity, and ease of transfer of corporate assets, in

addition to tax exemptions.8

Existing research on OFCs centers on corporate and personal taxation. Hines and Rice

(1994) study the use of “tax haven” affiliates by US corporations. They find that these

affiliates account for about 20% of all US foreign direct investment, and suggest that they are

motivated by low tax rates. Desai, Foley, and Hines (2006) detail the characteristics of US

multinationals that use OFCs and the tax advantages these firms enjoy. Dyreng and Lindsey

(2009) and Durnev, Li, and Magnan (2012) associate significant tax saving with use of OFC

subsidiaries by non-OFC firms.

In addition to tax benefits, Morriss (2010) notes that OFCs help US corporations lower the

cost of insurance and employee healthcare.9 At the same time, he mentions prominent events

in which OFCs are associated with local government corruption, the narcotics trade, financial

fraud, and tax evasion. There is continued concern about the uses and abuses of OFCs given

involvement in recent scandals (International Monetary Fund, 2006, 2008; van der Does de

Willebois, Halter, Harrison, Park, and Sharman, 2011; Shaxson, 2011) and continued easy

8 See http://www.bviifc.gov.vg/FinancialSectors/CorporateBusiness/Advantages/tabid/157/Default.aspx. For a

specific example, see the discussion of Michael Kors Holdings incorporating in the British Virgin Islands while

being headquartered in Hong Kong at

http://dealbook.nytimes.com/2011/12/20/the-benefits-of-incorporating-abroad-in-an-age-of-globalization/ . 9 An extreme example is hedge funds setting up offshore reinsurance units that invest back into the hedge fund.

See “With Lax Regulation, a Risky Industry Flourishes Offshore, in New York Times Dealbook, 5th September

2012, (http://dealbook.nytimes.com/2012/09/04/with-lax-regulation-a-risky-industry-flourishes-offshore/).

7

creation of shell companies and other offshore entities. While there is almost no academic

research on OFCs, what little exists suggests that OFC incorporation is associated with

poorer-quality firms (Durnev, Li, and Magnan, 2011) and weaker returns at merger and

acquisition events (Col and Errunza, 2013). Evidence on announcement effects and other

dimensions of valuation surrounding the incorporation event in OFCs is mixed.10 Adding to

the controversy, some OFCs were heavily involved in the US securitization boom which fed

the recent financial crisis (Lane and Milesi-Ferretti, 2010). OFCs certainly offer substantially

different conditions to foreign companies relative to their home environments.

Whether OFCs facilitate healthy competition between jurisdictions or a “race to the

bottom” continues to be debated.11 In the model of Bar-Gill, Barzuza, and Bebchuk (2006),

reincorporation improves the welfare of shareholders on some dimensions (such as Delaware’s

court system or an OFC’s tax system) at the cost of freeing insiders to expropriate more wealth.

Competition from other jurisdictions limits the revenue that an OFC can extract from firms.12

10 Desai and Hines (2002) find an average announcement response of 1.7% for 19 US firms moving

incorporation to an OFC from 1993 to 2002. Durnev, Li, and Magnan (2012) find that small positive

announcement effects for registering or establishing a subsidiary in an OFC vanish after 2002 while valuation is

lower except for UK and US firms establishing OFC subsidiaries. 11 For example, Hong Kong’s Jardine group of companies reorganized under a Bermuda-domiciled holding

company in 1984. Aspects of Bermuda law were thought to enhance the Keswick family’s control of the group,

which could protect the company once control of Hong Kong returned to China in 1997 or could increase the

power of insiders relative to minority shareholders. 12 For example, the maximum annual corporate franchise fee is about $30,000 in Bermuda and only a few

thousand dollars per year in the British Virgin Islands or the Cayman Islands (www.taxrates.cc). For comparison,

Delaware’s maximum annual fee is $180,000 (corp.delaware.gov/frtaxcalc.shtml). Public information on the

specific legal expenses of incorporation or re-incorporation of a substantial listed public company does not seem

8

Thus, the costs and benefits associated with a firm’s choice of jurisdiction for incorporation are

complex.

An anecdote illustrates the ambiguous potential impact of re-incorporation in an OFC.

Tyco International, a US listed corporation, moved its legal domicile from Bermuda to

Switzerland in 2009. The company’s 2009 10-K annual report states that the change in

incorporation will enhance “our ability as a Swiss company to take advantage of the tax

treaties between Switzerland and the United States”. On the same page, “there continues to be

negative publicity regarding, and criticism of, companies that conduct substantial business in

the U.S. but are domiciled abroad”. The document goes on to say “Because of differences

between Swiss law and Bermuda law and differences between the governing documents of

Swiss and Bermuda companies, it may not be possible to enforce court judgments obtained in

the United States…in Switzerland”. Thus, incorporation in an OFC can convey benefits such

as tax advantages or can harm shareholders by reducing their ability to control management

or damaging the company’s reputation.

A recent development is the increase in problems associated with OFC-incorporated

Chinese companies listed in US and other non-Chinese stock markets. The growth of China’s

available See, for example, http://www.calstrs.com/Newsroom/whats_new/cominghomeISS.aspx concerning

Tyco International. Tyco’s Form S-4 dated 10th December 2008 states (page 23) “…we do not expect these costs

to be material…” Consent of the shareholders and directors must be obtained. According to law firm websites, a

straightforward incorporation of a small firm would cost several thousand dollars in legal fees. See “The

incorporation business: They sell sea shells”, The Economist, 7th April 2012,

http://www.economist.com/node/21552197/print.

9

economy has outpaced legal, regulatory, and disclosure practices. Therefore, OFCs may

provide Chinese firms with a more secure and predictable legal system. Recently, however,

several Chinese corporations that listed in the US have suffered from management problems

and scandals (see, for example, Hung, Wong, and Zhang, 2011). Regulators explicitly

mention the entry of problematic Chinese companies to US stock markets using reverse

mergers (Public Company Accounting Oversight Board, 2011; United States Securities and

Exchange Commission, 2011). Furthermore, some of the OFC-related structures employed by

Chinese firms to list a holding company in the U.S. appear to violate Chinese law.13 Ang,

Jiang, and Wu (2012) explore the characteristics of US listed Chinese firms that have been

involved in scandals. They find that listing by reverse merger, greater earnings management,

and weaker corporate governance predict greater likelihood of scandal. Many of the recent

US listings from China are incorporated in offshore financial centers, particularly the

Cayman Islands.14

3. Empirical design

3.1 Testable hypotheses

13 See “China’s Forbidden Investment, 1st March 2012,

http://www.rkmc.com/publications/articles/china-s-forbidden-investment. 14 See Darrough,, Huang, and Zhao (2012), Jindra, Voetmann, and Walkling (2012), and Givoly, Hayn, and

Lourie (2012) for further evidence on Chinese firms listed in the US and, in particular, on the associations

between reverse listing and firm quality.

10

Doidge, Karolyi, and Stulz (2004) present a simple model to explain whether a non-US

firm chooses to list on a US stock market. In deciding whether to list shares in the US, a

non-US firm’s controlling shareholders face a trade-off. The firm can grow faster by raising

additional capital under the extensive investor protection of the US legal and regulatory system,

though this constrains the ability of controlling shareholders to expropriate minority

shareholders.

Similarly, we can imagine that incorporation in an offshore financial center moves a firm

to a less-demanding environment that offers potential expropriation benefits to insiders.

Therefore, we adapt the setting, notation, and solution of Doidge, Karolyi, and Stulz (2004) to

the case where investor protection afforded by incorporation at home, ρhome, can exceed the

investor protection when incorporated in an offshore financial center, ρofc. Controlling

shareholders are entitled to a fraction, k, of the firm’s total cash flow. The firm realizes cash

flow of C, plus potential additional cash flow from growth opportunities, z, and potential

additional cash flows, α, from benefits of OFC incorporation such as lower taxes, fewer legal,

regulatory, and administrative burdens, and other cost reductions mentioned earlier. The firm’s

insiders then selects a fraction, f, of firm cash flow that they expropriate beyond the amount

received from k times the total cash flow.15 Expropriation imposes a cost which is quadratic in

the fraction expropriated and linearly increasing in the quality of investor protection, ρ, in the

15 C is assumed exogenous. A more complex model would allow managerial decisions including legal domicile

to affect the distribution and value of C.

11

jurisdiction where the firm chooses to incorporate. Furthermore, if the firm is incorporated not

at home but in an OFC, it loses the opportunity to finance growth opportunities worth z but

gains the cost efficiencies and tax benefits worth α.

Thus, controlling shareholders enjoy the following gain if incorporated at home:

k[(C+z) – f (C+z) – ½ b f2 ρhome(C+z)] + f(C+z) (1)

where b is a constant greater than zero. In (1), the term in brackets is the cash flow of the firm

minus expropriation and minus its deadweight cost. The term, f(C+z), on the right is the benefit

from expropriation enjoyed by controlling shareholders. The gain to controlling shareholders if

incorporated in an offshore financial center is:

k[(C+ α) – f (C+ α) – ½ b f2 ρofc(C+ α)] + f(C+ α) (2)

Solve the maximization problem over f in both home and OFC incorporated situations and

substitute the solution into the original expressions to yield the total gain of the controlling

shareholders if incorporated at home:

k(C+z) + ½ [(1-k)2/bρhomek](C+z) (3)

and their total gain if incorporated in an OFC:

k(C+ α) + ½ [(1-k)2/bρofck](C+ α) (4)

Controlling shareholders will choose to incorporate in an OFC if their gain, (4) exceeds that for

incorporation at home, (3). Let ϴ equal the parameter ½ (1-k)2/bk:

k(C+z) + (ϴ /ρhome)(C+z) < k(C+ τ ) + (ϴ/ρofc)(C+α) (5a)

12

The left-hand side is the payoff from incorporating at home and financing growth opportunities.

The right-hand side is the payoff from greater expropriation by incorporation in an OFC.

Rearranging yields:

k(α -z) + ϴ[((C+τ) /ρofc) – ((C+z) /ρhome)] > 0 (5b)

The comparative statics in Doidge, Karolyi, and Stulz (2004) imply, first, that growth

opportunities dissuade firms from incorporating in an OFC, and, second, that expropriation

opportunities encourage firms to incorporate in an OFC. This is confirmed formally in (5b).

Controlling insiders will select incorporation in an OFC if growth opportunities, z, are low

relative to cost savings, α, and expropriation is much easier in an OFC (that is, 1/ρhome versus

1/ρofc).

We can also see how value differs from the point-of-view of minority outside shareholders

who do not enjoy any of the expropriation benefits of control. The firm’s cash flow to

controlling shareholders if incorporated at home, after they select the optimal f, is:

(C+z)(1 – [(ϴ/ρhome)(1+k)/(k(1-k)] (6a)

If incorporated in an OFC, the firm’s cash flow after controlling shareholders have selected the

optimal level of f is:

(C+ α)[1 – (ϴ/ρofc)(1+k)/(k(1-k))] (6b)

The results in Doidge, Karolyi, and Stulz (2004) imply that firms incorporated in an OFC sell at

a discount relative to firms incorporated at home when investor protection is greater at home,

13

and that discount is increasing in growth opportunities and the distance between ρhome and ρofc.

In our case, the premium for incorporating at home, (6a) minus (6b), is slightly more

complicated because there is a potential cost saving benefit to OFC incorporation in our model.

OFC incorporation is value destroying when a firm that incorporate in an OFC has lower value

than if it incorporates at home:

(C+z)(1 – [(ϴ/ρhome)(1+k)/(k(1-k)] > (C+ α)[1 – (ϴ/ρofc)(1+k)/(k(1-k))] (7)

In equation (7), if the potential cost savings of incorporating in OFC, α, is equal to the cash

flow increase from growth opportunities, z, then the only determinant of the value difference is

the investor protection in the OFC versus at home. In this case, when total cash flow is the same

for OFC incorporation as home incorporation, there is value destruction if and only if the

investor protection at home is stronger than in offshore, ρhome> ρofc. In contrast, if the difference

between home and OFC investor protection is small, equation (7) holds if the cash flow from

additional financed growth opportunities, z, is much larger than the potential cost savings, α,

that is offered by OFC incorporation.

In thinking about how incorporation in an OFC can affect value, it may also be the case

that OFC incorporation offers a simple, low-cost, effective legal environment that is

value-enhancing for firms:

(C+z)(1 – [(ϴ/ρhome)(1+k)/(k(1-k)] < (C+ α)[1 – (ϴ/ρofc)(1+k)/(k(1-k))] (8)

14

Equation (8) holds when either the investor protection of the OFC is actually stronger than the

investor protection of the home country, or if the cash flow benefit from cost savings from OFC

incorporation exceeds the expropriation costs and the cash flow from growth opportunities.

However, investor protection in OFCs is perceived to be weaker than investor protection in

typical developed home countries. In this case, value creation from OFC incorporation arises

from potentially large cost savings, α, that dominate the small differences in investor

protection.

With the theoretical findings of Doidge, Karolyi, and Stulz (2004) in mind, we offer the

following testable hypotheses. Some of the predictions compare OFC-incorporated firms to

control firms that share their home country. Other predictions imply consequences from a

change in incorporation to or from an OFC. We begin with a simple null:

H0: Incorporation in an OFC is irrelevant and, after controlling for firm characteristics,

there is no valuation difference between firms from non OFC countries that incorporate in

an OFC versus firms from a non OFC country that incorporate in their own country.

Next we compare OFC incorporated firms to non OFC incorporated firms under the

assumption that OFCs offer a weaker legal and regulatory environment:

15

H1: Incorporation in an OFC allows controlling shareholders to avoid home country legal

and regulatory discipline and expropriate more benefits from minority shareholders.

The first specific implication of H1 and our adaptation of the models of Doidge, Karolyi, and

Stulz (2004, 2007) concerns which firms choose to incorporate in an OFC:

H1a: Firms that choose to incorporate in an OFC have weaker governance and lower

growth opportunities than other firms.

A second detailed implication of H1 concerns firm value and OFC incorporation:

H1b: Firm value is lower for OFC incorporated firms. Firm value declines when a firm

re-incorporates in an OFC and increases when a firm re-incorporates away from an OFC.

We can also offer cross-sectional predictions based on firm and home country characteristics:

H1c: The effects predicted by H1a and H1b are heightened for firms located in high

quality environments that choose to incorporate in an OFC, for firms that incorporate in an

OFC with a particularly weak legal and regulatory regime

16

H1 and its related hypotheses are based on the idea that the OFC offers an undemanding

environment that weakens firm value and quality because ρofc is less than ρhome. In contrast, we

can imagine how incorporation in an OFC enhances value:

H2: Incorporation in an OFC offers a firm a more efficient legal and regulatory

environment that enhances firm value.

A firm operating in an underdeveloped legal and regulatory environment with can find that

high growth opportunities incorporation in an OFC enables raising funds externally. 16

Furthermore, moving incorporation from a demanding and intrusive home environment to a

lower-quality OFC can lower the cost of compliance, thereby enhancing the raw cash flow, C,

available to shareholders.17

As was the case for H1, H2 implies additional detailed hypotheses. In brief, H2a states

that OFC incorporation will be selected by firms with the most to gain from a more cost

efficient environment, both in terms of tax savings and lower regulatory requirements. H2b

follows: firms that enhance their cash flow by incorporating in an OFC have higher value.

16 For example, Allen, Qian, and Qian (2005) and Bailey, Huang, and Yang (2011) describe the reliance on

retained earnings that results from limited opportunities to raise funds externally.in China’s capital market. 17 See, for example, Doidge, Karolyi, and Stulz (2010) on foreign firms terminating their SEC registration when

faced with the cost of complying with the Sarbanes-Oxley Act.

17

H2c follows: value-enhancement from OFC incorporation will be particularly large for firms

with poor home country environment or substantial growth opportunities.

Finally, we view the irrelevance (H0), value destruction (H1), and value creation (H2)

theories of OFC incorporation through the lens of the investment decisions of professional

investors:

H3: The signs of the valuation effects of OFC incorporation predicted by H0, H1, or H2

are mirrored in patterns in the holdings of institutional investors.

3.2 Data

3.2.1 Firms and firm characteristics

We identify firms to include in the study as follows. First, we aggregate firms from

Datastream, Worldscope, ADR, and CompustatNA. The resulting firm names include the

current constituent list of all active, dead, and suspended firms from Datastream, all active

ADRs from adr.com and adrbnymellon.com, all firms on CompactD Worldscope CD-ROMs

from January 1992 to July 2006, and all firms on Compustat North America from January 1980

to January 2012. Country of incorporation is the country code (the first two digits) of the ISIN

identifier, except for Compustat NA where we use the “fic” variable. For ADRs, we use the

ISIN code of the underlying firm. To identify the address country, the variables “Nation” in

18

Datastream, “loc” in Compustat North America, and “Country” in ADRs and Worldscope

CompactD are used.

Given each firm is now associated with a country of incorporation and a country of

address, we next identify those firms with country of incorporation among the OFC countries

listed in International Monetary Fund (2006). This yields a total of 7,990 firms. We then

eliminate firms that have the same country of address and country of incorporation. This

excludes local OFC companies, leaving 1,372 distinct firms.18 We also identify candidates for

our control sample. We begin with all firms in Datastream with the same countries of address

as the previously-identified OFC-incorporated firms, then eliminate firms for which country of

incorporation differs from address country. This yields a total of 15,816 non-OFC firms. Next,

each of the OFC and non-OFC firms is matched with its financial variables. We obtain annual

firm-level financial variables Total Assets, Sales, Long Term Debt, Total Debt, Shareholder’s

Equity, and Market Value of Equity from Datastream and Compustat North America for

January 1980 to December 2011.19 Firms with missing financial series for any of the above

variables are discarded.

We gather or construct firm-specific characteristics as follows. For each OFC and

control firm, we construct annual balance sheet and income statement variables such as Total

19 Firms are matched by isinno in Datastream and gvkey in CompustatNA. Each firm-year observation includes ,

the spot foreign exchange rate for the currency as of the fiscal year end date so financial variables can be expressed

in US dollars.

19

Assets and capital expenditure (capex) in US dollars. Tobin’s q is computed for each firm-year

as (Market Value of Equity + Total Assets – Shareholder’s Equity) divided by Total Assets. We

also collect Insider Holdings (that is, the fraction of shares held by insiders) for both OFC and

potential control firms through Datastream. From SDC, we measure the extent to which each

firm uses capital markets to raise debt and equity and to obtain IPO characteristics. From

Thompson Reuters 13F mutual fund data, we gather information on holdings of each firm by

US institutional investors.

3.2.2 Country characteristics

We collect standard proxies for the legal and investment environment in the address

countries of our OFC-incorporated and control firms. They can be thought of as specific

dimensions of the environment faced by corporations in a particular country or as general

indicators of the overall quality of that environment. Following from Doidge, Karolyi, and

Stulz (2004, 2007), Bailey, Karolyi, and Salva (2006), and other authors, we begin with the

index of anti-director rights from La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1998).

The anti-director rights measure does not completely serve our purpose because many of our

most interesting and controversial firms come from the People’s Republic of China. The La

Porta et al (1998) indexes do not include this country given its complex and continuing

evolution away from a centrally-planned command economy. Allen, Qian, and Qian (2005)

20

discuss several dimensions of China’s legal and regulatory environment and compare to

countries that are included in the La Porta et al (1998) indexes. For the anti-director rights

index, Allen, Qian, and Qian (2005) estimate an index value of 3 for China while Pistor, Raiser,

Gelfer (2000) estimate an index value of 3.45 for Russia. We use these values into tests that

employ the anti-director rights index.

We collect additional country characteristics as follows. First, Djankov, LaPorta,

Lopez-de-Silanes, and Shleifer (2003) measure the time, in days, to complete two basic legal

procedures (collecting on a bounced check, evicting a tenant for nonpayment of rent) across

dozens of countries including China and many of our OFCs. They reflect the quality and

efficiency of the workings of a country’s legal system. Given that these two series are highly

correlated (ρ=0.4531), we select one, the eviction measure. Second, the World Bank’s ‘‘Doing

Business’’ indicators (DB) for disclosure, legal rights, credit information, director liability, and

shareholder suits (www.doingbusiness.org) measure other facets of the quality of the business

environment in a country annually starting with 2004. Given the often high correlations among

these measures, the eviction measure, and the anti-director rights index, we select one, the

disclosure quality indicator.

While these measures hold generally for developed nations, they may be poor proxy for

more complex and less transparent regulatory regimes, such as China. For example, Djankov et

al (2003) reports values for the time it takes to collect on a bad check or to evict a tenant for

21

non-payment of rent, and their tables include China. However, checks are not as widely used in

China as in other economies and evictions do not often involve formal legal proceedings.

Moreover, the environment perceived by investors and the measures tabulated by these sources

can depend on whether the investor is a local or a foreigner.

3.3 Methodology

We test our hypotheses from Section 3.1, we begin with several specifications as follows.

We regress Tobin’s q or the one-year IPO return20 on firm and home country characteristics

and an OFC dummy variable:

qit = α0 + α1Dofc,it + β’xit + δ’cit + εit (9)

Dofc,it is the OFC dummy which equals one if firm i is incorporated in an OFC in year t. xit is a

vector of firm characteristics, and cit is a vector of country characteristics. The observations

are firm-years for the two types of firms. Sample firms are those located in a non-OFC country

but incorporated in an OFC country, such as a US firm incorporated in the Cayman Islands. In

addition, our OFC sample also includes a number of firms located in an OFC country but

incorporated in a different OFC country. Control firms are both located and incorporated in non

20 See, for example, Welch (1991) for a cross sectional regression approach.

22

OFC countries. Firms which are both located and incorporated in the same OFC are excluded.

This excludes the relatively large number of firms that are both located and incorporated in

Hong Kong, Ireland, Singapore, and Switzerland. Regressions intended to explain IPO returns

also include a few characteristics of offerings.

An extension of (9) is a two-stage Heckit procedure to control for a specific self-selection

issue. The classic “treatment effects” problem for (9) is that an apparent relationship between

individual firm Tobin’s q and incorporation in an OFC can induced by the decision about where

to incorporate, rather than from OFC incorporation itself. The effect is akin to that of an

omitted variable in (9) if the decision to incorporate, Dofc,it, is a function of firm characteristics,

xit, and country characteristics, cit. Given an omitted variable, the error terms in (9) and the

selection equation (which explains the unobserved value driving Dofc,it) are correlated.

Therefore, we estimate a two-stage procedure following Doidge, Karolyi, and Stulz (2004) and

Bailey, Karolyi, and Salva (2006). The first stage is a Probit regression to identify factors that

predict whether a firm chooses to incorporate in an OFC. The dependent variable is the OFC

dummy. The independent variables are country and firm characteristics. The second stage

regresses firm-specific Tobin’s q on the OFC dummy variable and country and firm

characteristics as in (9), plus the inverse Mills ratio from the first stage.21 Another single-stage

21 We run the Heckit for Tobin’s q yearly, then report average coefficients: unlike Doidge, Karolyi, and Stulz

(2004) who use data for one year, 1997, we have data from a number of years. We cannot do this for IPO

returns due to a much smaller sample size and, therefore, we do not estimate the Heckit for IPO return.

23

variation on (9) adds interaction terms between the explanatory variables and the OFC dummy.

This decomposes the OFC valuation discount or premium measured by the slope, α1, on the

OFC dummy in (9).

In addition, we find a few dozen firms which experienced an OFC reincorporation event,

that is, either moving incorporation to or from an OFC. We conduct a simple event study on

Tobin’s q before versus after the reincorporation. Finally, we study holdings of US institutional

investors using a specification similar to (9) to determine whether professional investors

distinguish OFC and non-OFC incorporated firms in a manner related to our hypotheses.

4. Empirical results and discussion

4.1 An overview of the data

Table 1 tabulates country of incorporation, country of address, and listing exchange for

the OFC-incorporated firms we have identified. Panel A indicates a total of 1372 distinct

OFC-incorporated firms. It is evident that country of incorporation is heavily concentrated in

two jurisdictions, Bermuda and the Cayman Islands, which account for over 75% of the sample

firms. The British Virgin Islands is a distant third, with less than 5% of the OFC sample firms.

Among the home countries for these OFC incorporated firms, Hong Kong firms account for

more than half of the OFC sample, Chinese firms about a fifth, and UK and USA firms each

less than 10% of other OFC sample firms. Most Hong Kong address firms are incorporated in

Bermuda while most China address firms are incorporated in the Cayman Islands. Most UK

24

address firms incorporate in the nearby British Crown Dependencies of Jersey, Isle of Man, and

Guernsey. Interestingly, the majority of OFC firms incorporated in the Marshall Islands are

Greek shipping firms.22 These findings suggest that certain jurisdictions cater to the needs of

firms from specific home countries or perhaps even specialize in particular clients or

industries.

Panels B and C of Table 1 summarize the exchanges where OFC-incorporated firms are

listed. We note several prominent patterns in these two panels. First, listing on formal US

exchanges (Panel B) is unusual except for Chinese firms incorporated in the Cayman Islands.

Panel C indicates several hundred firms headquartered in Hong Kong but incorporated in

Bermuda or the Cayman Islands are listed in Hong Kong. Most unexpected is the huge number

of Hong Kong and China address firms which are OFC incorporated and listed on Germany’s

Berlin or Frankfurt stock exchanges.

While Table 1 presents a geographical summary of the firms we have identified, our

goal is to examine associations between financial performance and the firm’s choice of legal

domicile. Matching financial information to the firms yields missing observations that reduce

the number of available firms. Because firms can have missing observations for some variables

and years, firm-year observations with missing Total Assets, Shareholder’s Equity, and Market

Value are excluded as they are necessary to compute Tobin’s q. Furthermore, we eliminate any

22 The Marshall Islands offers special provisions for the incorporation of maritime shipping companies.

25

firm with Total Assets less than 10 million US dollars to avoid both very small firms and large

outliers when computing Tobin’s q. Finally, we remove resulting extreme observations of

Tobin’s q by dropping the top and bottom 1%, which include negative Tobin’s q. This yields a

total of 918 firms for the OFC incorporated sample and 7876 firms for the control sample.

Table 2 summarizes the characteristics of the firms used in subsequent empirical tests.

Panel A reports averages of Tobin’s q broken down by country of incorporation and country of

address. Substantial differences across the rows and columns are evident. Hong Kong address

firms that incorporate in an OFC have higher q than control firms, except for those that

incorporate in the British Virgin Islands, which display an extremely low average q of 0.4770.

In contrast, OFC incorporated Chinese address firms typically have lower q than the average of

2.8680 for control Chinese firms, with a particularly low average, 0.4543, for firms

incorporated in the British Virgin Islands. Low q for UK and US address firms incorporated in

the British Virgin Islands are also evident, though the contrast is not as great as for Hong Kong

and China address firms and the numbers of UK and US firms are much lower than for Hong

Kong and China. The gap between OFC sample and control q is also much less stark for UK

and US firms.23

Panel B of Table 2 summarizes basic descriptive financial variables to better understand

the firms in our two samples. Relative to control firms, OFC incorporated firms have lower

23 The Appendix includes Supplement to Table 2 which tables summary statistics on other country characteristic

26

average 3-year sales growth. This is consistent with H1a which predicts that OFC

incorporation is more likely for firms with lower growth opportunities. Firm size as measured

by log of USD sales is similar across control and OFC incorporated groups, which combined

with the finding of lower growth opportunities suggests “cash cows” are more likely to

incorporate in an OFC. Leverage is smaller for Bermuda, British Virgin Islands, and Cayman

Islands incorporated firms relative to control firms, but larger for firms incorporated in other

OFCs.

Table 2 also allows us to understand some of the structure of missing observations among

the firm-years of data. For example, Panel A shows 437 Bermuda-incorporated firms that have

at least partial data for Sales Growth, Log Sales, and Debt to Equity while Panel B shows the

number of non-missing observations for these variables range from 2908 for 3-year Sales

Growth to 3811 for Debt to Equity. However, the pattern of missing observations differs across

series. While the hypothetical maximum number of firm-year observations is 83,081 (the sum

of the bottom row of Panel B), the number of firm-years with complete data on these three

characteristics is only 69,159.

4.2 OFC incorporation and Tobin’s q

This section report results of cross section and time series regressions following Doidge,

Karolyi, and Stulz (2004) and Bailey, Karolyi, and Salva (2006). Our dependent variables are

27

Tobin’s q and, and for those firms that went public during our sample period, IPO returns.

Independent variables are characteristics of the firm and of its home country.24

Table 3 presents estimates of the single-stage regression, Equation (9), to explain firm

Tobin’s q. In Panel A, specification (1) includes only a constant and the OFC dummy, and we

find a strong negative coefficient of -0.0948 (t-statistic = -11.64) on the OFC dummy but a very

small r-squared. This implies that firms incorporated in an OFC have a significantly lower

Tobin’s q relative to counterparts incorporated domestically. In particular, OFC incorporation

is associated with about 9.5% lower valuation. This is consistent with H1, which predicts that

OFC incorporation facilitates managerial expropriation and reduces firm value. When we add

year and industry fixed effects, specification (2), the coefficient on the OFC dummy becomes

even more negative and statistically significant, -0.1241 (t-statistic = -15.39), while the

r-squared rises to over 11%. Next we add firm and country explanatory variables.

Specification (3) adds country characteristics. The anti-director rights and eviction measures

have negative and highly significant slope coefficients while the disclosure index has a

significantly positive slope. This indicates that Tobin’s q is typically higher for firms from

countries with low investor protection, an efficient legal system, or high quality corporate

disclosure. The OFC dummy remains significantly negative (-0.1303 t=-15.51) with the

24 Although Sales Growth is one of several possible proxies for growth opportunities (H1a), we exclude the

variable from regressions due to extensive missing observations which disproportionately affect the OFC

sample.

28

inclusion of the three country characteristics. The negative slope on AD Index is anomalous.

Note, however, that there is considerable multicollinearity among the country characteristics.

For example, the anti-director rights index has a correlation of -37% with the speed of evicting

a tenant and 51% with the World Bank disclosure index.

Specifications (4) and (5) separately add two proxies for the firm’ growth opportunities,

capital expenditure and capital raising, plus the fraction of shares held by insiders. Capital

expenditure and capital raising both have strongly significantly positive slope coefficients,

which is consistent with the intuition behind Tobin’s q and our interpretation of these variables

as proxies for growth opportunities. With capital expenditure included, (4) retains a

significantly negative slope on OFC dummy (-0.0536, t=-5.62). When replaced by capital

raised, (5) shows a reduction in the significance of the slope on OFC dummy (-0.0219, t=-1.90)

but this estimate is based on a substantially reduced number of observations due to missing

values for capital raised. The slope coefficient on Insider Holdings is significantly negative,

ranging from -0.1880 (t=-16.15) to -0.1943 (t=-12.04) across the two specifications. This

indicates that a one percent increase in insider holdings is associated with valuation that is

almost 1/5 percent lower.

A concern with these regressions is the possibility that the very large Tobin’s q (2.8680 in

Panel A of Table 2) of Chinese control firms tilt the regression estimates given the large number

of control firms (419 of 7876) from China. Therefore, specifications (6) and (7) use only China

29

address firms (including Hong Kong firms, which are often addressed in Hong Kong but

operate and are controlled in China). Comparing the specifications, we see that the slope on

OFC dummy flips from strongly negative (-0.6295, t=-24.34) to strongly positive (0.1079,

t=3.48) when capital expenditure and insider holdings are included. Slopes on capital

expenditure and insider holdings are positive and negative respectively, as is the case in the

corresponding specification, (4), using all firms. This evidence is consistent with H2: after

controlling for other characteristics, firms from a less developed environment appear to benefit

from incorporation in an OFC.

Panel B of Table 3 presents additional specifications to test whether high-level

cross-listing in the U.S. stock market affects the relationship between value and OFC

incorporation. Interestingly, the slope coefficient on the DR dummy in specifications (8), (9),

and (10) comes in significantly negative, ranging from almost minus eight percent to almost

minus ten percent. The size and strength of the negative slope on the OFC dummy is reduced or,

in the case of specification (10), vanishes entirely. Specifications (11) and (12) confine the

test to China and Hong Kong firms only. The coefficient on DR dummy becomes many times

more negative and more statistically significant. Based on the estimate (12), cross-listed

China and Hong Kong firms have Tobin’s q on average 66.68% lower than other firms. At the

same time, we see that, with insider holdings included (specification 12) the slope on OFC

dummy becomes significantly positive (0.1151, t=3.74), echoing specification (7) in Panel A.

30

Unreported results which exclude China and Hong Kong address firms find that the negative

slope on DR dummy becomes statistically insignificant. Thus, there is something unusual

about Chinese firms that cross-list in the US, particularly those that incorporate in OFCs.

Most interestingly, it appears that OFC incorporation adds value for China and Hong Kong

firms once other characteristics are controlled for.

While Table (3) shows substantially lower Tobin’s q for firms incorporated in OFCs, these

single equation pooled regressions do not account for potential self-selection bias: firms

choose whether or not to incorporate in an OFC. To correct for this, we estimate Heckit two

stage regressions as previously described. The first-stage probit describes the decision to

incorporate off-shore while the second stage includes the inverse Mills Ratio to control for

self-selection and measure the additional effect of OFC incorporation on Tobin’s q.

Highlights of the estimates presented in Table 4 are as follows. The estimates of the probit

equation show that firms with growth opportunities as measured by capital expenditure are

significantly more likely to incorporate in an OFC. This is not consistent with our predictions,

H1 and H1a, that firms with growth opportunities will avoid OFC incorporation in order to

maintain their ability to raise capital and exploit those growth opportunities. By contrast,

firms with high insider holdings, a governance proxy, are more likely to incorporate in an OFC.

This is consistent with H1. Finally, firms from countries that score high on investor protection

(anti-director rights index) are more likely to incorporate in an OFC. Thus, the probit estimates

31

show evidence of both expropriation (H1) and value-enhancing (H2) motives for OFC

incorporation. The probit regressions are based on an average of 1400 observations per year

and return an average pseudo r-squared in excess of 20 percent.

The probit supports the second-stage Heckit specification which also appears in Table 4.

The results of the second pass are qualitatively similar though statistically weaker than those of

the single-stage regression result presented in Table 3. In particular, the slope on the OFC

dummy remains negative and increases in size relative to Table 3 but is only marginally

statistically significantly (-0.2274, t=-1.80). The slope on the inverse Mill’s ratio is only

marginally statistically significant (-0.8910, t=-1.69), suggesting that the impact of selection

bias in the single-stage regressions of Table 3 is not particularly strong.

Table 5 returns to the single-stage pooled regression specification for Tobin’s q and

further decomposes the significantly negative slope on OFC dummy found in Table 3. The

slope on OFC dummy becomes significantly positive, with a value about 0.65 and t-statistic of

7 or 8 for specifications (1) and (2) that do not include DR dummy. Slopes on interactive terms

indicate that higher firm capex, home country anti-director index, or home country legal

inefficiency (days to eviction) are associated with lower slope on OFC dummy. Positive slopes

on OFC dummy indicate that, in the absence of other factors, OFC incorporation alone is not

value destroying as suggested by H1. The lower valuation of OFC firms reported in earlier

tables is due to OFC incorporated firms having large negative sensitivities to country and firm

32

level characteristics. For example, the negative coefficient on the capex interaction term

indicates that OFC incorporated firms experience a significant valuation penalty if capex is

large, relative to firms incorporated at home. Negative slopes on the interactive term for insider

holdings indicate that, if governance problems are relative large, OFC incorporation is indeed

value-destroying. The interactive coefficients for anti-director rights and eviction indicate

that OFC incorporation destroys value for firms that originate in high quality home

environments. These effects are also consistent with OFC incorporation enhancing value for

firms from weak home country environments.

Table 5 also includes two specifications, (3) and (4), with the US cross-listing dummy

(DR). Echoing the findings of Table 3, one of the two slopes on DR dummy is significantly

negative. The interactive term for DR times OFC dummy is negative and strongly significant,

ranging from -0.1909 (t=-3.82) to -0.2820 (-4.73). This suggests that the combination of US

cross-listing and OFC incorporation is associated with particularly low valuation, probably due

to the Chinese firms as detected in Table 3. In particular, valuation is twenty to thirty percent

lower for firms that select both OFC incorporation and high-level US cross-listing.

Our investigation of the impact of OFC incorporation on value can also make use of

events when a firm moves its legal domicile to or from an OFC. Therefore, we construct an

additional sample of firms that switched incorporation from their country of address to an

offshore financial center. The number of such events is small, slightly more than 100, so we

33

relegate this table to the appendix and only discuss its findings here briefly. There is enormous

variation in Tobin’s q around these events, indicating high volatility in both market and book

values from year to year. If we include only firms with positive book value of equity,25 mean

Tobin’s q for all reincorporating firms in OFC jurisdictions decline from 2.65 two years before

the reincorporation to 1.25 two years after. This is consistent with OFC incorporation

facilitating expropriation and destroying value, H1. We also decompose the summary statistics

for reincorporation by the three most popular OFC jurisdictions. It is difficult to draw

conclusions given the small number of observations in each, but there is no clear evidence that

reincorporation in an OFC is associated with increasing or decreasing value as measured by

Tobin’s q.

4.3. OFC incorporation and IPO returns

We also view the value destruction (H1), value creation (H2), or irrelevance (H0) of

incorporation in an OFC through the lens of initial public offering underpricing. Like Tobin’s q,

IPO returns represent investor’s consensus valuation and allow us to compare firms that

incorporate in OFCs and firms that incorporate at home. In particular, we look at percent

returns for IPOs to evaluate the success of the capital raising event. If managers and

controlling shareholders have private information about a firm’s quality, large and positive one

25 See Brown, Lajbcygier, and Li (2008) for the difficulties that negative book value of equity firms pose for

researchers. The table indicates a small number of such firm-years in our sample.

34

year IPO returns show that, as time passes and more information about the firm is revealed,

investors increase their valuation of a high quality firm. Furthermore, in some models, a high

quality firm signals its prospects to uninformed investors by offering shares at a discount, the

cost of which is expected to be recovered when the firm’s prospects are revealed and

subsequent seasoned equity offerings sell at higher valuations.26

Table 6 presents summary statistics for IPOs of our OFC and control sample firms that

went public during our sample period. Panel A shows that, across all IPOs (both OFC

incorporated firms and control firms), the median one year IPO return is 10.61% with a 1% to

99% range of -87.56% to 372.92%. The median amount raised is about 33 million US dollars

and management and underwriting fees typically sum to less than three percent. We compare

these statistics to those for IPOs of firms incorporated in the three most popular OFCs. For 35

Bermuda incorporated IPOs, the median one year IPO return is -10.71%, though the median

amount and management and underwriting fees are comparable to those of the entire sample.

For 11 British Virgin Islands incorporated IPOs, the median one year IPO return is a startling

-40.60%, the median amount raised is also about 30 million dollars, but management and

underwriting fees are typically about 75 basis points higher than those of the entire sample. For

87 IPOs of Cayman Islands incorporated firms, the median one year IPO return is -5.79%, the

median amount raised is also about 37 million dollars, but management and underwriting fees

26 See, for example, Allen and Faulhaber (1989), Welch (1989), and Grinblatt and Hwang (1989).

35

are typically lower than those of the entire sample. While these results are based on only a

small sample of IPOs and many IPO offering characteristics are missing, they suggest that

OFC-incorporated IPOs are significantly value-destroying, in contrast to typical IPOs. This is

consistent with the expropriation hypothesis, H1.

Given the general interest and controversy surrounding Chinese firms and the concern

that our statistics may mask findings specific to those firms, we present separate statistics for

China and Hong Kong address firms in Panel B of Table 6. As a subset of the results of Panel A,

these statistics are based on fewer observations, 18 from Bermuda, 5 from the British Virgin

Islands, and 82 from the Cayman Islands. The median one-year IPO return across all China and

Hong Kong address IPOs is a respectable 23.43%, and the median amount raised is about 55

million dollars. However, median one-year IPO returns are much poorer for China and Hong

Kong address firms incorporated in principal OFCs: 4.62% for Bermuda, -40.6% for British

Virgin Islands, and -0.68% for Cayman Islands. Again, these statistics are based on small

sample sizes but they suggest that new equity offerings with OFC domicile can be noticeably

poor performers.

Table 7 presents a more formal pooled regression comparison of one-year OFC and

non OFC incorporated IPO returns. The findings for the IPO returns broadly parallel those for

Tobin’s q. Specifications with OFC dummy only, OFC dummy and fixed effects, and OFC

dummy with firm characteristics, country characteristics, and fixed effects show statistically

36

significantly negative slopes on OFC dummy. Specification (4), for example, indicates that

OFC incorporation is associated with a -30.04% (t=-2.13) lower IPO return than other offerings.

The negative slope becomes insignificant in Specification (6), which is based on far fewer

observations so as to include a measure of offering management fees. Specifications (7) and (8)

are based on IPOs of China and Hong Kong address firms only. They confirm substantially

lower (minus 14% to minus 29%) one-year IPO returns for firms that choose OFC

incorporation.

On balance, the results for one-year IPO returns are more uniformly negative on OFC

incorporation than the earlier findings for Tobin’s q. While it may be the case that some firms

(from weak home environments or with relatively few governance problems) can benefit from

incorporating in an OFC, the findings for firms going public are consistent with value

destruction, H1.

4.5. OFC incorporation and US institutional investor holdings

Our tests to this point have concentrated on observed measures of corporate value, Tobin’s

q and one-year IPO returns, and their relation to incorporation in an offshore financial center.

We can think of the stock market information underlying Tobin’s q and the IPO returns as

reflecting the consensus judgment of stock markets on OFC incorporation and other firm

characteristics. For a different angle on how investors assess the OFC incorporation decision,

37

we next examine differences in US institutional investor holdings of OFC versus non-OFC

incorporated firms.

Table 8 presents estimates of pooled regressions along the lines of equation (9) and our

previously reported and described regressions to explain Tobin’s q and IPO return. In contrast

to findings for the valuation measures, there is no clear pattern in the sign and significance of

the slope on OFC dummy for US institutional holdings: in Panel A, the slope ranges from

significantly negative in the specification with OFC dummy and fixed effects only to

significantly positive in the specification that includes capital expenditure and country

characteristics. In specifications (4) and (5), it appears that US institutions shun firms with high

insider holders, favor firms from high quality home countries as measured by the anti-director

rights index, and dislike firms from countries that score poorly on disclosure quality. As with

some of our Tobin’s q findings, this suggests that incorporation in an OFC is not, by itself,

necessarily value-destroying but needs to be assessed in light of other firm and home country

characteristics. Panel B presents additional specifications that include the US cross-listing

dummy variable. The slope coefficient on the ADR dummy is significantly positive in all three

specifications. This indicates that US institutions prefer cross-listed firms.

The regressions reported in Table 8 use total institution holdings from the 13F data, which

aggregates different types of institutions such as insurance companies, banks, pension funds,

and mutual funds. Aggregate institutional holdings can mask the differences in behavior

38

towards OFC firms by different types of institutions. To measure how different types of

institutional investors respond to OFC firms, Table 9 uses the classifications of institution type,

investor horizon, and investment style from Bushee (1998, 2001).27 Each institution in the 13f

database is associated with the appropriate classification. For each stock-institution pair, we

compute the number of years the institution holds the stock and the average holding (as a

percentage of the number of shares outstanding) for the years when the institution had a

position in the stock. This yields years held and average percentage holding for each

institution-stock pair.

Next, we average the number of years institutions held the ith stock across all institutions

that held it. This is then averaged across stocks, separately for OFC incorporated stocks,

non-OFC incorporated stocks, and various categories of institutions. For example, the first

entry in Panel A of Table 9 indicates the average holding period, 6.19 years, across all OFC

incorporated stocks held by institutions classified as banks. Other entries in that row indicate

that bank holding periods for non-OFC incorporated stocks are slightly higher, 6.31 years, but

that this is not significantly different from zero (t-test is -1.18).

27 Institution types are Banks, Corporate or Private Pension Funds, Independent Investment Advisor or Mutual

Funds, Insurance Company, Public Pension Funds, and University Endowment. Investment styles are

Quasi-index, Dedicated or Long Term Investors, and Transient or short term investors. From Bushee (2001):

“Dedicated institutions are characterized by large average investment in portfolio firms and extremely low

turnover, consistent with a ‘relationship investing’ role and a commitment to provide long term capital” while

“Quasi-Indexers are also characterized by low turnover, but…diversified holdings consistent with a passive,

buy-hold of investing portfolio funds in a broad set of firms”. Finally, the mutual fund types are Large Growth,

Large Value, Small Growth, and Small Value.

39

Finally, we compute the average percent institutional holding of the ith stock across all

institutions that held it, for each institution type category. We do this for all OFC incorporated

stocks, and separately for non-OFC incorporated stocks. In the first row of Panel A of Table 9,

for example, the average institution’s holding of an OFC-incorporated stock is 0.22% of a

stock’s outstanding shares which is not statistically distinguishable from the average holding,

0.38%, of non-OFC stocks.

Panel A of Table 9 presents statistics by institution type. For the average number of years

invested in OFC stocks versus non-OFC stocks, there are two particular types of institutions

that stand out. First, public pension funds hold non-OFC stocks for an average of 7.57 years,

which is significantly longer than for OFC stocks (6.27 years). Second, independent

investment advisors or mutual funds typical hold non-OFC stocks significantly shorter (4.54

years) than OFC stocks (4.68 years). Independent investment advisors or mutual funds tend to

hold both OFC and non-OFC firms for a shorter time than public pension funds. For example,

public pension funds typically hold OFC firms 1.59 years longer than independent investment

advisors or mutual funds, and the difference is highly significant with a t-statistic of 7.87.

Throughout Panel A, differences across institution types in holding periods are often most

prominent when independent investment advisors or mutual funds are compared to public

pension funds. Independent investment advisors or mutual funds hold OFC firms for a

significantly longer time than non-OFC stocks, while public pension funds prefer the opposite.

40

Columns 3 and 4 present average percent holding, and we again find prominent

differences in the treatment of OFC versus non-OFC firms when we compare public pension

funds to independent investment advisors or mutual funds. Both public pension funds and

independent investment advisors or mutual funds comprise a significantly lower percent

holding of OFC stocks than non-OFC stocks. Independent investment advisors or mutual funds

have average percent holding 0.39% of OFC stocks and 0.52% of non-OFC stocks, a difference

which is statistically significant (t-statistic of -7.43). Furthermore, independent investment

advisors or mutual funds hold a statistically significantly higher percentage of OFC firms

(0.392%) than public pension funds (0.186%). In summary, it appears that some institutions

that are answerable to a specific clientele (public pension funds) are more averse to investing in

OFC incorporated firms than institutions that invest on behalf of a large, diverse clientele

(mutual funds). Moreover, these results suggest that differential treatment of OFC firms exist

for some institutional types, which can be obscured in the aggregate holdings data.

Panel B of Table 9 presents average years held and percentage held statistics by the

institution’s investment horizon as classified by Bushee (1998, 2001). Long run investors are

identified as “Dedicated”, short run investors are “Transient”, and “Quasi-indexers” appear to

follow a benchmark. Among the statistics reported in Panel B of Table 9, there is no

statistically significance difference between average years held of OFC firms versus non-OFC

firms across the three investor horizon categories. However, there is a significant difference

41

between average percent held of OFC (0.35%) versus non-OFC (0.39%) firms for Transient

investors, with a t-statistic of 2.79. This suggests that short horizon investors with speculative

or market timing motives appear to differentiate OFC firms, as reflected in their significantly

lower ownership in OFC firms versus non-OFC firms. Across investment horizon groups,

Dedicated investors comprise significantly more (2.14%) of the owners of OFC firms than

Transient investors (0.35%), with a t-stat of 25.03 for the difference. Similarly, Dedicated

investors on average hold OFC firms for 5.83 years as compared to an average holding period

of 4.09 for Transient investors. Differences between Dedicated investors and Transient

investors are perhaps not surprising given their differing investment horizons. However, in

general, holding periods of OFC versus non-OFC firms does not differ substantially across

horizon groups.

Panel C of Table 9 presents statistics by institutional investor style (Large Growth, Large

Value, Small Growth, and Small Value) following Bushee (1998, 2001). This set of categories

allows us to view potential firm size effects that can be obscured by the Dedicated, Transient,

and Quasi-indexer categories used in Panel B. Most investment style categories hold OFC

firms for a significantly shorter period than non-OFC firms, with the exception of Large

Growth investors. For example, Small Growth investors hold OFC stocks for an average of

4.51 years and non-OFC stocks for 4.67 years, with a t-statistic of -2.21 for the difference.

There is more variety in percent held across style categories. Large Growth investors hold on

42

average 0.29% of ownership in OFC firms and 0.37% of ownership in non-OFC firms, with a

t-statistic of -3.52 for the difference. Percent ownership of OFC firms is lower than that of

non-OFC firms for all investment style categories except for Large Value which is

insignificant.

In Panel A, independent investment advisors or mutual funds are willing to hold OFC

stocks slightly longer than non-OFC stocks, though in smaller amounts given the average

percentage holding. In Panel B, the OFC holdings of Transient investors echo this finding.

Collectively, this suggests that relatively nimble investors are more comfortable with OFC

stocks In the mutual fund literature, there is evidence that funds that trade frequently (Chen,

Jegadeesh, and Wermers, 2000) or have high turnover (Wermers, 2000) typically display at

least marginally better performance. Cremers and Petajisto (2009) find that funds that score

highly on implementing stock selection enjoy better performance. Yan and Zhang (2009) find

that the trades of institutions which they classify as short term predict future stock returns.

Their predictions do not appear to reverse eventually, are stronger for growth and small cap

stocks, and also have power to forecast earnings surprises. They describe short-term

institutions as “better informed and they trade actively to exploit their informational

advantage”. On the other hand, broad examinations of the mutual fund industry such as Gruber

(1996) and French (2008) report underperformance in excess of 60 basis points per year.

While there is no well-established consensus on which institutional types, if any, are skilled

43

or informed, we find that mutual fund managers do not strongly distinguish or penalize OFC

incorporated firms relative to other firms.

For a final look at how institutional investors view OFC firms, we sought records of

voting on corporate proposals to move legal domicile to an OFC. We immediately encountered

data limitations. First, we have a sample of only slightly more than 100 reincorporation events,

and, among those, only 11 have any US institutional ownership. Second, US institutions were

required to begin filing form N-PX starting August 31st 2004, and almost all of our 11

reincorporation events with US institutional participation occur prior to that time. A few

institutional investors make available online a general description of their philosophy

regarding proxy voting and include information on how they view reincorporation. For

example, AllianceBernstein states “We perform a case-by-case review of proposals that seek

shareholder approval to reincorporate in a different state or country…” and indicates that

benefits to reincorporation in a jurisdiction like Bermuda must be weighed against the cost to

investor protection. 28 There are only a few news stories on institutions objecting to

reincorporation. In 2002, Stanley Works withdrew a plan to reincorporate in Bermuda in

response to political pressure and shareholder objections.29 Several pension funds objected to

28 See their proxy voting manual for March 2013 at:

https://www.alliancebernstein.com/abcom/Our_Firm/Content/CGDocs/Proxy%20Voting%20Manual.pdf 29

http://articles.chicagotribune.com/2002-08-03/business/0208030184_1_stanley-works-reincorporation-nabors-in

dustries

44

Tyco’s Bermuda reincorporation in 2003, although this is related to Tyco’s more fundamental

governance problems at that time. 30 Thus, we have no overwhelming evidence that

institutional investors sought to block reincorporation in OFCs.

5. Summary and conclusions

This paper examines a rarely-explored dimension of the choices corporations make

concerning their legal, regulatory, and disclosure environment. Offshore financial centers

present both the hope of a more efficient, low cost legal home and the fear of a poorly-regulated

environment that benefits managers and other insiders at the expense of ordinary shareholders.

Our results suggest that, while in general the impact of OFC incorporation on valuation appears

to be negative, it is necessary to consider individual firm characteristics and the firm’s home

country regulatory environment to assess the marginal impact of OFC incorporation. We find

that OFC incorporation can destroy value for growing firms from a high quality home country

environment or can add value for firms from some less-developed environments. Thus, it

seems there is nothing inherently naughty or nice about the use of OFCs as legal domiciles by

corporations. More fundamental characteristics such as the absence of growth opportunities,

tight control by insiders, and the legal environment in the home country determine whether

offshore incorporation is enhances or destroys value for shareholders. In this respect, the

choice of legal domicile mirrors many other corporate decisions.

30 http://teamster.org/content/wsj-tyco39s-bermuda-incorporation-challenged-shareholders

45

References

Allen, Franklin, and Faulhaber, Gerald R., 1989, Signalling by underpricing in the IPO market, Journal of

Financial Economics 23, 303 - 323,

Allen, Franklin, Qian, Jun, and Meijun Qian, 2005, Law, finance, and economic growth in China, Journal of

Financial Economics 77, 57 – 116.

Ang, James S., Jiang, Zhiqian, and Wu, Chaopeng, 2012, Good Apples, Bad Apples: Sorting Among

Chinese Companies Traded in the US, unpublished Florida State University working paper.

Bailey, W , Huang, V., and Yang, Z., 2011, Bank loans with Chinese characteristics: Some evidence on inside

debt in a state controlled banking system, Journal of Financial and Quantitative Analysis 46, 1795 – -

1830.

Bailey, Warren, Karolyi, G. Andrew, and Salva, Carolina, 2006, The Economic Consequences of Increased

Disclosure: Evidence from International Cross-Listings, Journal of Financial Economics 81, 175-213.

Bar-Gill, Oren, Barzuza Michal, and Bebchuk, Lucian, 2006, The Market for Corporate Law, Journal of

Institutional and Theoretical Economics 162, 134 – 171.

Bebchuk, Lucian A. and Cohen, Alma, 2003, Firms’ Decisions Where to Incorporate, 46 The Journal of Law

and Economics 46, 383 - 485.

Bickley, Christopher, 2004, Bermuda, British Virgin Islands, and Cayman Islands Company Law, Hong

Kong: Sweet & Maxwell Asia.

Brown, Stephen, Lajbcygier, Paul, and Li, Bob, 2008, Going Negative: What to Do with Negative Book Equity

Stocks, Journal of Portfolio Management 35, 95 – 102.

Bushee, Brian, 1998, The influence of institutional investors on myopic R&D investment behavior,

Accounting Review 73, 305 - 333.

46

Bushee, Brian, 2001, Do Institutional Investors Prefer Near-Term Earnings over Long-Run Value?,

Contemporary Accounting Research 18, 207 – 246.

Chen, Hsiu-Lang, Jegadeesh, Narasimhan, and Wermers, Russell, 2000, The Value of Active Mutual Fund

Management: An Examination of the Stockholdings and Trades of Fund Managers, Journal of Financial

and Quantitative Analysis 35, 343 – 368.

Col, Burcin, and Errunza, Vihang, 2013, Havenly Acquisitions, unpublished McGill University working

paper (February).

Comment, Robert, and Schwert, G. William, 1995, Poison or placebo? Evidence on the deterrence and

wealth effects of modern antitakeover measures, Journal of Financial Economics 39, 3-43.

Cremers, K. J. Martijn, and Petajisto, Antti, 2009, How Active Is Your Fund Manager? A New Measure That

Predicts Performance, Review of Financial Studies 9, 3329 - 3365.

Daines, Robert M., 2001, Does Delaware Law Improve Firm Value?, Journal of Financial Economics 62,

525-558.

Darrough, Masako N., Huang, Rong and Zhao, Sha, 2012, The Spillover Effect of Chinese Reverse Merger

Frauds: Chinese or Reverse Merger?. Unpublished Baruch College working paper available at SSRN:

http://ssrn.com/abstract=2144483

Desai, Mihir A., Foley, C. Fritz, and Hines, James R. Jr., 2006, The Demand for Tax Havens, Journal of

Public Economics 90, 513-531.

Desai, M. A., and J. R. Hines, 2002, Expectations and expatriations: Tracing the causes and consequences of

corporate inversions, National Tax Journal 55, 409 - 440.

Djankov, S., LaPorta, R., Lopez-de-Silanes, F., and Shleifer, A., 2003, Courts, Quarterly Journal of

Economics 118, 453-517.

Djankov, S., La Porta, R., Lopez-de-Silanes, F., and Shleifer, A., 2008, The Law and Economics of

Self-Dealing, Journal of Financial Economics 88, 430 – 465.

Doidge, Craig, Karolyi, G. Andrew, and Stulz, René M., 2004, Why are foreign firms listed in the U.S. worth

more? Journal of Financial Economics 71, 205-238.

47

Doidge, Craig, Karolyi, G. Andrew, and Stulz, René M., 2007, Why do countries matter so much for corporate

governance?, Journal of Financial Economics 86, 1-39.

Doidge, Craig; Karolyi, Andrew; Stulz, René M., 2010, Why Do Foreign Firms Leave U.S. Equity

Markets?, Journal of Finance 65, 1507-1553.

Durnev, Art, Li, Tiemei, and Magnan, Michel, 2011, Beyond Tax Avoidance: Offshore Firms, Institutional

Environment, and Financial Reporting, unpublished University of Iowa working paper.

Durnev, Art, Li, Tiemei, and Magnan, Michel, 2012, Are Offshore Firms Worth More?, unpublished

University of Iowa working paper.

Dyreng, Scott, and Lindsey, Bradley P., 2009, Using Financial Accounting Data to Examine the Effect of

Foreign Operations Located in Tax Havens and Other Countries on US Multinational Firms' Tax Rates,

Journal of Accounting Research 47, 1283 – 1316.

Fernandes, Nuno, Lel, Ugur, and Miller, Darius, 2010, Escape from New York: The Market Impact of

Loosening Disclosure Requirements, Journal of Financial Economics 95, 129-147.

French, K.R., 2008, Presidential Address: The cost of active investing, Journal of Finance 63, 1537 – 1574.

Givoly, Dan, Hayn, Carla, and Lourie, Ben, 2012, Importing Accounting Quality: The Case of Foreign

Reverse Mergers, unpublished Pennsylvania State University and University of California, Los Angeles

working paper (May).

Grinblatt, Mark, and Hwang, Chuan Yang, 1989, Signaling and the Pricing of New Issues, Journal of Finance

44, 393 - 420.

Gruber, M.J., 1996, Another puzzle: the growth in actively managed mutual funds, Journal of Finance 51, 783

– 810.

Hines, Jr., James R., and Rice, Eric M., 1994, Fiscal Paradise: Foreign Tax Havens and American Business,

Quarterly Journal of Economics 109, 149-182.

48

Houston, J., Lin, C., and Ma, Y., 2012, Regulatory Arbitrage and International Bank Flows, Journal of

Finance 67, 1845 – 1895.

Hung, Mingyi, Wong, T. J., and Zhang, Fang, 2011, The Value of Relationship-based and Market-based

Contracting: Evidence from Corporate Scandals in China, unpublished Chinese University of Hong

Kong working paper (May).

International Monetary Fund, 2006, Offshore Financial Centers: The Assessment Program—A Progress

Report, Monetary and Financial Systems Department (February).

International Monetary Fund, 2008, Offshore Financial Centers: A Report on the Assessment Program and

Proposal for Integration with the Financial Sector Assessment Program, Monetary and Capital Markets

Department and the Legal Department (May).

Jindra, Jan, Voetmann, Torben and Walkling, Ralph A., 2012, Reverse Mergers: The Chinese Experience,

unpublished Drexel University working paper. Available at SSRN: http://ssrn.com/abstract=2105814.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998, Law and finance, Journal of Political

Economy 106, 1113–1155.

Lane, Philip R., and Milesi-Ferretti, Gian M., 2010, Cross-Border Investment in Small International Financial

Centers, IMF Working paper 10/38.

Malatesta, Paul H, and Walkling, Ralph A., 1988, Poison Pill Securities: Stockholder Wealth, Profitability, and

Ownership Structure, Journal of Financial Economics 20, 347-376.

Morriss, Andrew P., ed., 2010, Offshore Financial Centers and Regulatory Competition, Washington, DC:

The AEI Press.

Pistor, K., Raiser, M., and Gelfer, S., 2000, Law and finance in transition economies, The Economics of

Transition 8, 325 – 368.

Public Company Accounting Oversight Board (PCAOB), 2011, Activity Summary and Audit Implications

for Reverse Mergers1 Involving Companies from the China Region: January 1, 2007 through March

31, 2010, Research Note # 2011-P1

49

Shaxson, Nicholas, 2011, Treasure Islands: Uncovering the Damage of Offshore Banking and Tax Havens,

New York: Palgrave MacMillian.

Subramanian, Guhan, 2004, The Disappearing Delaware Effect, Journal of Law, Economics & Organization

20, 32 - 59 .

Suss, Esther C., Williams, Oral, and Mendis, Chandima, 2002, Caribbean Offshore Financial Centers: Past,

Present, and Possibilities for the Future, IMF Working Paper 02/88.

United States Securities and Exchange Commission, 2011, Reverse Mergers, Investor Bulletin (9th June).

van der Does de Willebois, Emile, Halter, Emily M., Harrison, Robert A., Park, Ji Won, and Sharman,

J. C., 2011, The Puppet Masters: How the Corrupt Use Legal Structures to Hide Stolen Assets and What

to Do About It, World Bank Stolen Asset Recovery Initiative report.

Wald, John K., and Long, Michael S., 2007, The Effect of State Laws on Capital Structure, Journal of Financial

Economics, 83, 297 - 320.

Welch, Ivo, 1989, Seasoned Offerings, Imitation Costs, and the Underpricing of Initial Public Offerings,

Journal of Finance 44, 421 - 449.

Welch, Ivo, 1991, An Empirical Examination of Models of Contract Choice in Initial Public Offerings,

Journal of Financial and Quantitative Analysis 26, 497 – 518.

Wermers, R., 2000, Mutual fund performance. an empirical decomposition into stock-picking talent, style,

transactions costs, and expenses, Journal of Finance 55, 1655 – 1695.

Yan, Xuemin (Sterling) Yan, and Zhang, Zhe, 2009, Institutional Investors and Equity Returns: Are

Short-term Institutions Better Informed?, Review of Financial Studies 22, 893 - 924.

Zoromé, Ahmed, 2007, Concept of Offshore Financial Centers: In Search of an Operational Definition, IMF

Working Paper 07/87.

50

Table 1. The distribution of country of incorporation, country of address, and exchange listing for sample firms

This table summarizes the country of incorporation and country of address of sample firms. Panel A tabulates sample firms without double-counting cross listings. The

supplement to Table 1 Panel A (at the end of the paper) details the distribution of sample firms for countries of incorporation or countries of average classified as “Others” in

Panel A. Sample firms are incorporated in an OFC but with address not in an OFC as identified from Datastream, www.adr.com, Worldscope, or CompustatNA.

Panel A: Unique firms

Country of Incorporation

Address

Country Bermuda

Cayman

Islands

British

Virgin

Islands

Marshall

Islands Jersey

Isle of

Man Guernsey Bahamas Panama

Netherlands

Antilles Cyprus Others Total

Hong Kong 449 291 7 2 0 0 0 0 0 0 0 1 750

China 31 189 13 0 1 1 0 0 0 0 0 3 238

UK 11 3 18 1 27 28 12 0 0 0 3 7 110

USA 25 11 4 13 0 0 0 3 7 4 0 5 72

Singapore 41 2 0 0 0 0 0 0 0 1 0 0 44

Greece 2 1 0 31 1 0 0 0 0 0 0 0 35

Canada 4 7 6 1 2 0 0 3 0 0 0 1 24

Norway 8 2 0 0 0 0 0 0 0 0 4 2 16

Others 25 12 6 6 5 3 3 4 3 4 2 10 83

Total 596 518 54 54 36 32 15 10 10 9 9 29 1372

51

Table 1 continued.

Panel B: Listings on US exchanges

Country of Incorporation

Exchange Country of address Bermuda Cayman Islands British Virgin Islands Other

NYSE/Amex Hong Kong 1 1 0 1

China 0 19 2 3

UK 1 0 0 2

US 6 3 0 10

Other 5 0 0 19

NASDAQ Hong Kong 3 2 2 7

China 0 38 4 2

UK 1 0 0 1

US 3 2 0 4

Other 3 3 0 24

Over the counter Hong Kong 10 8 0 0

China 0 10 0 0

UK 0 0 0 1

US 5 1 3 6

Other 9 5 2 23

52

Table 1 continued.

Panel C: Listings on non US exchanges

Country of Incorporation

Exchange Country of address Bermuda Cayman Islands British Virgin Islands Other

London Hong Kong 8 0 1 0

China 0 2 0 1

UK 8 3 4 18

US 0 0 0 0

Other 1 0 0 6

Hong Kong Hong Kong 273 174 1 2

China 4 6 0 0

UK 0 0 0 0

US 0 0 0 0

Other 0 0 0 0

Berlin Hong Kong 117 91 2 0

China 22 72 6 2

UK 6 0 21 55

US 2 0 0 2

Other 7 2 3 18

Frankfurt Hong Kong 333 228 7 2

China 30 213 12 0

UK 0 2 16 73

US 23 9 4 20

Other 46 17 13 132

53

Table 2. Summary statistics on individual company characteristics

Sample firms are incorporated in an OFC but with address not in an OFC as identified from Datastream, www.adr.com, Worldscope, or CompustatNA. Their financials are

collected from Datastream and CompustatNA. Control firms have Datastream country of address in one of the countries of address in the OFC-incorporated sample but are not

incorporated in an OFC. Both sets of firms are screened to exclude firms with Total Assets less than 10 million US dollars or no data on Total Assets, Sales, Long Term Debt,

Total Debt, Shareholder’s Equity, or Market Value of Equity. We remove resulting extreme observations of Tobin’s q by dropping the top and bottom 1%, Each variable in the

table is averaged across firm-years by firm, and the table in turn reports averages across firms.

Panel A: Tobin’s q

Country of Incorporation

Control Bermuda Cayman Islands British Virgin Islands Others

Address

Country

Number

of Firms

Average

Tobin's

Q

Number

of Firms

Average

Tobin's

Q

Difference

with

Control

Number

of Firms

Average

Tobin's

Q

Difference

with

Control

Number

of Firms

Average

Tobin's

Q

Difference

with

Control

Number

of Firms

Average

Tobin's

Q

Difference

with

Control

Hong Kong 72 0.8779 352 1.0974 -0.2195 197 1.3998 -0.5219 7 0.4770 0.4009 2 0.9408 -0.0629

China 419 2.8680 25 0.8125 2.0555 121 1.5871 1.2809 11 0.4543 2.4137 2 0.4877 2.3803

UK 625 1.3682 6 1.1517 0.2165 1 1.9194 -0.5512 7 1.0524 0.3157 46 1.3740 -0.0059

US 2561 1.2979 14 1.2562 0.0417 3 1.1793 0.1186 3 1.3084 -0.0105 15 1.3509 -0.0530

Singapore 291 1.2012 12 1.4074 -0.2062 1 0.8738 0.3274 NA NA NA NA NA NA

Canada 312 1.2262 3 1.3357 -0.1096 3 1.0060 0.2202 4 1.4334 -0.2073 NA NA NA

Norway 78 1.1122 8 1.1180 -0.0057 2 1.4902 -0.3780 NA NA NA 5 1.9055 -0.7932

Netherlands 141 1.3477 2 1.9152 -0.5675 NA NA NA NA NA NA 4 1.4101 -0.0624

Others 3377 1.2513 15 1.1788 0.0725 6 0.7285 0.5228 4 0.5912 0.6601 37 1.2864 -0.0352

Total 7876 437 334 36 111

54

Table 2 continued.

Panel B: Other firm characteristics

Country of Incorporation

Control Bermuda Cayman Islands British Virgin Islands Other

Number

of

available

firm-years Average

Number

of

available

firm-years Average

Difference

with

Control

Number

of

available

firm-years Average

Difference

with

Control

Number

of

available

firm-years Average

Difference

with

Control

Number

of

available

firm-years Average

Difference

with

Control

Sales

Growth 64,969 2.5534 2,908 0.9679 -1.5855 1,163 0.4834 -2.0700 93 0.3385 -2.2149 537 0.7782 -1.7752

Log Sales

USD 75,720 12.1356 3,718 12.0095 -0.0828 1,792 12.1450 0.0475 158 12.1969 0.1107 758 11.9202 -0.2101

Debt to

Equity 76,363 1.1632 3,811 0.3107 -0.8524 1,809 0.3402 -0.8229 175 0.4329 -0.7302 840 1.2873 0.1242

Total

possible

firm-years 76,438

3,811

1,810

175

847

55

Table 3. Single-stage regressions to explain individual firm Tobin’s q

This table reports regression estimates for specifications with Tobin’s q as dependent variable and characteristics of the firm and the firm’s address country as

independent variables. These include a dummy set to 1 for firms incorporated in an OFC; CAPEX divided by total assets; equity and debt capital raised divided by

total assets, insider holdings as a fraction of equity, the Anti-Director Rights index of La Porta et al with values inserted for China and Russia as detailed in the text (AD

index); estimated time to evict a tenant of Djankov et al (2008); time series average World Bank Doing Business Protecting Investors - Extent of disclosure index

indicator; DR dummy indicating a Level II or III cross listing on a US stock exchange; and year and industrial sector fixed effects. Observations are firm-years.

Panel A

China or Hong Kong address only

1 2 3 4 5 6 7

Slope t-test slope t-test Slope t-test Slope t-test Slope t-test Slope t-test Slope t-test

Intercept 1.2834 557.15 1.3013 42.61 1.3910 41.61 1.4058 18.02 1.2158 14.25 1.7790 79.24 1.0144 8.05

OFC dummy -0.0948 -11.64 -0.1241 -15.39 -0.1303 -15.51 -0.0537 -5.62 -0.0220 -1.90 -0.6295 -24.34 0.1079 3.48

Capex - - - - - - 0.0338 14.58 - - - - 0.3014 13.04

Capital raised - - - - - - - - 125.188 20.16 - - - -

Insider holdings - - - - - - -0.1880 -16.15 -0.1943 -12.04 - - -0.1514 -2.53

AD index - - - - -0.0300 -10.75 -0.0091 -2.47 -0.0146 -2.96 - - - -

Evict - - - - -0.0004 -15.85 -0.0002 -5.97 -0.0005 -1.27 - - - -

Disclosure - - - - 0.0142 11.23 -0.0040 -2.35 -0.0016 -0.69 - - - -

Fixed effects No Yes Yes Yes yes No Yes

Adj. R-squared 0.0016 0.1122 0.1178 0.1275 0.1380 0.0791 0.1477

Observations 83081 83081 80389 45138 26012 6887 5227

56

Table 3 continued.

Panel B: Specifications with cross-listing dummy

China or Hong Kong address only

8 9 10 11 12

Slope t-test Slope t-test Slope t-test Slope t-test Slope t-test

Intercept 1.3960 41.74 1.4117 18.1 1.2275 14.38 1.7805 79.73 1.0782 8.62

OFC dummy -0.1235 -14.48 -0.0424 -4.29 -0.0096 -0.8 -0.6110 -23.67 0.1151 3.74

Capex - - 0.0336 14.49 - - - - 0.2852 12.41

Capital raised - - - - 124.567 20.06 - - - -

Insider holdings - - -0.1934 -16.53 -0.2009 -12.37 - - -0.2526 -4.19

AD index -0.0306 -10.95 -0.0105 -2.83 -0.0166 -3.33 - - - -

Evict -0.0004 -16.06 -0.0002 -6.17 -0.0001 -1.5 - - - -

Disclosure 0.0140 11.02 -0.0047 -2.7 -0.0025 -1.07 - - - -

DR dummy -0.0774 -4.61 -0.0979 -4.47 -0.0881 -3.55 -0.6168 -8.65 -0.6668 -9.55

Fixed effects yes Yes Yes No Yes

Adj. R-squared 0.1181 0.1279 0.1383 0.0889 0.1624

Observations 80389 45138 26012 6887 5227

57

Table 4. Two-stage regressions to explain individual firm Tobin’s q

This table reports regression estimates for specifications with Tobin’s q as dependent variable and characteristics of the firm and the firm’s address country as independent variables.

These include a dummy set to 1 for firms incorporated in an OFC; CAPEX divided by total assets; equity and debt capital raised divided by total assets, insider holdings as a fraction

of equity, the Anti-Director Rights index of La Porta et al with values inserted for China and Russia as detailed in the text ; estimated time to evict a tenant of Djankov et al (2008); time

series average World Bank Doing Business Protecting Investors - Extent of disclosure index indicator; number of underwriters; underwriter fees as fraction of amount raised; and year

and industrial sector fixed effects. Observations are firm-years. First pass dependent variable for Probit is OFC dummy while second pass dependent variable is Tobin’s q. Each

specification is run separately for each year and the table reports averages of resulting coefficients.

Averages of yearly regression output

First-pass Probit Second-pass Heckman

Slope z-test Slope t-test

Intercept -4.134 -12.67 4.0143 2.03

OFC dummy - - -0.2274 -1.80

Capex 1.974 3.55 0.2937 1.41

Insider holdings 1.145 6.83 -0.6630 -1.94

AD index 0.866 12.26 -0.2731 -2.05

Evict - - 0.0001 1.94

Disclosure - - -0.0237 -1.37

Inverse Mills Ratio - - -0.8910 -1.69

Fixed effects No - Yes

Pseudo R-squared 0.2087 -

Adj. R-squared - - 0.2965

Observations 1400 1400

58

Table 5. Single-stage regressions with interactive terms to decompose individual firm Tobin’s q

This table reports regression estimate for specification with Tobin’s q as dependent variable and characteristics of the firm and the firm’s

address country as independent variables. These include a dummy set to 1 for firms incorporated in an OFC; CAPEX divided by total

assets; equity and debt capital raised divided by total assets, insider holdings as a fraction of equity, the Anti-Director Rights index of La

Porta et al with values inserted for China and Russia as detailed in the text (AD index); estimated time to evict a tenant of Djankov et al

(2008); time series average World Bank Doing Business Protecting Investors - Extent of disclosure index indicator; number of

underwriters; underwriter fees as fraction of amount raised; and year and industrial sector fixed effects. Observations are firm-years.

1 2 3 4

Slope t-test Slope t-test Slope t-test Slope t-test

Intercept 1.3737 17.59 1.1557 13.52 1.3674 17.51 1.1548 13.51

OFC dummy 0.6443 8.91 0.6633 7.39 1.0616 12.02 1.2607 10.55

Capex 0.0517 16.78 - - 0.0513 16.66 - -

Capital raised - - 110.093 15.97 - - 109.940 15.95

Insider holdings -0.1664 -13.64 -0.1959 -11.45 -0.1684 -13.78 -0.1965 -11.47

AD index 0.0014 0.35 0.0028 0.53 0.0005 0.12 0.0017 0.32

Evict -0.0002 -5.02 0.0001 1.31 -0.0002 -5.1 0.0000 1.13

Disclosure -0.0046 -2.56 -0.0035 -1.41 -0.0047 -2.63 -0.0035 -1.42

DR dummy - - - - -0.1129 -3.31 -0.0624 -1.6

Capex x OFC

dummy

-0.0435 -9.31 -

-

-0.0433 -9.27 -

-

Capital raised x

OFC dummy -

-

73.0696 4.76 -

-

67.8844 4.42

Insider holdings x

OFC dummy

-0.0302 -0.73

0.1713 3.37

-0.0988 -2.33

0.0936 1.8

AD index x OFC

dummy

-0.1048 -8.39

-0.1107 -7.89

-0.0999 -8

-0.1095 -7.81

Evict x OFC

dummy

-0.0009 -5.21

-0.0014 -6.79

-0.0013 -7.43

-0.0019 -8.85

Disclosure x OFC

dummy

-0.0027 -0.39

-0.0042 -0.5

-0.0337 -4.26

-0.0492 -4.8

DR dummy x OFC

dummy - - - -

-0.1909 -3.82

-0.2820 -4.73

Fixed effects Yes Yes Yes Yes

Adj. R-squared 0.1312 0.1420 0.1327 0.1440

Observations 45138 26012 45138 26012

59

Table 6. Comparison of IPOs incorporated in an OFC to other IPOs This table summarizes data on initial public offerings of sample of OFC and control firms. Period is 1990 to June 2012. Source of offering data is SDC.

Panel A: Medians (first row of each section) and other statistics on offering characteristics for all IPOs

Domicile Percent IPO

return 1 day

Percent IPO

return

90 days

Percent

IPO return

1 year

Principal

amount

million USD

Management

fee percent

Underwriting

fee percent

Over-

allotment

Percent

Number of

lead

managers

Lock-up

dummy

Venture capital

dummy

All 13.75 15.74 10.619 33.374 1.4 1.455 0 1 1 0

1% -25.301 -54.294 -87.56 1.253 0 0.352 0 1 0 0

99% 358.189 298.345 372.92 1386 3 5 15 10 1 1

Observations 941 933 941 941 233 207 941 941 429 930

Bermuda 6.82 4.857 -10.714 33.809 1.3485 1.6135 0 2 0 0

1% -23 -52.27 -76.595 6.4 0.25 1.407 0 1 0 0

99% 135.714 207.894 205.555 1050 1.407 5 15 11 1 1

Observations 35 35 35 35 6 6 35 35 30 35

British Virgin Islands 2 1.92 -40.6 30.956 2.087 1.692 0 1 1 0

1% -13.333 -34.833 -92.173 2 1.412 1.412 0 1 0 0

99% 32 39.428 17.479 225.09 2.308 1.971 15 3 1 0

Observations 11 11 11 11 3 3 11 11 11 11

Cayman Islands 9.482 7.4305 -5.797 37.998 0.64 0.64 3.439 2 0 0

1% -25.301 -59.23 -88.455 0.988 0.5 0.5 0 1 0 0

99% 202.5 185 410.489 828.704 1.5 1.5 15 7 1 1

Observations 87 84 87 87 9 9 87 87 66 87

60

Table 6 continued.

Panel B: Medians (first row of each section) and other statistics on offering characteristics for China or Hong Kong address IPOs only

Domicile Percent IPO

return 1 day

Percent IPO

return

90 days

Percent

IPO return

1 year

Principal

amount

million USD

Management

fee percent

Underwriting

fee percent

Over-

allotment

Percent

Number of

lead

managers

Lock-up

dummy

Venture capital

dummy

All 44.222 42.847 23.433 54.991 1.2 1.4 0 1 0 0

1% -23.75 -52.27 -85.961 2.587 0.25 0.5 0 1 0 0

99% 428.817 344.887 318.704 1990.74 2.308 2.25 15 6 1 1

Observations 309 306 309 309 15 15 309 309 88 309

Bermuda 17.1235 5.5375 4.627 33.5915 0.807 1.9885 0 2 0 0

1% -23 -52.27 -76.595 -6.4 0.25 1.727 0 1 0 0

99% 135.714 207.894 205.555 128.686 1.364 2.25 12.706 3 1 1

Observations 18 18 18 18 2 2 18 18 14 18

British Virgin Islands 1.92 0.74 -40.6 8.625 2.087 1.692 0 2 1 0

1% -13.333 -34.833 -56.72 2 1.412 1.412 0 1 0 0

99% 22.79 18.84 17.39 160 2.308 1.971 15 3 1 0

Observations 5 5 5 5 3 3 5 5 5 5

Cayman Islands 9.4775 6.999 -0.68 37.8155 0.64 0.64 5.231 2 0 0

1% -25.301 -59.23 -88.455 0.988 0.5 0.5 0 1 0 0

99% 202.5 185 410.489 828.704 1.5 1.5 15 7 1 1

Observations 82 79 82 82 9 9 82 82 61 82

61

Table 7. Single-stage regressions to explain individual firm initial public offering (IPO) return

This table reports regression estimates for specifications with one year percent IPO return as dependent variable and characteristics of the firm and the firm’s address country as independent variables.

These include a dummy set to 1 for firms incorporated in an OFC; CAPEX divided by total assets; equity and debt capital raised divided by total assets, insider holdings as a fraction of equity, the

Anti-Director Rights index of La Porta et al with values inserted for China and Russia as detailed in the text (AD index); estimated time to evict a tenant of Djankov et al (2008); time series average

World Bank Doing Business Protecting Investors - Extent of disclosure index indicator; number of underwriters; underwriter fees as fraction of amount raised; and year and industrial sector fixed

effects. Observations are firm-years.

China or Hong Kong address only

1 2 3 4 5 6 7 8

slope t-test slope t-test slope t-test Slope t-test Slope t-test slope t-test slope t-test slope t-test

Intercept 33.4708 9.9 -26.0128 -0.36 -108.3147 -0.87 -92.9098 -0.73 -81.1708 -0.63 -81.6443 -0.29 46.8482 8.29 126.0545 0.173

OFC dummy -20.8483 -2.52 -21.3802 -2.47 -25.5835 -1.96 -30.0474 -2.13 -27.5609 -1.93 4.3858 0.06 -29.4219 -3.05 -14.1073 0.612

Capex - - - - 0.3267 0.34 - - - - - - 11.0309 0.47

Capital raised - - - - - - 5913.55 0.84 6162.932 0.88 15695.16 1.07 - -

Insider holdings - - - - -45.838 -2.36 -40.842 -1.88 -44.121 -2.02 -64.609 -1.12 - - -60.8064 0.18

AD index - - - - 3.4387 0.68 3.8994 0.73 3.4968 0.65 21.1031 0.64 - - - -

Evict - - - - 0.0698 1.25 0.0671 1.14 0.0459 0.75 0.0998 0.32 - - - -

Disclosure - - - - 9.9475 3.00 8.8151 2.68 8.2160 2.47 -3.8852 -0.16 - - - -

Underwriters - - - - - - - - -2.9780 -1.27 - - - - - -

Fees - - - - - - - - - - -77.2925 -0.39 - - - -

Fixed effects No yes yes Yes Yes yes no yes

Adj. R2 0.0056 0.1209 0.0974 0.1010 0.1023 -0.1026 0.0274 0.3121

Observations 941 941 506 462 462 130 309 101

62

Table 8. Single-stage regressions to explain U.S. institutional holdings

This table reports regression estimates for specifications with U.S. institutional holding (13f data) as percent of total equity or numbers dependent variable and

characteristics of the firm and the firm’s address country as independent variables. These include a dummy set to 1 for firms incorporated in an OFC; CAPEX divided

by total assets, equity and debt capital raised divided by total assets, insider holdings as a fraction of equity, the Anti-Director Rights index of La Porta et al with values

inserted for China and Russia as detailed in the text (AD index); estimated time to evict a tenant of Djankov et al (2008); time series average World Bank Doing

Business Protecting Investors - Extent of disclosure index indicator; DR dummy indicating a Level II or III cross listing on a US stock exchange; and year and industrial

sector fixed effects. Observations are firm-years. It is not possible to report regressions for “China or Hong Kong address only” because all such firms in our sample

with 13f data for our time period were OFC incorporated.

Panel A

1 2 3 4 5

Slope t-test slope t-test slope t-test slope t-test Slope t-test

Intercept 0.4065 146.52 0.5520 9.93 0.5516 7.24 0.5259 9.65 0.0577 0.84

OFC dummy -0.0174 -1.23 -0.1638 -13.01 0.0261 1.78 0.0324 2.32 -0.0079 -0.49

Capex - - - - - - 0.0050 0.45 - -

Capital raised - - - - - - - - -18.9134 -2.8

Insider holdings - - - - - - -0.5764 -47.64 -0.6505 -33.86

AD index - - - - 0.1494 14.51 0.1430 12.56 0.1683 12.26

Evict - - - - -0.0004 -3.68 -0.0001 -1.17 0.0002 1.73

Disclosure - - - - -0.0877 -18.18 -0.0728 -14.84 -0.0619 -11.18

Fixed effects No yes Yes yes Yes

Adj. R2 0.0000 0.2411 0.2819 0.4329 0.3920

Observations 13648 13648 13551 9799 5414

63

Table 8 continued. Panel B: Specifications with cross-listing dummy

6 7 8

slope t-test slope t-test Slope t-test

Intercept 0.5570 7.31 0.3209 5.75 0.1497 2.14

OFC dummy -0.0284 -1.22 -0.0723 -2.69 -0.1726 -5.11

Capex 0.1516 14.69 0.0095 0.86 - -

Capital raised - - - - -15.7680 -2.33

Insider holdings - - -0.5751 -47.57 -0.6490 -33.87

AD index - - 0.1475 12.92 0.1703 12.43

Evict -0.0005 -4.55 -0.0003 -2.71 -0.0001 -0.68

Disclosure -0.0885 -18.33 -0.0737 -15.03 -0.0623 -11.27

DR dummy 0.0731 3.02 0.1319 4.56 0.2071 5.54

Fixed effects yes yes yes

Adj. R2 0.2824 0.4340 0.3953

Observations 13551 9799 5414

64

Table 9. Duration and size of U.S. institutional investor holdings sorted by OFC versus non OFC incorporated stock and by institutional investor type

The statistics in this table are computed as follows. We exclude Investment Company and Miscellaneous because the categories are too general.

Panel A: Sorted by Bushee (1998, 2001) institution type

Average years held Average percent held

Institutional investor type 1. OFC stocks

2. Non OFC

stocks 1 – 2 t-test 3. OFC stocks

4. Non OFC

stocks 3 – 4 t-test

a. Bank 6.19 6.31 -0.12 -1.18 0.22 0.38 -0.16 -1.06

b. Corporate pension fund 6.16 5.81 0.34 1.25 0.16 0.15 0.02 0.26

c. Independent investment advisor

(mutual fund) 4.68 4.54 0.14 3.89 0.39 0.52 -0.12 -7.43

d. Insurance company 6.36 6.61 -0.25 -1.32 0.30 0.31 -0.01 -0.11

e. Public pension fund 6.27 7.57 -1.30 -4.99 0.19 0.28 -0.09 -3.00

f. University and foundation endowments 5.90 5.21 0.69 1.31 0.11 0.20 -0.09 -0.70

t(a – b) 0.11 9.01 -1.45 1.21 2.56 -0.39

t(a – c) 15.82 106.75 -2.89 -6.27 -7.6 -0.37

t(a - d) -0.76 -8.14 0.6 -2.31 1.43 -0.58

t(a – e) -0.27 -26.88 4.27 1.0 1.59 -0.19

t(a – f) 0.43 11.0 -1.25 1.02 1.02 -0.07

t(b – c) 6.11 33.3 0.93 -2.93 -17.36 1.35

t(b - d) -0.54 -13.24 1.7 -2.21 -9.21 0.26

t(b – e) -0.29 -26.29 4.23 -0.57 -10.79 1.75

t(b - f) 0.35 6.09 -0.55 0.46 -2.1 0.68

t(c – d) -10.85 -83.79 2.73 1.93 17.36 -1.9

t(c – e) -7.87 -95.48 7.7 3.44 16.31 -0.38

65

t(c – f) -2.43 -9.64 -1.22 1.58 7.83 -0.15

t(d – e) 0.29 -17.7 3.29 2.56 2.22 1.41

t(d – f) 0.64 13.67 -1.42 1.47 3.36 0.48

t(e – f) 0.56 21.89 -2.88 1.44 4.29 -0.04

66

Table 9 continued.

Panel B: Sorted by Bushee (1998, 2001) investor horizon

Average years held Average percent held

Institutional investor

style 1. OFC stocks 2. Non OFC stocks 1 – 2 t-test 3. OFC stocks 4. Non OFC stocks 3 - 4 t-test

a. Quasi-Index 5.866547 5.835567 0.03098 0.61 0.3243% 0.4415% -0.117% -2.77

b. Dedicated 5.834928 5.593026 0.241902 0.253 2.143% 2.3254% -0.182% -0.7

c. Transient 4.094468 4.07414 0.020328 0.611 0.3502% 0.3929% -0.043% -2.79

t(a – b) 0.13 6.27 -0.92 -26.82 -44.8 0.31

t(a – c) 23.79 118.24 0.15 -1.52 3.84 -1.35

t(b - c) 10.55 55.94 1.4 25.03 98.06 -1.43

Panel C: Sorted by Bushee (1998, 2001) investment style

Average years held Average percent held

1. OFC stocks 2. Non OFC stocks 1 – 2 t-test 3. OFC stocks 4. Non OFC stocks 3 - 4 t-test

a. Large Growth 5.896529 5.646518 0.250011 3.1 0.2906% 0.375% -0.084% -3.52

b. Large Value 5.75135 5.872775 -0.121425 -1.76 0.2389% 0.2977% -0.059% -0.78

c. Small Growth 4.511445 4.670569 -0.159124 -2.21 0.5537% 0.8295% -0.276% -6.79

d. Small Value 4.323077 4.459075 -0.135998 -2.58 0.4475% 0.5688% -0.121% -4.66

t(a – b) 1.28 -11.01 3.47 2.39 3.71 -0.27

t(a – c) 11.97 44.13 3.76 -7.77 -42.91 4.17

t(a - d) 15.94 64.23 4.15 -5.47 -23.6 1

t(b – c) 11.8 57.21 0.36 -11.81 -22.61 2.12

t(b – d) 15.98 81.27 0.16 -9.2 -15 0.76

t(c - d) 2.16 11.72 -0.27 3.22 24.48 -3.38

67

Supplement to Table 1 Panel A. Other countries of incorporation or address

Other Country of Incorporation

Address

Country AG AI BB BZ CK CN GI HK IE LI LU MC MT SE SG

HK 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0

CN 2 0 0 0 0 0 0 1 0 0 0 0 0 0 0

GB 0 0 0 1 0 0 3 0 1 0 1 0 1 0 0

US 0 2 0 0 1 0 0 0 1 1 0 0 0 0 0

SG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

GR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CA 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0

NO 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0

AG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

AT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

AU 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

BE 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0

BM 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0

BZ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

CH 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0

DE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

DK 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

FR 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0

IE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

IL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

JP 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

KR 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

LU 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0

MX 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

MY 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

NL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

PE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

RU 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

TW 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2

TZ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

VE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

VG 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

VN 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

ZA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

68

Supplement to Table 1 Panel A continued.

Other  Country of Incorporation 

Address 

Country  BM  KY  VG  MH  JE  IM  GG  BS  PA  AN  CY 

AG  0  0  1  0  0  0  0  0  0  0  0 

AT  0  0  0  0  1  0  0  0  0  0  0 

AU  7  0  1  0  1  0  0  0  0  0  0 

BE  0  0  0  0  0  0  0  0  0  0  1 

BM  0  1  0  6  0  0  0  0  0  0  0 

BZ  3  0  0  0  0  1  0  0  0  0  0 

CH  1  1  0  0  0  0  1  0  0  0  0 

DE  0  0  0  0  0  0  1  0  0  0  0 

DK  0  0  0  0  0  0  0  1  0  0  0 

FR  0  0  0  0  0  0  0  1  0  0  0 

IE  2  0  0  0  1  0  0  0  0  0  0 

IL  0  0  0  0  0  0  0  0  1  0  0 

JP  0  1  0  0  0  0  0  0  0  0  0 

KR  0  2  0  0  0  0  0  0  0  0  0 

LU  1  1  0  0  0  1  0  0  0  0  0 

MX  0  0  0  0  0  0  0  0  1  0  0 

MY  2  0  0  0  2  0  0  0  0  0  0 

NL  2  0  0  0  0  0  0  1  0  4  0 

PE  1  0  0  0  0  0  0  1  1  0  0 

RU  0  0  0  0  0  0  0  0  0  0  1 

TW  3  3  0  0  0  0  0  0  0  0  0 

TZ  1  0  0  0  0  0  0  0  0  0  0 

VE  0  1  0  0  0  0  0  0  0  0  0 

VG  0  0  0  0  0  1  0  0  0  0  0 

VN  0  2  0  0  0  0  0  0  0  0  0 

ZA  2  0  4  0  0  0  1  0  0  0  0 

69

Supplement to Table 3. Correlations among country characteristics

Correlations are computed with one observation per country.

English law

dummy

AD

index

Milken capital

index

Check

Evict

Regulatory

Quality

AD index 0.8249

Milken capital index 0.5355 0.4755

Check -0.2587 -0.2876 -0.4884

Evict -0.4892 -0.3775 -0.3596 0.3586

Regulatory Quality 0.2994 0.2 0.8757 -0.2907 -0.2122

Disclosure 0.5137 0.5177 0.0996 -0.2826 -0.3015 -0.0959

70

Supplemental Table. Tobin’s q and reincorporation from non OFC country to OFC country

This table summarizes the valuation of companies that moved incorporation to an OFC during the period 1990 to June

2012. “Year t” refers to t years before or after the reincorporation. Book value of equity less than zero tends to inflation

Tobin’s q. Sources of data include Mergentonline and Datastream.

OFC domicile Year -2 Year -1 Year 0 Year +1 Year +2

All

Mean Tobin’s q 2.64 20.31 2.23 20.38 2.98

Median 1.18 1.00 1.03 1.02 1.07

Minimum 0.36 -2.68 -1.00 -0.75 0.21

Maximum 98.08 1870.62 110.05 2463.00 52.46

Observations 82 98 120 129 132

Mean for book equity > 0 2.65 1.16 1.27 1.18 1.25

Observations with book equity < 0 3 5 6 9 9

Bermuda

Mean Tobin’s q 1.18 1.16 1.17 1.11 5.20

Median 1.03 0.93 0.97 1.00 1.08

Minimum 0.36 -0.74 0.02 -0.75 0.21

Maximum 4.10 4.22 4.49 3.72 52.46

Observations 17 21 25 34 36

Mean for book equity > 0 1.18 1.16 1.17 1.11 1.06

Observations with book equity < 0 0 0 0 0 3

British Virgin Islands

Mean Tobin’s q 49.97 374.27 27.72 492.95 0.48

Median 49.97 0.83 0.92 0.67 0.33

Minimum 1.86 -2.68 -1.00 0.18 0.29

Maximum 98.08 1870.62 110.05 2463.00 0.98

Observations 2 5 4 5 4

Mean for book equity > 0 49.97 0.18 -0.38 0.44 0.48

Observations with book equity < 0 0 1 2 1 0

Cayman Islands

Mean Tobin’s q 0.51 0.51 0.40 0.38 0.36

Median 0.51 0.51 0.40 0.38 0.36

Minimum 0.39 0.35 0.24 0.15 0.23

Maximum 0.62 0.68 0.55 0.61 0.49

Observations 2 2 2 2 2

Mean for book equity > 0 0.505157 0.5141377 0.397052 0.380423 0.3571353

Observations with book equity < 0 0 0 0 0 0

71

Supplemental Table This table documents the sample of firms moving legal domicile for which ownership information for years zero to plus two of re-incorporation is available.

Re‐incorporation from 

Re‐incorporation to  BVI  Bermuda  Canada  Hong Kong  Ireland Cayman Islands Liberia  Netherlands New Zealand  Norway SG UK US  Total 

BVI  0  0  0  0  0  0  0  0  0  0  0  0  1  1 

Bermuda  0  0  0  8  0  8  0  0  2  0  1  2  5  26 

China  0  0  0  1  0  0  0  0  0  0  0  0  0  1 

Cyprus  0  0  0  0  0  0  0  0  0  1  0  0  0  1 

Ireland  0  4  0  0  0  1  0  1  0  0  0  0  1  7 

Cayman Islands  3  1  0  1  0  0  0  0  0  0  0  1  4  10 

Luxembourg  0  0  1  0  0  0  0  0  0  0  0  0  0  1 

Marshall Islands  0  0  0  0  0  0  1  0  0  0  0  1  0  2 

Switzerland  1  5  1  0  0  1  0  0  0  0  0  0  3  11 

UK  0  0  0  0  1  0  0  0  0  0  0  0  0  1 

US  4  1  0  0  0  1  0  0  0  0  0  0  0  6 

Total  8  11  2  10  1  11  1  1  2  1  1  4  14  67