Docstoc

Commercialization Improve the Performance of Stock Exchanges

Document Sample
Commercialization Improve the Performance of Stock Exchanges Powered By Docstoc
					    Can Commercialization Improve the Performance of Stock Exchanges Even
                          Without Corporatization?



                             Isaac Otchere* and Erin Oldford
                                Sprott School of Business
                                   Carleton University
                                  Ottawa, ON K1V 2H1
                                         Canada




Abstract

A significant number of stock exchanges have demutualized and gone public in recent
years. We examine the performance of mutual, demutualized, and publicly-listed
exchanges to ascertain whether corporatization leads to performance improvement
beyond what commercialization generates. We find evidence of performance
improvements along the exchange governance continuum, suggesting that corporatization
leads to increased operational efficiencies and profitability. However, our robustness test
reveals a puzzle as we do not find significant incremental improvement in performance
for exchanges that have gone through the three phases. We conclude that
commercialization provides sufficient freedom for exchanges to fully exploit monopoly
rents before going public.


Keyword: stock exchanges, demutualization, corporatization, self-listing, operating
performance

JEL Classification: G15, G32, L2


*Corresponding author: email: iotchere@sprott.carleton.ca
     Can Commercialization Improve the Performance of Stock Exchanges Even
                           Without Corporatization?

1.     Introduction

Stock exchanges have traditionally been organized as mutual, not-for-profit

organizations. However, in the past decade technological developments, in combination

with the liberalization of capital markets, have created a much different operating

environment for exchanges, one that questions the effectiveness of the traditional

operating model and the mutual ownership governance structure. As a mutually-owned

entity, the exchange’s mandate is to serve its members’ interests. In the current

competitive environment, exchanges are forced to operate with much more efficiency and

flexibility; however, the user-owned nature of the mutual form constrains management’s

ability to effectively respond to the competition. Consequently, a number of stock

exchanges have responded to the challenges by changing their governance structure and

business model from mutual (not-for-profit) to demutualized (for-profit) entities.

Demutualization entails commercialization of the exchange's business and focuses

managers' attention on profitability and efficiency considerations, while preserving

private ownership. The World Federation of Exchanges (WFE, 2006) reports that in

1998, 38% of its member had a for-profit mandate. Since then, commercialization has

risen markedly to 75% of WFE-member exchanges in 2006.

       Increasingly, exchanges have taken a further step by corporatizing their structure,

going public and listing the shares on publicly traded exchanges, sometimes on the same

exchange that they organize, a phenomenon dubbed self-listing (Otchere, 2006). By

corporatizing the exchange, the owner and user interests are completely decoupled,

thereby bestowing on managers the freedom to pursue business strategies that increase



                                                                                        2
the firm’s competitiveness and maximize value for shareholders. The major stock

exchanges, including NYSE, the London Stock Exchange, Hong Kong Stock Exchange,

Australian Stock Exchange, and the NASDAQ have all corporatized their structure and

are now publicly traded firms.

       Despite the growing popularity of exchange corporatization, important questions

remain, such as how exchanges will deal with the conflicts of interest that arise, and

whether self-listing is necessary at all. The first question has occupied the attention of

policy makers and regulators (see International Organization of Securities Commissions

2006). The second question which has received less attention in the academic literature is

relevant because as more and more exchanges contemplate a change in organizational

form with the view to making the entity more efficient and competitive, it is important to

understand what form of exchange governance structure is most likely to improve

performance of the exchange and whether it is beneficial from operating efficiency

perspective for exchanges to go beyond demutualization to corporatize their structure.

The question is even more important given the significant direct and indirect costs of

going public.

       The few empirical studies focusing on the performance of listed stock exchanges

(including Mendiola and O’Hara, 2003, Otchere, 2006, and Aggarwal, 2002) focus on

comparing the pre- and post-privatization operating performance of self-listed exchanges.

These studies provide evidence to show that self-listing leads to increase in efficiency,

profitability, and capital expenditure. However, for exchanges that have demutualized

and become self-listed, it is unclear whether the improvements in performance occur

because the exchange pursues a new business model – i.e., profit maximization objective




                                                                                        3
(commercialization) or because of capital market monitoring and threat of takeover that

result from self-listing (i.e., corporatization). Theoretically, corporatization can help

exchanges to operate more profitably and efficiently because it subjects managers to the

pressures of financial markets and the attendant monitoring and discipline of returns-

oriented investors. However, by demutualizing and remaining privately owned,

exchanges would be able to exploit many profitable entrepreneurial ventures that they

would otherwise not have been able to undertake (Otchere, 2006) without necessarily

going public. The improvements documented for self-listed exchanges could thus be

attributable to the change in business model (demutualization) and not corporatization

and the attendant self-listing per se. In this study, we analyze the performance of mutual,

demutualized but customer-controlled, and publicly-traded exchanges to ascertain

whether improvements in performance are due to commercialization of the entity’s

objective and the attendant change in business model from mutual to profit maximization

(demutualization) or to corporatization.

       We show that both demutualized but member-owned exchanges and publicly

traded exchanges exhibit higher levels operating efficiency than mutually-owned

exchanges. The profit-motive that emanates from exchange commercialization allows the

managers to operate more efficiently. Second, we compare the performance of

corporatized exchanges with that of demutualized exchanges and find that the operating

performance of corporatized exchanges is better than that of exchanges that demutualize

but remain member-owned, thus suggesting that corporatization as a governance device

that leads to increased operational efficiencies and profitability. However, our robustness

test reveals a puzzle. Although self-listed exchanges perform better than demutualized




                                                                                         4
exchanges, when we compare the post-corporatization performance of listed exchanges to

their pre-corporatization (demutualization) performance we do not find evidence of

significant incremental gains in efficiency and profitability for the exchanges that

corporatize their structure. This finding suggests that the profit-mandate that accompanies

demutualization provides sufficient freedom to fully exploit ‘monopoly’ rents before

corporatizing. Therefore, the improved operating performance of self-listed exchanges

over that of demutualized exchanges perhaps is not due to corporatization and market

monitoring per se. What then motivates exchanges to corporatize their structure? Further

analysis shows that corporatization allows investors to harvest their investments and earn

significant returns in excess of market benchmarks. By going public, the true value of

owners’ investments can be determined in the market place.

       This paper contributes to existing literature in a number of ways. We analyze the

performance of stock exchanges in a continuum of exchange governance structure,

namely, mutual, demutualized but customer-controlled, and publicly traded exchanges.

The research design allows us to address the question of whether self-listing is the

ultimate solution to competition problems or whether a demutualized structure can help

improve the efficiency and performance of these enterprises even if they remain under

private ownership. This analysis is important and has important policy implications

because self-listing involves significant costs including the threat of takeover which can

sometimes be distracting for management (Otchere, 2006). This point is well-illustrated

by the reoccurring hostile bids for the London Stock Exchange (LSE).1 The results of this

study will help other exchanges determine if eventual IPO is warranted. Beyond the value


1
 The hostile yet unsuccessful bids for the LSE include takeover attempts by OM Gruppen in 2000,
Deutsche Bourse in 2004, Macquarie Bank in 2005, and NASDAQ in 2006 and 2007.


                                                                                             5
of shedding light on the question of whether performance improvements could be

achieved without corporatization, our study also contributes to the organizational

structure and performance literature. Our sample and test design provide a setting to test

the ownership-structure hypothesis after controlling for the effects of competition. We

find evidence consistent with the assertion that structure affects performance.

       The remainder of the paper is organized as follows. In Section 2, we discuss the

theoretical underpinnings of the exchange governance continuum that informs the

discussion of the difference in performance of exchanges. A review of the sample, data

and methodology is presented in Section 3. Testable hypotheses are discussed in Section

4. The results are presented and discussed in Section 5, while summary and conclusion

are provided in Section 6.


2.0    The Exchange Governance Continuum

Extant literature has put forth a number of hypotheses to explain the demutualization of

stock exchanges, but two interrelated theories are particularly useful in explaining the

current trend of exchange commercialization. The first, economic disturbance theory,

suggests that business strategies and organizational structures adapt over time as

exogenous economic shocks disturb the competitive environment (Otchere, 2006).

Shocks to the exchange industry have reshaped both how exchanges compete and with

whom they compete. Recent shocks that have affected the industry include globalization,

technological developments, and the attendant growth in electronic communication

networks (ECNs) which have heightened the level of competition for order flow among

exchanges.




                                                                                        6
         The development of sophisticated trading capabilities has also literally changed

the face of stock markets, with electronic trading platforms replacing the traditional,

physical exchange floors. In the developed world, virtually all exchanges have now

replaced trading floors with electronic trading systems, which comes as no surprise given

the well-documented improvements in efficiency when exchanges switch to electronic

trading (see Bessembinder & Kaufman, 1997 and Domowitz & Steil, 1999).

Advancements in communication, information and network technology have also

facilitated the break-down of barriers to entry and have further increased competition

among stock exchanges and electronic communication networks (ECNs). ECNs offer

investors expanded trading opportunities and undermine the exchanges’ traditional

monopolistic stronghold by offering speedy and cost-efficient services made possible by

technologically-advanced electronic trading platforms. These developments have forced

exchanges to change their governance structure in order to adapt better to changing

operating conditions.

         The second driver of exchange governance conversion relates to the efficiency

motive, an extension of agency theory. The efficiency motive suggests that firms will

operate under the most efficient governance structure, one that minimizes contracting and

other costs because an efficient organizational form leads to a more competitive firm.2

Based on the theory of economic disturbance and the pursuit of economic efficiency,

three forms of stock exchanges governance structures are currently being used by

exchanges, namely, mutually-owned cooperative structure (Stage 1), demutualized but

privately-held structure (Stage 2), and publicly-held structure (Stage 3). The primary

2
 Contracting costs associated with exchange governance structures include i) the cost of paying monopoly
prices for trading services and ii) the cost of under-consumption of trading services due to high prices set
by the exchange.


                                                                                                           7
distinguishing features of the exchange structure continuum are better visualized in

Figure 1. An exchange is deemed to be mutual and ‘member-owned’ if the rights of

members are coupled with ownership rights. A demutualized exchange is one whose

members’ rights have been decoupled from ownership rights, and the exchange has taken

on commercial, profit-oriented objectives. An exchange is considered self-listed if it has

demutualized and its shares are publicly-listed. In the following section, we discuss what

motivates exchanges to move along the governance continuum.


2.1    Mutual Ownership

Stock exchanges have traditionally been organized as mutual entities. Under this

governance structure, the exchange is owned only by members, so ownership and the

right to consume the services of the exchange are bundled (Hart and Moore, 1996; Di

Noia, 1998). It is the members, and therefore customers, who have the residual rights of

control over the exchange (Hart and Moore, 1996). The mutual form of organization has

prevailed for centuries because exchanges were able to maintain their monopoly status.

Hitherto, the state of technological development restricted communications across

physical distances and curbed any real competition among exchanges. Therefore, until

more recently, the mutual structure has been the most efficient and the favored

governance structure.

       As the operating environment is changing, the benefits of mutual ownership are

diminishing, and exchanges are increasingly abandoning the traditional mutual ownership

model. The change can be attributed to a number of factors. Over the past decade,

competition has increased sharply. In a more competitive environment, exchanges are

forced to operate with much more efficiency and flexibility, but the user-owned nature of



                                                                                        8
the mutual form constrains management’s ability to respond to the needs of competition

(Domowitz and Steil, 1999; Hart and Moore, 1996). In fact, empirical evidence (e.g.

Krishnamurti, et al., 2003, and Otchere, 2006) suggests that mutual ownership does not

optimize operational performance of the exchange, hence the recent trend among stock

exchanges to convert to a demutualized, profit-oriented structure.


2.2    Demutualization

Demutualization involves replacing mutual ownership of seats on the exchange with

shares. Inherent in demutualization is a fundamental shift in the exchange’s raison d’être.

The exchange becomes a for-profit (commercial) entity, a dramatic departure from the

not-for-profit mandate of the mutual form. The demutualized exchange remains a private

entity, with a clear separation between the trading rights (now owned by the exchange)

and ownership rights (Aggarwal, 2002). From an efficiency perspective, exchanges are

increasingly seeking an organization structure that will allow them greater freedom to

respond to the challenges they face and enhance the profitability of the firm. Prior

researchers have identified and modeled the benefits of privately-held, profit-oriented

exchange to include increased efficiency (Hart and Moore, 1996; Serisfoy, 2007) and

increased competitiveness through technological sophistication (Domowitz and Steil,

1999; Krishnamurti, et al., 2003). The literature suggests that the mutual structure no

longer fits the current economic environment. Demutualization, with the attendant

decoupling of ownership and trading rights, provides managers the flexibility to compete

effectively; hence the growing trend in exchanges demutualization (see Figure 2).




                                                                                         9
2.3      Self-Listing

A growing number of exchanges have corporatized their structure by issuing shares to the

public. Most of these exchanges have conducted initial public offerings, with the notable

exception of the NYSE which became publicly-traded exchange in a backdoor listing

transaction through its acquisition of Archipelago Exchange. One important benefit of

self-listing is that managers of the publicly-listed exchange are exposed to the discipline

of the market for corporate control and the scrutiny of market participants. As suggested

by Otchere (2006), public-listing increases the market value of the business because it

provides liquidity for the shares and increases exposure of the exchange to analyst and

institutional investor scrutiny and this facilitates efficient pricing of the exchange owners’

equity positions. Furthermore, the listed exchange has an expanded set of choices for

raising capital in the future. The few studies that analyze the performance of listed

exchanges (e.g., Aggarwal and Dahiya, 2006; Otchere, 2006; Serisfoy, 2007) document

significant improvements in the financial and operating performance of these exchanges.

These findings and the increasing trend of corporatization perhaps suggest that the

benefits of self-listing outweigh the costs of going public.3



3
  There are two important costs associated with self-listing. First, by self-listing, the exchange exposes itself
to the market for corporate control. The market for corporate control typically acts as a positive disciplinary
force, incentivizing management to maximize firm value. However, in the current environment of levels of
consolidation in the industry, the market for corporate control is seen as a serious threat to the
independence of even the best-run exchange, and the attendant unrelenting pressures from bidders distract
exchange managers. While most exchanges have at least a 5% ownership restriction bylaw (Aggarwal,
2006), the potential for takeover is a real cost of going public, as exemplified by the reoccurring hostile
bids for the London Stock Exchange (Otchere, 2006). Second, self-listing creates (perceived or real)
conflicts of interest due to the exchange’s position as a market regulator. In its 2006 Consultation Report,
the International Organization of Securities Commissions (IOSCO) identifies instances where conflicts of
interest are heightened when an exchange self-lists. One key issue raised by the exchange corporatization is
the ability of a self-listed exchange to effectively act as its own regulator. Researchers including Lee (2002)
and Otchere, (2006) however, argue that such conflicts of interest are not of great concern given that
market integrity is such a valuable asset for exchanges that they will act in a manner that is consistent with
best practices.


                                                                                                             10
4.     Testable Hypotheses

Following from the foregoing discussion, we argue that during periods of low levels of

competition, exchanges with these governance structures would adopt similar rent

seeking behavior as there would be no significant incentives for any of these exchange

structures to adopt distinctly different strategy. Consequently, one form of exchange

ownership might not achieve any significantly better performance compared to the others.

However, in an environment where competition is keen, Parker and Hartley (1991) argue

that ownership structure matters and that efficiency gains might lie in ownership status.

In the current environment of increased competition in the exchange industry and the

migration of order flow, demutualization and the attendant change of business model

(commercialization) can provide incentives for management to be efficient. We

hypothesize that demutualized exchanges will be more efficient than mutual exchanges.

Formally, hypothesis 1 is stated as


Hypothesis 1: The operating efficiency of the privately-held, demutualized exchange will
      be better than that of the mutually-owned, cooperative exchange.


Drawing on Parker and Hartley (1991), one can be argued that given profit maximization

as the objective function, when both demutualized but private exchange and publicly

traded exchanges are exposed to the same competitive pressures and market signals, they

would be expected to yield similar performance in terms of allocative efficiency,

regardless of their ownership structure. However, we argue that publicly traded

exchanges will be able to respond to competitive pressures better than demutualized but

privately held exchanges because of the complete decoupling of ownership from control

which gives managers of self-listed exchanges the freehand to pursue profitable ventures.



                                                                                      11
So long as the demutualized exchange is owned by the same owners/members and,

therefore subject to the same interference as mutual or cooperative exchanges, there may

be some hindrance to the efficient operation of the exchange and the coupling problem of

membership and control associated with mutual exchanges will not be fully addressed by

demutualization. For one thing, managers of privately-held firms are incentivized by the

profit motive that characterizes demutualization but to a lesser extent than managers of

publicly-held firms who are constantly scrutinized by the market; hence, demutualized

exchanges may not be able to compete as efficiently as their public traded counterparts.

In addition, though managers of demutualized exchanges have much more operating

freedom than before, the exchanges are still restrained in their growth potential by their

limited access to capital. Based on the foregoing, we expect the demutualized but

customer-owned exchanges to underperform their publicly-traded counterparts.4

Formally, we state hypothesis 2 as follows:


Hypothesis 2: The operating performance of the publicly-held, self-listed exchange is
better than that of demutualized, privately-held exchange.


4.0     Data and Methodology

4.1     Data

This study focuses on the three governance structures used in the exchange industry as

outlined in Figure 1. Demutualization and corporatization among WFE members have

increased significantly in recent years. The changing trend in the exchange governance


4
   For listed exchanges, the firm could still be owned by the former owners and other investors, but the
trading of the firm’s shares on the stock exchange and the attendant monitoring of performance by market
participants will compel managers to respond positively to market monitoring with improved performance
(Otchere, 2006).



                                                                                                     12
structure is better visualized in Figure 2. In the 1990s, almost all the WFE members were

mutual exchanges. In recent years, the number of mutual exchanges has fallen markedly,

while the number of profit-oriented exchanges has increased significantly. For the

purpose of analyzing the performance of exchanges, we formed three portfolios of

exchanges to correspond with the three stages of exchange governance structure. Our

sample consists of 12 mutually-owned exchanges (Stage 1), 12 demutualized but

privately-held listed exchanges (Stage 2), and 21 publicly-listed exchanges (Stage 3).5

The exchanges in each portfolio are listed in Appendix 1.

        For the portfolio of listed exchanges,6 three years of pre-and post-event data is

collected around two event dates: date of demutualization and date of listing. For the

portfolio of demutualized exchanges, three years of pre-and post-event data is collected

around the date of demutualization. For the portfolio of mutually-owned exchanges, three

years of data is collected, and the observation window is selected based on a process of

matching mutually-owned exchanges to demutualized and listed exchanges. We include

firms for which we have at least two years of pre- and two years of post-event data. This

process resulted in a total of 186 firm-year observations. The operating performance data

and stock return data come from Datastream and the exchanges’ prospectuses. The

5
  The number of publicly-listed exchanges is higher than the demutualized and mutually-owned exchanges
sub-samples because while all publicly-listed exchanges report financial data, only some demutualized and
mutually-owned exchanges voluntarily report financial data and this restricts the sampling frame for the
two non-public forms of exchange governance.
6
  The sample is not limited to stock exchanges but it includes derivative exchanges that have also changed
their governance structure. Although they have different product lines and different economics, these
exchanges perform similar roles. Both stock and derivative exchanges perform the role of price discovery,
risk management and reduction of transaction costs (Otchere (2006). They are exposed to similar industry
shocks and have responded to these shocks with changes in their governance structures. Whilst these
exchanges may have different economics, for example, futures exchanges have proprietary products given
the nature of futures clearing and the equity and options trading are cutthroat commodity business with very
narrow profit margins, an examination of the changes in performance will not be affected by the nature of
the product.



                                                                                                        13
product market data come from WFE website, and macroeconomic data come from

International Financial Statistics.

         Summary statistics for the three groups of exchanges are presented in Table 1.

The statistics show that the performance of the sample increases in the degree of

corporatization. Publicly traded exchanges are larger and more profitable than

demutualized exchanges, and demutualized exchanges are also larger and more profitable

than mutual exchanges. The listed exchanges generate more revenues and net income

than demutualized and mutually-owned exchanges.7 In summary, revenue, net income

and turnover velocity follow a similar pattern, increasing with the degree of

corporatization, except ROA which is higher for the mutual exchanges.

                                    [Fix Table 1 here]

4.2      Methodology

The aim of this study is to document changes in operating performance at each stage of

the exchange governance continuum. We examine several dimensions of operating

performance. First, extant literature suggests that demutualization and self-listing provide

managers the freehand to pursue profitable business ventures (see Hart & Moore, 1996;

Krishnamurti, et al., 2003 for demutualized exchanges; and Otchere, 2006 and Serisfoy,

2007 for self-listed exchanges). We examine return on assets (ROA), return on equity

(ROE), and net income margin to ascertain differences in profitability and efficiency in

the use of assets exist among the three organizational structures along the exchange

governance continuum especially between demutualized and self-listed exchanges, and

less so for mutual exchanges as the objective of the latter is not-for-profit. We analyze

7
  The fact that demutualized exchanges are more profitable than mutual exchanges is not surprising since
by definition, mutual exchanges are generally not-for-profit entities. In fact, a mutual exchange may be just
as profitable but the profits may be passed on to the members through lower fees.


                                                                                                         14
operating efficiency ratio (total revenue/total expenses) to determine whether the

operating efficiency of exchanges changes along the exchange governance continuum.

Second, improvements in technology have changed the cost structure of the exchanges,

with the costs of an additional trade and of listing a company approaching zero (Lee,

2002). We examine the ratio of traditional sources of revenue (listing, trading, and data

dissemination) to total revenue with the view to documenting any changes in these

sources and the exchanges’ ability to adjust to changing cost and revenues sources. We

test the difference in the median ratios using the Wilcoxon matched-pair signed rank test

statistics and the difference in mean ratios using the matched-pair means test. The

measures of operating performance and the predicted sign are summarized in Appendix 1.


5.0    RESULTS

5.1    Demutualization versus mutual ownership

We expect that following demutualization and the adoption of a profit motive, managers

of demutualized exchanges will have greater freedom to pursue profitable business

opportunities and become more efficient. When we compare the operating performance

of the demutualized exchange sample to that of mutual exchanges, we find evidence that

demutualization improves the exchange’s operating efficiency (see Table 2). The

demutualized exchanges mean ROA, ROE, and net income margin are 1%, 4% and 5%

respectively better than those of the mutual exchanges. These differences are significant

at 5%. As argued earlier, it is not surprising that demutualized exchanges are more

profitable than mutual exchanges since profit motive is absent for mutual exchanges

whose mandate is to serve the members’ interests. Nonetheless, the demutualized

exchanges are much more efficient than mutual exchanges, with the former attaining a



                                                                                      15
level of operating efficiency that is 21% higher than that of their mutually-owned

counterparts.

                              [Fix Table 2 here]

5.2    Self-listing versus demutualization

Going public enables the exchange to gain access to the public capital markets, allows the

firm’s performance to be monitored by analysts and institutional investors, and exposes

the exchange to pressures from the market for corporate control which in turn elicit

higher levels of management discipline. We conjecture that these capital market effects

can lead to an increase in the operating performance of the corporatized exchange. As a

result, we expect that listed exchanges would exhibit superior operating performance.

Given the counterfactual situation of not knowing what the performance of the publicly

traded exchanges would have been had the corporatization not occurred, we use the

sample of demutualized but privately-owned exchanges as a benchmark to determine

whether corporatization generates incremental benefits beyond what demutualization

generates. Any significant difference in performance of the listed exchanges would be

attributed to corporatization and the effects of capital market monitoring.

       The results of the analysis are reported in Table 3. Panel A shows the mean ratios

whereas Panel B shows the median ratios. We observe that the corporatized exchanges

outperformed demutualized exchanges on the profitability and efficiency measures. The

ROE of publicly traded exchanges of 18.6% is significantly better than that of

demutualized exchanges of 17.1% (t-statistic for the difference is 2.47). Similarly, the

operating efficiency of corporatized exchanges is better than that of demutualized

exchanges. However, their rate of growth in capital expenditure lags behind that of




                                                                                       16
demutualized exchanges. Listing revenues generated by the corporatized exchanges

continues to be a significant component of their total revenue than that of the

demutualized exchanges. While a similar pattern is observed for the median ratios, the

difference in performance is not as significant as the mean results. In sum, consistent with

our hypotheses 1 and 2, we find evidence of performance improvements along the

exchange governance structure continuum, i.e., corporatized exchanges perform better

than demutualized exchanges and demutualized exchanges in turn perform better than

mutual exchanges.

                                      [Fix Table 3 here]

5.3     Robustness tests

We perform two additional tests to ascertain the robustness of our results. The first test

focuses on the pre- (mutual) and post-demutualization period operating efficiency of the

sub-sample of exchanges that have demutualized but remained private. The second

robustness test focuses on the listed exchanges sub-sample. This sub-sample has gone

through the three phases as mutual exchanges, demutualized exchanges, and then self-

listed exchanges presents a natural laboratory to test the robustness of the results based on

the three samples of mutual, demutualized and publicly-listed exchanges. We expect to

observe improvement in operating efficiency and profitability as the exchange moves

along the continuum from mutual to demutualized and self-listing.


5.3.1   Pre- and post-demutualization performance of demutualized exchanges

To check the robustness of our results that demutualized exchanges perform better than

mutual exchanges in terms of operating efficiency, we compare the pre- and post-

demutualization performance of the demutualized exchanges to ascertain whether the



                                                                                          17
performance of this sub-sample improved when they demutualized. Recall that we expect

the post-demutualization operating efficiency of this sub-sample to be better than the pre-

demutualization performance because the change in business model (commercialization)

will compel the firm to become more efficient once these exchanges adopt a profit

motive. The results of this test are presented in Panel A of Table 4 which shows the pre-

and post-demutualization performance of the demutualized exchanges. Consistent with

our earlier results, the performance of this sub-sample improved significantly when

exchanges demutualized. The most marked improvements attained by this sub-sample

when it demutualized are in the profitability measures. The mean ROA almost doubled,

whereas ROE, and net income margin improved by about 61% over the pre-

demutualization performance (all of these are statistically significant at conventional

levels).

                                [Fix Table 4 here]

5.3.2      Performance of self-listed exchanges along the exchange governance continuum

Recall from the analysis above that the performance of stock exchanges improves as the

exchanges moves along the continuum. The corporatized exchanges present a setting for

us to test the robustness of our results that the operating performance of exchanges

improves along the governance continuum. We ascertain the robustness of our results

pertaining to hypothesis 2 by analyzing the performance of publicly traded exchanges

along the governance continuum to ascertain whether the performance the sub-sample at

the corporatized phase > demutualized phase > mutual phase. The results presented in

Panel B Table 4 show that the listed exchanges have achieved remarkable improvements

in profitability, efficiency and growth in investments upon demutualization. The




                                                                                        18
magnitudes of the change is worth noting: the mean ROA improved by 74%, mean ROE

increased by 72%, the median net income margin rose by 62% and the median operating

efficiency by 83%. These results add support to the finding that by replacing the

members’ interest mandate with profit maximizing objective, the exchanges are able to

pursue profitable business ventures that enhance the profitability and efficiency.8

         The evidence documented earlier also shows that the sample of publicly traded

exchanges performs better than the sample of demutualized but private exchanges. A few

studies have also documented significant improvements in operating performance of self-

listed exchanges. As argued earlier, the source of the profitability and efficiency gains is

not clear. If the exchanges operated with a profit motive even in the hands of members,

then the anticipated gains after corporatization could be minimal. This is because the

corporatized exchanges will have less room to realize gains from unexploited

opportunities if the firm operated in a relatively entrepreneurial fashion before going

public. Thus, there would be no significant difference in post-listing profitability and

operating efficiency once the exchange subsequently goes public. To provide robustness

for our earlier results, we examine the operating performance of the corporatized

exchanges when they self-list with their performance during the demutualization phase.

         Focusing on the post-listing phase and demutualization phase performance of the

self-listed exchanges (see Table 4), we find no significant improvements in performance

of the exchanges that corporatize their structure and become publicly traded. Only the

ratio of trading revenue-to-total revenue improves when the exchanges become publicly

traded. The evidence is consistent with the assertion that stock exchanges that


8
 The profitability results should be interpreted with caution in light of our earlier discussion about the
mandate of mutual exchanges.


                                                                                                             19
demutualize and operate competitively before self-listing tend to yield less incremental

operating efficiency gains when they subsequently corporatize. It appears that such

exchanges fully exploit monopoly rents in the product market prior to corporatization.


5.4    Performance of demutualized exchanges and self-listing exchanges during
       the demutualized phase

We have shown that the operating performance of the listed exchanges is better than that

of demutualized but customer-owned exchanges. However, the robustness test based on

the self-listed exchange sample that has passed through the three phases shows surprising

results in the sense that the post-self-listing performance of this subsample is not

statistically different from the pre-self listed (demutualized phase) performance. The

question that we explore here, with the view to reconciling this inconsistency, is how

different are the exchanges that go on to corporatize their structure from those

demutualized exchanges that remain private? To answer this question, we compare the

performance of the corporatized exchanges at the time of their demutualization with that

of exchanges that are, or have remained, demutualized but private. The results of the tests

are presented in Table 5. Interestingly, we find that the sub-sample of exchanges that

subsequently become publicly listed exchanges exhibits significantly higher levels of

profitability and efficiency than those that remain demutualized. Also, their growth in

capital expenditure is much higher than that of exchanges that remain demutualized,

perhaps suggesting that exchanges that go on to corporatize equip themselves with the

tools of competition prior to going public. Exchanges that corporatize also tend to

generate more listing and information revenue but generate less income from trading

revenue than exchanges that remain demutualized.




                                                                                         20
                              [Fix Table 5 here]

       These findings and the results reported in Table 4 suggest that though self-listed

exchanges do not exhibit significant incremental operating performance improvements

after they subsequently corporatize, they attain high levels of performance during the

demutualization phase. It may be that managers of these exchanges intending to go public

prepare the firm for the rigors of the market at the demutualization phase. We surmise

that the exchanges that proceed to self-list are those that have already achieved improved

performance at the demutualization phase and this increases managers’ confidence in the

exchange’s ability to compete as a publicly-listed entity. Either way, the exchanges that

go on to corporatize are distinctly different (at least in terms of operating performance)

from their peers that remain demutualized but private.


5.5.   Cross-sectional regression analysis

In this section, we ascertain the impact of the change in the governance structure on the

exchange performance in a more robust manner by performing cross-sectional regression

analysis of operating performance using the following regression.

       Operating performanceit = αi + βi ExchangeDumit + λ Controlit + εit

where exchange dummy is the governance structure dummy and it is either LIST

(dummy), or DEM (dummy) or both. The DEM (dummy) takes on a value of 1 if the

exchange is demutualized and zero, otherwise. Likewise, LIST (dummy) takes on a value

of 1 if the exchange has corporatized and self-listed, and zero otherwise. Control consists

of variables that extant literature suggest can affect performance of exchanges and it

includes total assets (a proxy for size), number of firms listed on the exchange, trading




                                                                                        21
volume (dollar value and number of shares), and country-specific characteristics.9 We

include various revenue sources as independent variables in the profitability regressions

because they affect the profitability of exchanges. The figures are denominated in US

dollars. In all the regressions, the variables of interest are LIST (dummy) and DEM

(dummy). The regressions are based on firm observations of the subsamples at the

respective position on the continuum. The results of the regressions are presented in

Table 6. Panel A shows the results for the combined sample, Panel B presents the

regression results for the demutualized and self listed exchange subsamples during the

demutualization period; Panel C shows the results for the demutualized exchanges and

the self-listed exchanges subsample, while Panel D shows the results for the self-listed

exchanges over the three phases of the governance continuum.10

         We find that for the combined sample of mutual, demutualized and self-listed

exchanges, the demutualization dummy is positive and significant in the ROA and

operating efficiency regressions. This confirms the earlier results that demutualized

exchanges are more efficient. The coefficient of the DEM (dummy) is also positive in the

trading revenue regression, suggesting that demutualized exchanges derive more income

from trading revenue than mutual exchanges. Interestingly, the demutualization dummy

coefficient is negative in the listing revenue regression. This result is consistent with the

suggestion that because of the desire to entice more firms to list on the exchanges and

attract order flow, profit-motivated exchanges offer incentives for firms to list on their




9
  Country specific variables account for variability due to the country’s legal system, regulatory system,
level of economic activity, and level of development.
10
   The results are based on time series cross sectional data. Though not shown here, there is a significant
variability in the data from year to year; hence the t-statistics are not biased.


                                                                                                              22
exchanges (Otchere, 2008). The mutual exchanges however, continue to rely significantly

on listing revenue.

       The LIST (dummy) is positive and significant in the ROE, net interest margin and

operating efficiency regressions. This finding also confirms our earlier univariate results

that corporatized exchanges generate incremental improvements in profitability and

operating efficiency as compared to the non-corporatized exchanges. In terms of product

market variables, the coefficient of LIST (dummy) is significantly positive in the listing

revenue and significantly negative in the trading revenue regression, suggesting that

corporatized exchanges generate more income from listing revenue than customer-

oriented exchanges but less from trading revenue once they go public. In Panel B, we

estimate the regression for the demutualized and the corporatized sub-samples at the time

of demutualization. The LIST (dummy) which takes on a value of 1 if the exchange is

self listed and zero otherwise, is significantly positive in the operating efficiency, listing

revenue and information revenue regressions. This suggests that when the listed

exchanges were demutualized, they were more efficient and also generated greater

income from these sources than the exchanges that are currently demutualized.

                               [Fix Table 6 here]

       Turning to the results for the demutualized and self listed sub-samples presented

in Panel C of Table 6, we find that the LIST dummy is positive and significant in the

ROA, ROE, and operating efficiency regressions. These results confirm our earlier

findings that the corporatized exchanges perform better than their demutualized but

private counterparts. Lastly, for the corporatized exchanges themselves that have gone

through the three phases, we find that the demutualization dummy is significant in the




                                                                                           23
profitability and operating efficiency regressions, but the LIST dummy is generally not

significant, thus confirming our earlier results of improved performance at the

demutualization stage and no significant improvement in operating performance when

these exchanges subsequently self-list (the dummy is only significant in the net income

margin regression. Since there are economies of scale in stock exchange activities (see

Malkamaki, 1999), and also given the size differential between the listed and non-listed

exchanges, it is possible that non-linearity could lead to specification problems with the

linear OLS model. For robustness test purposes, we re-estimated this regression as a

quadratic function on the size variable, Ln(Total Assets), and this regression yielded very

similar results.


5.7.    Corporatizing financial markets: the role of corporate governance, incentives
        and market monitoring on performance

Our analysis of the pre- and post-listing operating performance of corporatized stock

exchanges shows that the self listed exchanges’ operating performance does not

significantly improve when they convert from a demutualized structure to a corporatized

structure. If corporatization does not generate substantial improvements in performance,

what is it that propels the exchanges to corporatize their structure given the significant

direct and indirect costs of going public, including the threat of takeover? Drawing on

Aron’s (1991) moral hazard argument, we offer a reason why exchanges might want to

corporatize. Demutualized exchanges with many investment opportunities or growth

options cannot provide sufficient incentives for managers by distributing either shares or

stock options. Efficient incentives can be provided when shares in the firm are publicly

traded. For corporatized exchanges, management is evaluated on a daily basis through the




                                                                                        24
company’s stock price. This immediate visible evaluation can boost performance as

managers, recognizing the scrutiny of the market, are likely to make important strategic

decisions that focus on appropriate value drivers, and be rewarded through stock options.

           Furthermore, although demutualization allows exchange owners to convert their

seats to shares, the true value of the shares is difficult to ascertain because there is no

formal market for the privately-held shares (Chesini, 2001). Going public provides

owners with a mechanism to unlock the true value of their investment and increase the

liquidity of their holdings. In this section, we evaluate the long run buy-and-hold

abnormal returns (BHARs) and cumulative abnormal returns (CARs) returns to the

corporatized exchange shareholders using the market adjusted model, with the main

index of the country as the benchmark index.11 The results, which we present in Table 7,

show quite convincingly that upon listing there are significant financial benefits that

owners of the exchange can realize. No matter the holding period examined, from one

year to five years, we find that investors earn economically and statistically significant

returns in excess of the market benchmarks. Based on the buy-and-hold returns,

shareholder of listed exchanges earned over 130% market-adjusted returns in the third

year of corporatization. These results are consistent with those of Otchere (2006) and

Mendiola and O’Hara (2003) who use a much smaller sample. The results provide

compelling evidence that there are substantial value discovery effects resulting from

corporatization.

                                    [Fix Table 7 here]

           Having documented that the corporatized exchanges have realized significant

excess returns, we attempt to link the stock market performance of these exchanges to
11
     To conserve space, the BHAR and CAR estimation procedures have been omitted.


                                                                                        25
corporate governance, compensation and monitoring variables. We use measures that

prior research suggests affect stock market performance. We estimate a regression of the

3-year CARs on the variable component of executive compensation, fixed component of

executive compensation,12 board size, board effectiveness as measured by the frequency

of attendance at meetings, and the volume of shares traded on the exchange. We use the

percentage of institutional investors and foreign trade as a crude proxy for monitoring

since data on the number of analysts were not available. The proportion of shares owned

by institutional holders is used as proxy for monitoring because if these institutional

investors hold a significant number of the shares of a firm, they will have strong

incentives to monitor the performance of the firm. We include volume of shares traded as

an independent variable because the revenues and therefore performance of the exchange

will be influenced by the volume of trading activities. The proportion of variable

compensation (including stock options) is included in order to ascertain whether the

performance of the exchange is positively related to performance-based incentive

measures. We include board size and proxies for board effectiveness as control variables

as prior research shows that these variables affect performance of firms. The results of

the regression are presented in Table 8.

                         [Fix Table 8 here]

We find that the coefficient of executive compensation, in particular the variable

component is significant in most of the regressions. The variable component (comprising

primarily of stock options) is positively related to the long run stock market performance.

As expected, trading volume positively affects stock returns; the greater the volume of


12
  Fixed and variable components do not add up to one because there are other components of
compensation that are not classified under these categories.


                                                                                             26
trade, the greater the trading revenue and cetaris paribus, the greater the returns. Our

proxies for monitoring, namely, institutional investors holding has the correct sign and is

significant, suggesting that the greater the percentage of institutional shareholding, the

greater the incentive to monitor the firm and hence, the greater the performance.

However, this variable loses its significance once we exclude our crude proxy for the

trading activities of foreign investors. The goodness of fit for the regression is high with

an adjusted R-squared of 96%. As a robustness test, we re-estimated the regression using

the five year returns and the results are qualitatively similar although the number of

observations is smaller. Overall, the results reported in this section suggest that the

economically significant long run returns realized by corporatized exchanges are related

to governance, executive compensation and monitoring variables.


6.     Summary and conclusion

Technological development and globalization have created a new operating environment

for exchanges, one that questions the effectiveness of the traditional mutual governance

structure. An increasing number of exchanges have therefore converted their mutual, not-

for-profit structure to one with a profit maximization mandate. Some exchanges have

taken a further step, fully corporatizing and listing their shares on the public market. Few

prior studies provide evidence that corporatization of the exchange’s mandate

significantly improves performance. It is not clear whether the performance

improvements are due to corporatization or the adoption of the profit motive when the

exchange demutualizes. We examine whether corporatization of the exchange is

necessary to improve the performance of the exchange.




                                                                                         27
       By employing a series of tests that compare operating performance of exchanges

along the governance continuum, we document several findings. First, we find that both

demutualized but member-owned exchanges and publicly traded exchanges exhibit

higher levels of profitability and operating efficiency than mutually-owned exchanges. As

to whether corporatization allows the exchange to generate incremental gains in

profitability and efficiency beyond what can be obtained under a demutualized structure,

we find that consistent with our hypothesis, publicly traded exchanges exhibit superior

operating performance than demutualized exchanges, thus suggesting that corporatization

is a governance device that leads to better operating performance. We perform robustness

test of our results using the corporatized exchanges that have gone through the three

phases of the governance continuum and compare their post-corporatization performance

to their pre-corporatization (demutualization) performance and find that the corporatized

exchanges do not exhibit evidence of significant incremental gains in efficiency and

profitability. Interestingly, however, we find evidence that listed exchanges achieved

higher levels of operating performance during the time they were demutualized than

exchanges that have remained demutualized. This suggests that perhaps the profit-

mandate associated with demutualization provides sufficient incentives to fully exploit

‘monopoly’ rents before corporatizing.

       Given that corporatization does not generate any significant incremental gain in

profitability and operating efficiency beyond what they realized at the demutualization

phase, we explore the question of what motivates exchanges to corporatize and subject

themselves to the discipline of the market for corporate control. We conclude that

corporatization, among others, brings about proper equity valuation of the exchanges’




                                                                                      28
franchise as the trading of the exchanges’ shares on the public stock exchange provide

owners with significant long-run excess returns. In sum, the profit mandate introduced by

commercialization provides managers with sufficient flexibility to compete, but the

prospect of going public incentivizes management to more aggressively improve

performance prior to listing. Corporatization, while not necessarily driven by promises of

operating performance improvements, appears to be sought by exchanges because of the

auxiliary benefits of listing such as proper valuation of the equity.




                                                                                       29
                                     References
Aggarwal, R., 2002. Demutualization and corporate governance of stock exchanges.
      Journal of Applied Corporate Finance 15(1), 105-113.

Aggarwal, R., Dahiya, S., 2006. Demutualization and public offerings of financial
      exchanges. Journal of Applied Corporate Finance 18(3), 96-106.

Aron, D.J., 1991. Using the capital market as a monitor: corporate spinoffs in an agency
       framework. Rand Journal of Economics 22(4), 505–518.

Bessembinder, H., Kaufman, H., 1997. A comparison of trade execution costs for NYSE
      and NASDAQ-listed stocks. Journal of Financial and Quantitative Analysis 32 (2)
      287-310.

Chesini., G., 2001. Changes in the ownership structure of stock exchanges: From
       demutualization to self-listing. Paper presented at the 10th International “Tor
       Vergata” Conference on Banking and Finance: Competition, Financial Integration
       and Risks in the Global Economy, University of Rome.

Di Noia, C., 2000. Customer controlled firms: The case of stock exchanges, EFMA
      Working Paper.

Domowitz, I., Steil, B., 1999. Automation, trading costs, and the structure of the
     securities trading industry. Brookings-Wharton Papers on Financial Services, 33-
     81.

Elliott, J., 2002. Demutualization of Securities exchanges: A regulatory perspective. IMF
         Working Paper WP/02/119.

Hansmann, H., 1994. The ownership of enterprise, New Haven, CT, Yale University
     Press.

Hart, O., Moore, J., 1996. The governance of exchanges: Members’ cooperatives versus
       outside ownership. Oxford Review of Economic Policy 12(4), 53-69.

International Organization of Securities Commissions Technical Committee, 2006.
        Regulatory issues arising from exchange evolution. Consultation Report.

Krishnamurti, C., Sequiera, J.M., Fangjian, F., 2003. Stock exchange governance and
       market liquidity. Journal of Banking and Finance 27 (9), 1859-1878.

Lee, R., 2002. The future of securities exchanges. Brookings-Wharton Papers on
      Financial Services, 1-33.

Macey, J., O’Hara, M., 1999. Globalization, exchange governance, and the future of
      exchanges. Brookings-Wharton Papers on Financial Services, 1-32.



                                                                                       30
Mendiola, J., O’Hara, M., 2003. Taking stock in stock markets: The changing governance
      of exchanges, working paper, Cornell University.

Otchere, I., 2006. Stock exchange self-listing and value effects. Journal of Corporate
       Finance 12 (5), 926-953.

Otchere, I., Abou-Zied, K., 2008. Stock exchange demutualization, self-listing and
       performance: The case of the Australian Stock Exchange. Journal of Banking and
       Finance 32 (4), 512-525.

Pagano, M., O. Randl., A.A. Roell, J. Zechner, 2001. What makes stock exchanges
      succeed? Evidence from cross-listing decisions. European Economic Review 45
      (4/6), 770-782.

Parker D, 1991, Privatization. A conceptual framework, in A. Ott and K. Hartley, (eds)
       Privatization and economic efficiency: A comparative analysis of developed and
       developing countries (pp 11-25) Brookfield, VT: Edward Edgar.

Parker, D, and Hartley, K, 1991, Status change and performance: Economic policy and
        evidence, in Ott, and K. Hartley (ed) Privatization and economic efficiency: A
        comparative analysis of developed and developing countries (pp 108-25)
        Brookfield, VT, Edward Edgar.

Pirrong, C., 1999. The organization of financial exchange markets: Theory and evidence.
       Journal of Financial Markets 2 (4), 329-357.

Serisfoy, B., 2007. Demutualization, outsider ownership, and stock exchange
       performance: Empirical evidence. Working Paper Series: Finance and Accounting
       157, Department of Finance, Goethe University Frankfurt am Main.

Subrahmanyam, A., Titman, S., 1999. The going-public decision and the development of
      financial markets. The Journal of Finance 54(3), 1045 – 1082.




                                                                                    31
FIGURES

Figure 1:                          Exchange Governance Continuum

                             Stage 1:                                     Stage 2:                                        Stage 3:

              Mutually-Owned                                      Demutualization /                                 Self-Listing /
                                                                  Commercialization                                Corporatization
 Mandate:
 Member we lfa re ma ximiza-                                    Mandate: Profit ma ximization                  Mandate: Profit ma ximization
 tion
                                                                Seats are exchanged for                        After IPO, shares are freely
 Seats can be traded but the                                    shares in a private place ment                 traded and widely-held
 transaction must be approved                                   and ownership of shares is
 by members                                                     restricted




Figure 2:                          The Evolution of Exchange Governance


                                            Governance Structures of WFE-Member Exchanges

                      120%


                      100%
   % of WFE members




                      80%

                                                                                                                           % mutually-owned
                      60%                                                                                                  % demutualized
                                                                                                                           % publicly-listed
                      40%


                      20%


                       0%
                              1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
                                                                        Years




                                                                                                                                    32
                                                         Table 1:          Descriptive statistics

       This table contains descriptive statistics of the sample. For the portfolio of listed exchanges, data is at year of listing; for the portfolio of
       demutualized exchanges, data is at the year of demutualization; and for the portfolio of mutually-owned exchanges, data is at the year matched
       to the listed exchanges. Columns 1, 2, and 3 display the mean and median financial statement values of Time since demutualization/listing,
       total assets, number of employees, net income, total revenue, ROA, volume of trading on the exchange, market capitalization of firms listed on
       the exchange, and turnover velocity. The next three columns, (2-1), (3-2), and (3-1) show the results from testing the difference in mean and
       median. The t-statistic and z-statistic are used to measure statistical significance for the difference in mean and median tests respectively. The
       symbols ***, ** and * indicate statistical significance at the 1%, 5%, and 10% level, respectively.

                                          Mutual Mean              Demutualized Mean                Self-list Mean                       t-statistic
                                           (Median)                    (Median)                       (Median)                          (z-statistic)
                                              (1)                         (2)                             (3)                  (2-1)        (3-2)           (3-1)
Time since demutualization /
listing                                                                   3.833                        5.000                                (0.90)
                                                                          2.000                        5.000                                (1.29)
Total Assets (US$)                       102,231,255.863             93,879,281.236              1,363,210,973.119            (-0.18)    (4.44)***      (4.38)***
                                         63,869,936.062              71,380,656.270               719,997,919.695              (0.25)    (3.78)***      (3.67)***
Number of Employees                          187.333                     273.800                      823.750                  (0.41)    (2.58)***      (2.61)***
                                             18.000                      106.000                      584.500                  (1.04)     (2.15)**       (2.12)**
Net Income (US$)                          2,410,350.830              11,910,084.425               104,959,949.602              (1.37)     (2.06)**       (2.30)**
                                          1,461,250.967               3,481,217.576                34,258,150.027              (0.55)    (3.18)***      (3.71)***
Total Revenue (US$)                      34,640,963.890              43,823,790.107               555,118,800.446              (0.35)     (2.23)**       (2.27)**
                                         10,389,105.058              20,111,683.229               155,865,000.000              (0.62)    (3.52)***      (3.71)***
ROA                                           0.090                       0.037                        0.060                  (-1.34)       (0.74)        (-0.95)
                                              0.084                       0.043                        0.056                  (-1.06)       (0.66)        (-0.79)
Trading Volume (US$)                  2,689,630,089,071.010      156,262,624,111,492.000     2,816,974,613,721,100.000         (1.45)     (1.97)**       (2.09)**
                                       19,321,730,818.799          516,189,613,605.585        328,600,000,000,000.000        (1.98)**     (2.13)**      (3.33)***
Market Capitalization (US$)           6,175,823,266,627.260        284,440,233,003.828         4,326,927,183,209.160          (-1.41)      (1.88)*        (-0.39)
                                       29,931,432,778.813          54,510,553,874.340           624,109,870,363.240            (0.19)     (2.33)**         (1.48)
Turnover Velocity                             0.495                       0.717                        0.969                   (0.70)       (0.70)        (1.87)*
                                              0.387                       0.326                        0.718                   (0.05)      (-0.79)         (1.59)




                                                                                                                                                                    33
          Table 2:            Difference in performance of demutualized and mutual exchanges
This table contains the results of H1 that the operating performance of demutualized exchanges is better than that of mutually-owned exchanges. Panel A
displays the results of the matched-pair difference in mean test, and Panel B shows the results of the matched-paired Wilcoxon signed rank test.. The t-statistic
and z-statistic are used to measure statistical significance for the difference in mean and median tests, respectively. The symbols ***, ** and * indicate
statistical significance at the 1%, 5%, and 10% level, respectively.


                                                                                                       Listing Revenue    Trading
                        Return on       Return on      Net Income        Operating       Growth in                                   Info Revenue to
                                                                                                            to Total    Revenue to
                         Assets          Equity         Margin           Efficiency       Capex                                       Total Revenue
                                                                                                           Revenue     Total Revenue

 Panel A: Difference in Mean Operating Performance of Demutualized versus Mutual Exchange (Demutualized Exchange Mean - Mutual
 Exchange Mean)
 Demutualized           0.109       0.171        0.295         1.709        1.731         0.087           0.557         0.128
 Mutual                 0.097       0.131        0.241         1.407        0.820         0.122           0.487         0.073
 t-statistics        (2.027)**   (1.989)**     (2.046)**    (2.015)**      (0.557)       (-1.541)        (0.358)       (1.089)

 Panel B: Difference in Median Operating Performance of Demutualized versus Mutual Exchange (Demutualized Exchange Median - Mutual
 Exchange Median)
 Demutualized           0.096       0.069        0.279        1.476           0.253       0.070           0.470          0.045
 Mutual                 0.075       0.158        0.254        1.359          -0.018       0.095           0.572          0.069
 z-statistics        (2.293)**    (-0.663)      (1.599)     (2.293)**       (1.524)      (-1.540)       (-0.652)       (-0.845)




                                                                                                                                                                    34
Table 3:          Difference in performance of publicly traded exchanges and demutualized but private exchanges

This table contains the results of H2 that the operating performance of publicly listed exchanges is better than that of demutualized exchanges.
Panel A displays the results of the difference in mean test, and Panel B displays the results of the Wilcoxon signed rank test. The t-statistic and
z-statistic are used to measure statistical significance for the difference in mean and median tests, respectively. The symbols ***, ** and *
indicate statistical significance at the 1%, 5%, and 10% level, respectively.


                                                                                              Listing Revenue    Trading
                      Return on     Return on     Net Income       Operating      Growth in                                 Info Revenue to
                                                                                                   to Total    Revenue to
                       Assets        Equity        Margin          Efficiency      Capex                                     Total Revenue
                                                                                                  Revenue     Total Revenue


 Panel A: Difference in Mean Operating Performance of Listed versus Demutualized Exchanges (Listed -Demutualized)
 Listed                 0.086       0.186        0.263           2.832        1.247        0.155           0.463                      0.153
 Demutualized           0.109       0.171        0.295           1.709        1.731        0.087           0.557                      0.128
 t-statistics         (-0.010)   (2.468)**     (-0.652)       (3.786)***   (-1.722)*    (2.554)***        (-0.764)                   (0.243)


 Panel B: Difference in Median Operating Performance of Listed versus Demutualized Exchanges (Listed -Demutualized)
 Listed                 0.077      0.165         0.242           2.373       -0.075        0.150           0.469                      0.115
 Demutualized           0.096      0.069         0.279           1.476        0.253        0.070           0.470                      0.045
 z-statistics         (-0.218)   (2.451)**      (1.335)       (2.545)***    (-0.928)      (1.642)         (-0.702)                   (1.603)




                                                                                                                                                      35
Table 4: Results of Robustness Tests

This table contains the results of the robustness test of H1 and H2. Panel A shows the results of the difference in performance of the demutualized
sample during their mutual to demutualization phases. Panel B displays the results of the operating performance of the publicly traded stock exchange
sample over the three governance stages from mutual to demutualization to self-listing. The t-statistics and z-statistics measure statistical significance
for the difference in mean and median tests, respectively. The symbols ***, **, and * indicate statistical significance at the 1%, 5%, and 10% level,
respectively.

                     Panel A: Demutualized Sample                     Panel B: Self-listed exchanges
                       Mutual    Demutualized                          Mutual [1]       Demutualized [2]     Listed [3]          [2-1]           [3-2]
                       Mean          Mean         (t-statistic)          Mean                 Mean             Mean         (t-statistic)   (t-statistic)
                      Median        Median        (z-statistic)         Median               Median           Median        (z-statistic)   (z-statistic)
 Return on Assets      0.055         0.109         (2.54)***             0.044                0.076            0.086          (2.00)**          (0.37)
                       0.058         0.096          (2.04)**             0.034                0.033            0.077          (-1.68)*          (0.25)
 Return on Equity      0.090         0.171           (1.89)*             0.098                0.168            0.186         (2.53)***          (0.77)
                       0.074         0.069         (-2.04)**             0.099                0.154            0.165         (2.60)***          (0.45)
 Net Income            0.183         0.295         (3.27)***             0.140                0.193            0.263          (2.21)**          (1.15)
 Margin                0.145         0.279         (2.75)***             0.107                0.173            0.242          (2.09)**          (1.36)
 Operating             1.684         1.709           (1.94)*             1.741                2.811            2.832           (1.87)*          (0.14)
 Efficiency            1.247         1.476          (2.13)**             1.284                2.363            2.373         (2.70)***          (0.59)
 Growth in Capex       0.536         1.731            (0.53)             2.755                2.811            1.247          (2.48)**         (-1.17)
                       0.156         0.253            (0.70)             0.609                -0.116          -0.075           (-0.67)          (0.47)
 Listing Rev. to       0.097         0.087           (-0.37)             0.167                0.188            0.155            (0.17)         (-0.55)
 Total Revenue         0.083         0.070           (-1.60)             0.161                0.097            0.150           (-0.14)          (0.31)
 Trading Rev. to       0.627         0.557           (-1.19)             0.325                0.440            0.463            (1.44)       (2.07)**
 Total Revenue         0.713         0.470         (-2.40)**             0.240                0.455            0.469            (1.48)       (2.05)**
 Info. Revenue to      0.051         0.128           (1.86)*             0.209                0.267            0.153            (0.89)         (-0.90)
 Total Revenue         0.054         0.045          (-1.84)*             0.194                0.141            0.115           (-0.56)         (-0.27)




                                                                                                                                                             36
Table 5: Performance of self-listed exchanges and demutualized exchanges during the demutualization phase

This table contains the results of a comparison of the operating performance of publicly traded exchange during their demutualization phase
and the performance of the currently demutualized exchange sample. Panel A presents the results of the difference in mean test, and Panel B
shows the results of the median test. The t-statistic and z-statistic are used to measure statistical significance for the difference in mean and
median tests, respectively. The symbols ***, ** and * indicate statistical significance at the 1%, 5%, and 10% level, respectively.

                                                                                                          Listing     Trading        Info
                                                                 Net
                                   Return on     Return on                  Operating     Growth in      Revenue      Revenue      Revenue
                                                               Income
                                    Assets        Equity                    Efficiency     Capex         to Total     to Total     to Total
                                                               Margin
                                                                                                         Revenue      Revenue      Revenue
   Panel A: Difference in Mean performance of Listed Exchange (at time of demutualization) and currently demutualized Exchange
   Listed: when demutualized         0.076         0.168        0.193         2.811         2.811         0.188        0.440        0.267
   Currently Demutualized            0.109         0.171        0.295         1.709         1.731         0.087        0.557        0.128
   t-statistic                     (-0.273)      (-1.782)*     (-0.387)    (2.563)***     (5.481)***     (1.280)      (-1.051)     (0.949)

   Panel B: Difference in Median performance of Listed Exchange (at time of demutualization) and currently demutualized Exchange
   Listed: when demutualized         0.033         0.154        0.173         2.363         -0.116        0.097        0.455        0.141
   Currently Demutualized            0.096         0.069         0.279        1.476          0.253         0.070        0.470       0.045
   (z-statistic)                    (0.319)       (1.488)      (-0.620)     (2.121)**      (-1.049)     (-2.032)**    (-0.677)   (2.526)***




                                                                                                                                                    37
               Table 6: Cross sectional regression of operating performance measures
    This table contains the results of a series of cross sectional regressions of operating performance measures on
    governance dummies and control variables. The dependent variables of these regressions include ROA, Net
    Income Margin, Operating Efficiency, Listing Revenue, Trading Revenue, and Information Revenue. DEM
    (dummy) and LIST (dummy) are indicator variables that take on a value of 1 if the exchange is demutualized and
    self-listed respectively, and zero otherwise. Panel A presents regressions on all portfolios of exchanges at each
    stage of governance; Panel B presents regressions on the listed and demutualized portfolios during
    demutualization, Panel C presents regressions on demutualized and listed portfolios at their currents ownership
    structures; and Panel D presents regressions on the listed portfolio at each stage of governance. We included a
    number of control variables that can influence the operating performance measures: size (total assets, number of
    firms listed), trading volume (volume in US dollars and volume by number of trades), country-specific controls
    (legal system, regulatory quality, growth in real GDP, level of economic development). The sources of income
    (listing, trading, and information revenue) were controlled for in the regressions with ROA, ROE, Net Income
    Margin, and Operating Efficiency as dependent variables. T-statistics are in parentheses, and the symbols ***, **
    and * indicate significance at the 1%, 5% and 10% level, respectively. Regressions are corrected for
    autocorrelation and heteroskedasticity.

                             ROA           ROE           NIM        Operating        Listing     Trading         Data
                                                                    Efficiency      Revenue      Revenue        Revenue
Panel A: Determinants of operating performance of mutual, demutualized, and listed exchanges
(Constant)                    0.481        -0.215         0.267        0.002          1.059       -1.021          0.569
                             (1.40)       (-0.26)        (0.66)        (0.00)      (4.46)***      (-1.05)       (2.00)**
DEM (dummy)                   0.110        0.076          0.103        0.680         -0.012       -0.032         -0.008
                            (2.31)**       (1.53)       (1.79)*      (2.31)**        (-0.84)      (-0.91)        (-0.88)
LIST (dummy)                 -0.009        -0.010         0.089        0.176          0.038        0.038          0.036
                             (-0.22)      (-0.22)        (1.51)      (2.36)**       (1.97)**       (0.97)      (2.86)***
Controls                       Yes          Yes            Yes          Yes            Yes          Yes            Yes
Adj. R2                       0.534        0.695          0.479        0.720          0.909        0.831          0.929
F-statistic                   8.044        15.166         6.806       16.230        102.471       57.006        133.748
p-value (F-statistic)      (0.00)***     (0.00)***     (0.00)***    (0.00)***      (0.00)***    (0.00)***      (0.00)***
N                              102          102            102          102            102          102            102
Panel B: Determinants of operating performance of demutualized and listed exchanges during demutualization
(Constant)                   -0.002        -0.125        -0.173        -0.261         0.120       -0.533         1.103
                             (-0.02)      (-0.92)        (-0.57)      (-0.17)        (1.50)       (-1.53)      (2.87)***
LIST(dummy)                   0.008        0.050         -0.012        0.616          0.005       -0.035         0.088
                             (0.52)      (2.97)***       (-0.31)    (3.27)***         (0.46)      (-0.76)       (1.73)**
Controls                       Yes          Yes            Yes          Yes            Yes          Yes           Yes
Adj. R2                       0.417        0.328          0.342        0.325          0.423        0.255         0.300
F-statistic                   1.213        2.809          2.991        2.770          5.866        2.738         3.425
p-value (F-statistic)         (0.29)     (0.00)***     (0.00)***    (0.00)***      (0.00)***    (0.01)***      (0.00)***
N                               79           79             79           79             79           79            79
Panel C: Determinants of operating performance of demutualized and listed exchanges
(Constant)                    1.758        -0.661        -0.666        -6.556        -0.018       -0.238          1.706
                             (1.60)       (-0.47)      (-2.41)**      (-1.28)        (-0.38)      (-1.27)      (3.81)***
LIST (dummy)                  0.418        1.606          0.156        0.562          0.540        0.208         -0.013
                             (0.86)        (0.71)      (2.71)***       (0.61)      (3.76)***       (0.45)        (-0.28)
Controls                       Yes          Yes            Yes          Yes            Yes          Yes            Yes
Adj. R2                       0.542        0.665          0.599        0.754          0.830        0.751          0.894
F-statistic                   5.997        9.546          7.662       13.933         33.264       23.575         58.301
p-value (F-statistic)      (0.00)***     (0.00)***     (0.00)***    (0.00)***      (0.00)***    (0.00)***      (0.00)***
N                               69           69             69           69             69           69             69
Panel D: Determinants of operating performance of listed exchanges at each stage of the governance continuum
(Constant)                    0.280        -0.074        -0.140        -1.606         1.210       -0.001         0.882
                             (0.90)       (-0.11)        (-0.44)      (-0.35)       (1.90)**       (0.00)      (2.69)***
DEM (dummy)                   0.092        0.070          0.042        0.709         -0.019       -0.002         -0.010
                              (1.67)*        (1.10)        (0.58)          (0.85)      (-1.12)          (-0.06)     (-0.93)
LIST (dummy)                   -0.031        -0.027        0.083           -0.038       0.036            0.033       0.032
                              (-0.83)       (-0.51)        (1.39)          (-0.06)    (1.90)*            (0.99)    (2.09)**
Controls                         Yes          Yes           Yes              Yes         Yes              Yes         Yes
Adj. R2                         0.544        0.669         0.553            0.694       0.817            0.791       0.942
F-statistic                     6.954       11.112         7.279           11.663      31.795           30.190     121.703
p-value (F-statistic)        (0.00)***     (0.00)***     (0.00)***       (0.00)***   (0.00)***        (0.00)***   (0.00)***
N                                75            75            75               75          75               75          75

               Table 7: Abnormal returns realized by self-listed exchanges
               This table contains the abnormal returns of the listed exchanges. Monthly returns were used to
               calculate the buy-and-hold abnormal return (BHAR) and the cumulative abnormal return (CAR)
               on an equally-weighted portfolio of listed exchanges. The windows tested are 1 year (0, 12
               month), 2 years (0, 24 months), 3 years (0, 36 months), 4 years (0, 48 months), and 5 years (0,
               60). Statistical significance was established using the t-statistic. The symbols ***, ** and *
               indicate statistical significance at the 1%, 5%, and 10% level, respectively.

                                                           BHAR                           CAR
                                                         (t-statistic)                (t-statistic)
                              1 Year                        0.5004                       0.3816
                                                          (3.22)***                    (3.56)***
                              2 Year                        0.9034                       0.5471
                                                         (3.320***                     (3.47)***
                              3 Year                        1.3629                       0.7508
                                                          (2.95)***                    (3.86)***
                              4 Year                        0.9619                       0.7871
                                                           (2.16)*                     (4.06)***
                              5 Year                        2.8458                       1.2222
                                                          (2.80)***                    (9.56)***




                                                                                                                    39
Table 8: Cross sectional analysis of the stock market returns of self-listed stock exchanges

       This table shows the results of the regression of the three-year cumulative abnormal returns of self-listed
       stock exchanges on governance and compensation variables after controlling for interest rates, trading
       volume and the legal system in place. T-statistics are in parentheses and the symbols ***, ** and *
       indicate statistical significance at the 1%, 5%, and 10% level, respectively.

                                              Regression 1     Regression 2     Regression 3     Regression 4
                                              Year 3 CAR       Year 3 CAR       Year 3 CAR       Year 3 CAR
                                               (t-statistic)    (t-statistic)    (t-statistic)    (t-statistic)
   Intercept                                      0.500            0.497            -8.400            -7.872
                                               (12.69)***       (12.61)***       (-2.83)***       (-3.01)***
   Var Comp/Total Comp                            0.469            0.571             0.664             0.339
                                                (3.98)***        (5.13)***        (3.60)***           (1.61)
   Total Comp/Net Income                                           0.251             0.265             0.126
                                                                 (2.91)***        (3.41)***           (1.35)
   Fixed Comp/Total Comp                                                             0.681             0.406
                                                                                  (4.25)***         (2.22)**
   Board Size                                                                        0.022             0.007
                                                                                  (2.53)***           (0.77)
   Effectiveness - % of meetings attended                                           -0.029             0.106
                                                                                    (-0.19)           (0.60)
   Effectiveness - Tenure on board                                                  -0.001             0.003
                                                                                    (-0.14)           (0.76)
   % foreign investors                                                                                 1.835
                                                                                                    (2.37)**
   % institutional ownership                                                        0.538              0.894
                                                                                   (1.10)            (1.77)*
   interest rate (lending rate)                                                    -1.147             -3.969
                                                                                  (-0.11)            (-0.43)
   Log (Volume Shares Traded)                                                       0.875              0.816
                                                                                 (3.45)***         (3.54)***
   DUM (common law=1)                                                               0.204              0.374
                                                                                   (0.60)             (1.23)
   Adj. R2                                        0.031           0.034            0.943            0.958
   N                                               186             186              186              186




                                                                                                                  40
APPENDIX 1: METRIC DEFINITIONS AND PREDICTED SIGN

                                                                             Predicted
     Metric Definitions                                                        Sign
                            a) ROA = Net Income / Total Assets                  +
     Profitability          b) ROE = Net Income / Shareholders’ Equity          +
                            c) Net Income / Total Revenue                       +
     Operating Efficiency   Total Revenue / Expenses                            +
                            a) Growth in Capital Expenditures =
     Investment in                                                              +
                            ( CapEx1 – CapEx0 ) / CapEx0
     Technology
                            b) Investment in Technology / Total Revenue         +
                            a) Listing Fee Income / Total Revenue               –
     Sources of Revenue     b) Trading Income / Total Revenue                   –
                            c) Sale of Information / Total Revenue              +
                            a) Trading Volume / GDP                             +
     Liquidity              b) Turnover Velocity = Trading Volume / Market
                                                                                +
                            Capitalization




                                                                                         41

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:9/27/2012
language:English
pages:41