Capital Market Reactions to the Passage of the Financial Services Modernization Act of 1999

Kenneth A. Carow Kelley School of Business – Indianapolis

Indiana University 801 W. Michigan Street

Indianapolis, IN 46202-5151

Email: kcarow@iupui.edu Phone: (317) 274-2783 Fax: (317) 274-3312

Randall A. Heron Kelley School of Business – Indianapolis

Indiana University 801 W. Michigan Street

Indianapolis, IN 46202-5151

Email: rheron@iupui.edu Phone: (317) 274-4984 Fax: (317) 274-3312

We thank Edward Kane for valuable comments on earlier drafts. We also thank First Indiana Bank for supporting this analysis with a faculty research grant.

1

Capital Market Reactions to the Passage of the Financial Services Modernization Act of 1999

Abstract

The Financial Services Modernization Act of 1999, also known as the Gramm-Leach-Bliley Act (GLBA), removed most of the remaining barriers between financial companies. Stock market reactions to the passage of GLBA vary across financial sectors and company size. Specifically, we find negative returns for foreign banks, thrifts and finance companies; insignificant returns for banks; and positive returns for investment banks and insurance companies. Additionally, larger non-depository firms have higher returns. The return variation reflects resolution of uncertainty surrounding the final provisions of GLBA, competitive pressures, and expectations of future business combinations. Potential gains from business combinations may arise from economies of scope, market power, and/or from an implicit extension of government guarantees to banking affiliates. JEL Classification: G21 Keywords: Deregulation, Financial Services, Insurance, Banking, and Investment Banking

This paper can be downloaded from the Social Science Research Network Electronic Paper Collection:

http://papers.ssrn.com/abstract=285509

1. Introduction

The passage of the Financial Services Modernization Act of 1999, frequently referred to

as the Gramm-Leach-Bliley Act (GLBA), eliminated most of the remaining barriers erected

between commercial banks and nonbank financial companies by the National Banking Act of

1864, by the Glass-Steagall Act of 1933, and the Bank Holding Company Act of 1956.

Although Congress had debated the deregulation of the financial industry ten times in the

previous two decades, prior legislative attempts to remove barriers between banks, investment

companies, and insurance companies always failed to pass (Gaetano (1998)). Despite repeated

Congressional failure to produce sweeping legislation to redefine the financial industry, many of

the barriers preventing banks from entering into the sale of insurance and securities were

removed long before the GLBA's passage in 1999 through decisions made by regulatory bodies

such as the Federal Reserve Board (FRB), the Office of the Comptroller of the Currency (OCC),

and the Office of Thrift Supervision (OTS) (for further discussion, see Broome and Markham

(2000)).

Even without legislative reform, it was obvious that the financial industry would continue

to consolidate and the distinction between traditional banking activities and nonbank financial

companies would continue to gradually erode. However, the enthusiastic capital market

response to the high-profile 1998 merger between Citicorp and Travelers Insurance Group to

form Citigroup intensified the pressure on Congress to pass legislation to supersede the

patchwork of rules and regulations that defined the financial industry. Given that banks could

already engage in a host of nonbank financial services, perhaps the greatest significance of the

Citicorp - Travelers merger was that it directly challenged regulators to address the manner in

which banks and nonbank financial companies could enter into business combinations to form

2

financial conglomerates. Recent studies by Carow (2001a) and by Johnston and Madura (2000)

document that large banks, insurance companies, and brokerage houses experienced significant

positive abnormal returns at the merger announcement. Both studies attribute this spillover of

favorable returns as evidence that the merger would force Congress to deliver the long-overdue

legislation to fully deregulate the financial industry, rather than simply debate the issue again,

and then table it for another year. A little over a year after the 1998 formation of Citigroup,

Congress passed the GLBA, which was signed by President Clinton on November 12, 1999.

This study investigates how the passage of the GLBA affected the stock prices of banks,

thrifts, finance companies, investment banks, and insurance companies (hereafter, collectively

referred to as Financial Companies). Everyone knew that Congress would consider bills to

reform the financial industry in 1999. The capital market reactions studied here document the

extent to which the GLBA’s legislative progress resolved uncertainties about the specifics of

financial reform and simultaneously updated expectations regarding the probability that a

substantive bill would finally pass. Three areas of uncertainty existed. The first issue was under

what rules financial companies would be allowed to form conglomerates. The second issue was

whether and how regulatory jurisdiction over financial conglomerates would be allocated

between different financial regulators. Finally, President Clinton had threatened to veto the bill

unless it upheld his concerns about regulating discrimination in credit via the Community

Reinvestment Act.

When passed, the GLBA expanded the mix of permissible financial activities that could

be carried out by a single company and removed many important restrictions on combinations

between financial companies. In rearranging competitive opportunities in the financial industry,

the GLBA allowed some financial companies to gain at the expense of others. In analyzing its

3

effects by industry sector, we seek to identify cases where downward price pressures from

increased competition offset or overcome upward price pressures from improved opportunities to

form profitable financial conglomerates.

Overall, we find that individual financial sectors reacted diversely to the passage of the

GLBA. The stock prices of investment banks and insurance companies increased significantly,

but the stock prices of banks were not materially affected. At the other end of the spectrum,

foreign banks, thrifts, and finance companies lost value. Further analysis indicates that larger

investment banks and insurance companies reacted more favorably to the GLBA than did their

smaller counterparts. These results are consistent with either of two mutually reinforcing

hypotheses: (1) that large banks would target large nonbank financial companies for acquisition

in search of scope economies, and (2) as Macey (2001) suggests, the GLBA was passed to serve

the interests of large financial conglomerates, including especially investment banks who were

interested in commercial banking. Benefits to large conglomerates might include an implicit

extension of “too-big-to-fail” guarantees to the nonbank activities of future financial

conglomerates and opportunities for financial conglomerates to amass significant market power

(see Kane, 2000).

The remainder of this study is organized as follows. Section 2 summarizes the regulation

that defined the boundaries between bank and nonbank activities prior to the passage of the

GLBA and then discusses some of the GLBA’s highlights. Section 3 summarizes related

literature and develops a foundation for our empirical tests. Section 4 details the key dates

associated with the GLBA’s passage, describes our sample, and discusses our methodology.

Section 5 presents our results and then Section 6 concludes.

4

2. The Regulatory Environment

2.1 Overview of the historical separation of banking and nonbanking activities

Most statutory barriers between commercial banks and nonbank financial companies are

contained in three pieces of legislation. Section 24 of the National Banking Act of 1864

permitted banks to exercise “all such incidental powers as shall be necessary to carry on the

business of banking.” Traditionally insurance was not considered incidental to banking, so

banks were prohibited from engaging in the business of insurance.1 In the wake of bank failures

and the stock market crash of 1929, the Glass-Steagall Act of 1933 sought to restore confidence

in the financial industry. This Act’s Section 20 prohibited a commercial bank from affiliating

with any entity that is "engaged principally” in the sale of securities. Banks responded by using

holding company affiliates to circumvent both sets of restrictions. As a nonbank corporate

entity, a holding company could in principle own a bank, an investment bank, and an insurance

company. To narrow this loophole, the Bank Holding Company Act of 1956 restricted corporate

affiliates to activities that are “closely related to banking.” Although banks subsequently lobbied

to expand their set of permissible activities beyond the “traditional banking activities” of taking

deposits and making loans, actual expansions into new areas were not the result of targeted

legislative change. Instead, as we discuss below, financial institutions entered into business

combinations and/or otherwise pressured regulators to gradually dismantle the barriers between

traditional bank and nonbank activities.

In recent years, the Federal Reserve allowed bank holding companies to underwrite and

deal in securities as long as the holding company was not “engaged principally” in the securities

underwriting business. These loophole affiliates are commonly called section 20 subsidiaries.

1 There are a few exceptions. For example, Section 92 of the National Banking Act of 1916 authorized national banks to sell insurance from banks located in towns with populations less than 5,000.

5

Immediately prior to the passage of the GLBA, the revenue from section 20 subsidiaries could

not exceed 25% of the subsidiary’s gross revenue.2

Similarly, on the insurance front, the Comptroller of the Currency took several steps to

enable banks to enter into insurance markets. As an example, in 1985 the Comptroller declared

annuities, traditionally a leading product of insurance companies, to be a general investment

product rather than an insurance product for regulatory purposes. In a similar vein, in 1986, the

Comptroller expanded the loophole that permitted banks to sell insurance products in towns with

populations under 5000 by interpreting the provision to authorize national banks to sell insurance

anywhere, as long as their insurance transactions were conducted by a subsidiary bank or branch

office in a town whose population did not exceed 5000. Although challenged by the insurance

industry, these rulings were upheld by the Supreme Court. Hence, years before the passage of

the GLBA, banking organizations were acting as agents in the sale of insurance policies

underwritten by unaffiliated insurance companies and competing directly with insurance

companies in writing annuities.

The easiest manner for companies to engage in both depository and nondepository

activities prior to the GLBA was to use the platform of a nonbank unitary thrift holding

company. With relatively few restrictions, nonbank holding company parents that owned a

single thrift were allowed to carry out commercial and financial activities so long as they were

not "detrimental to the safety and soundness of the thrift" (OTS (1997)). Because of the

extensive powers provided by the thrift charter and relatively few restrictions to chartering or

2 The revenue threshold started off as low as 5% when the Federal Reserve approved the requests of Bankers Trust, Citicorp, and J.P. Morgan to underwrite commercial paper, mortgage-backed securities, and municipal bonds in 1987. In 1989, the Federal Reserve increased the securities underwriting threshold to 10% and included corporate bonds and equities under the permissible underwriting activities of bank holding companies that met certain financial standards. The underwriting threshold was increased to 25% in 1996 (for further discussion, see Narayanan, Rangan, and Sundaram 2001).

6

acquiring a thrift, many financial companies (including especially insurance companies) and

several nonfinancial companies either acquired a thrift or applied for permission to establish a

unitary thrift holding company in order to offer a variety of financial products [see Tannenbaum

(1997) and Kane (1999) for further detail].

Two high-profile business combinations also challenged regulatory restrictions and

notably undermined efforts to segregate traditional bank and nonbank activities: BankAmerica's

acquisition of Charles Schwab and Company in 1981 [see Saunders and Smirlock (1987)] and

the 1998 merger of Citicorp and Travelers Group [see Carow (2001a) and Johnston and Madura

(2000)]. BankAmerica's acquisition of Charles Schwab and Company paved the way for banks

to offer discount brokerage services. The Citicorp - Travelers merger forced regulators to

reconsider the boundaries on the activities of financial conglomerates and the statutory

impermissibility of completing the proposed merger put Congress on a timetable for passing

legislation to streamline the maze of statutes, loopholes, and regulatory decisions that had long

fragmented the financial industry.

2.2 Highlights of the GLBA

The legislative process entailed a debate about how to redistribute jurisdiction among

existing financial regulatory agencies. The toughest issue was whether the GLBA should only

authorize financial holding companies (FHCs) to operate under Federal Reserve supervision, or

whether banking organizations would be permitted to choose between the FHC structure

regulated by the Federal Reserve, or to use a bank operating subsidiary whose activities would be

regulated by the OCC. The final legislation compromised on this issue. It favored the FHC

approach by specifying the Federal Reserve to be the primary regulatory authority for financial

holding companies, but requiring the Treasury and the Federal Reserve to jointly determine

7

permissible activities. This debate also extended to the roles of other regulatory agencies. In the

end, the GLBA establishes the Federal Reserve as the regulator of the consolidated FHC, but

limits the Federal Reserve’s authority over entities that are functionally regulated by other

regulators. If a dispute enters the courts, GLBA gives equal deference to the SEC, bank

regulators, and to insurance regulators. Moreover, the Federal Reserve must rely to the

maximum extent possible on examinations and reports prepared by functional regulators and not

apply capital standards directly to a regulated affiliate that is already in compliance with its

functional regulator.3 It remains to be seen whether these limits on the Federal Reserve’s ability

to intervene in the activities of security firms and insurance companies will result in fewer

regulatory conflicts and a smaller regulatory burden for financial companies that must comply

with multiple operating standards. The GLBA closed the unitary-thrift loophole for non-

financial firms. This provision applies to any financial company that is affiliated with a non-

financial firm. Although this action reconfirmed the principle of separating banking and

commerce, the law compromises the principle in a different way, through granting financial

holding companies merchant banking powers.

3. Related Literature

Academicians have long debated the merits of separating bank and nonbank financial

activities. Two areas receive the majority of attention in the literature. First, it is clear that

banks may benefit from diversification if nonbank financial services reduce their dependency on

loans as a dominant source of income (Saunders and Walters (1994), Hughes, Lang, Mester, and

Moon (1999), Szego (1986), and Rose (1989)). Second, nonbank services may permit banks to

utilize their extensive branch networks, back offices, and client relationships more efficiently

(Felgren (1985), Szego (1986), and Herring and Santomero (1990)). Given that most middle-

3 The Act does not prevent the Federal Reserve from establishing a consolidated capital requirement.

8

and high-income consumers have relationships with a bank, banks may have an information

advantage in assessing and accessing client resources (Herring and Santomero (1990), Ang and

Richardson (1994), Kroszner and Rajan (1994), and Puri (1994, 1996)). This advantage could

make it easier to market and service investment banking and insurance products, to adapt product

design to customer needs, and to price the quality of an issuer in the underwriting process

(Kroszner and Rajan (1994) and Puri, (1996, 1999)).

Empirical studies that examine earlier barrier-reducing events can provide a foundation

for predicting how the stock prices and the risk-taking by bank and nonbank financial companies

might change in response to the GLBA's passage. One such event was BankAmerica's

acquisition of Charles Schwab in 1981. This deal resulted in the entry of banks into the discount

brokerage arena. Saunders and Smirlock (1987) report that while bank prices and risk were

largely unaffected, small securities firms experienced a significant decrease in value, presumably

in response to the threat of increased future competition from banks. However, these authors

found no evidence to suggest an increase in the systematic risk exposure of money-center or

regional banks during the announcement period or in response to the deal’s ultimate approval by

the Federal Reserve.

Carow (2001b) reports similar findings analyzing how the stock prices of insurance

companies responded to OCC and Supreme Court rulings authorizing banks to sell annuities and

other insurance products. For both events, the stock prices of insurance companies declined

significantly when the Court ruled that banks could sell annuity products. The stock prices of

banks were unaffected.

Bhargava and Fraser (1998) and Narayanan, Rangan, and Sundaram (2001) examine how

the stock prices of commercial banks respond to decisions by the Federal Reserve Board

9

allowing them to engage in investment banking activities through section 20 subsidiaries. They

report that commercial banks earned significantly positive abnormal returns when the Federal

Reserve first granted permission for banks to enter into investment banking on a limited basis.

However, subsequent authorizations to underwrite corporate securities and to increase the

percentage of permissible revenues from underwriting activities produced negative abnormal

returns along with increased risk. Consistent with increased competition for investment banking

services, they find some evidence that the stock prices of investment banks decreased around the

announcements. As further documentation of competitive pressures, the evidence provided by

Puri (1996), Gande et al. (1997), and Gande, Puri, and Saunders (1999) suggests that

underwriting spreads and bond yield spreads are lowered when banks compete in underwriting

activities.

Johnston and Madura (2000) and Carow (2001a) examine how the announcement of the

Citicorp-Travelers merger affected the stock prices of financial companies. Johnston and

Madura (2000) report that insurance companies, brokerage firms, and large and medium sized

banks all experienced statistically significant positive stock price reactions to the announcement.

Large banks and large brokerage firms experienced the largest returns. They interpret the

positive stock price movements as predicting that regulators would be forced to allow future

combinations of financial companies. Based on the size effect, they suggest that larger

institutions are more likely to be involved in future business combinations, possibly because of

greater efficiencies to be realized from larger distribution networks. Carow (2001a) does not

find a statistically significant stock price reaction for commercial banks as a whole – though he

finds evidence of a positive size effect for banks. Abnormal returns for banks with assets greater

than $10 billion prove significantly higher than the abnormal returns of smaller banks. Carow

10

segregates the insurance companies in his sample and shows that the stock prices of life

insurance companies increased by 1.02% in response to the merger announcement, whereas the

returns for health and property/casualty insurers were insignificant. Carow attributes the gains to

anticipated synergies resulting from future combinations of large banks and life insurance

companies and/or a potential expansion of the government safety net to include the nonbank

activities of financial conglomerates.4

Except for the Citicorp-Travelers merger, the events analyzed focus on intrusions by

banks into the product markets of nonbank financial companies. These intrusions were

asymmetric in that banks gained additional powers, while the nonbank financial companies

involved suffered increased competition. These asymmetric intrusions generate patterns of stock

price responses in which nonbank financial companies lose value and the stock prices of banks

tend on average either to be unaffected or to increase slightly. This stream of research supports

Kane’s (1984) argument that the progressive fusion of the financial industry in the face of

barrier-reducing technological advance illustrates Baumol’s (1982) theory of contestable

markets.

The Citicorp-Travelers merger was a symptom of the ever-increasing competition in the

financial industry and signaled that a fairer and possibly more efficient way to create a financial

powerhouse was to combine large banks and large nonbank financial companies, rather than to

exploit the loopholes and exceptions that favored asymmetric bank entry into nonbank activities.

The qualitatively similar stock price reactions to prior occurrences of industry segments

“fusing” together provides a foundation for predicting that stock price reactions to the GLBA

would balance offsetting forces. On the one hand and consistent with contestable-markets

theory, all firms in the industry should experience downward price pressure due to increased

4 For further discussion on the expansion of the implicit government safety net, see Kane (2000).

11

future competition. On the other hand, individual firms whose characteristics positioned them to

jettison regulatory circumvention costs associated with loophole exploitation and to use the

relaxation of the remaining barriers to effect business combinations with first-mover advantages

should experience upward price pressure. The literature just reviewed suggests that large

financial companies may be able to develop economies of scope or scale associated with using

their extensive distribution channels and volumes of client information. Large firms may benefit

further if consolidation results in increased market power or an implicit expansion of the

government safety net (due too-big-to-fail considerations) to nonbank financial activities (Kane,

2000). At the same time, sectors of the financial industry where firms are likely to become

future takeover targets as a result of the deregulation should benefit from the relaxed restrictions

for business combinations.

4. Legislative dates, sample selection, and methodology

4.1 Legislative dates leading up to the passage of the GLBA

Our tests of GLBA's effects focus on the stock price movements of financial companies

in response to key legislative events leading up to the bill's passage. Because Congressional

committees had debated and approved similar deregulatory legislation ten times in the past

twenty years, we limit the analysis to legislative events that represent progress beyond

Congressional committee approvals.5 The six legislative events we analyze are listed in Table 1

and include: (1) the Senate approval of the Financial Services Modernization Act of 1999

(S.900) on May 6th, 1999, (2) the approval by the House of Representatives (H.R. 10) on July 1st,

1999, (3) the agreement to a compromise bill by the conference committee on October 22nd,

5Because similar bills had passed Congressional committees often in the past 20 years, passage by a committee may merely be a way of milking industry PACs and cannot be deemed a significant surprise to market participants. Therefore, our analysis begins with votes by the Senate, the House of Representatives, and their joint conference committees. Including standing committee votes prior to the Senate and House passage does not significantly affect

12

1999, (4) the signing of the final conference report on November 2nd, 1999, (5) the passage of the

Gramm-Leach-Bliley Act by the House and the Senate on November 4th, 1999, and (6) the

signing of the Gramm-Leach-Bliley Act by President Clinton on November 12th, 1999.6

4.2 Sample Selection

This study analyzes daily stock returns for a sample of 247 U.S. banks (3-digit SIC code

602 but excluding the 4-digit SIC code 6029), 10 foreign banks (4-digit SIC 6029), 145 thrifts

(3-digit SIC 603), 32 finance companies (3-digit SIC 610), 33 investment banks (3-digit SIC

620), 18 life insurance companies (3-digit SIC 631), 12 health insurance companies (3-digit SIC

632), and 55 property / casualty insurance companies (3-digit SIC 633). Sample firms are those

that met the following data availability criteria. Each firm must:

1. be traded on either the NYSE, ASE, or OTC, 2. have daily returns available on CRSP from 9/98 through 11/99, 3. be traded on at least 70% of the possible trading days, and 4. have balance sheet and income statement data available on Compustat. Table 2 provides descriptive statistics for the sectors of financial firms in the sample. For each

sector, the Table reports the proportion of firms with a unitary thrift holding company, the

proportion of firms with section 20 subsidiaries, the average market value of assets for the firms,

and the proportion of firms with greater than $10 billion in assets.

4.3 Methods

We measure the effect of the designated legislative events by means of a multivariate

regression model. The approach closely resembles the methods employed by Schipper and

Thompson (1985), Cornett and Tehranian (1989), Billingsley and Lamy (1992), Sundaram,

Rangan, and Davidson (1992) and Carow and Heron (1998) in prior studies of the wealth effects

of regulatory processes. The model incorporates the seemingly unrelated regression (SUR)

any of the conclusions drawn in this study.

13

framework where dummy variables assume values of one on the key legislative dates and are

zero otherwise.7 As compared to the traditional event-study techniques such as introduced by

Fama et al. (1969) and described in Brown and Warner (1985), SUR takes into account cross-

sectional correlations among the stock returns of the various firms examined.

Formally, we estimate the following system of n equations over a period beginning

September 1, 1998 and ending November 15th, 1999. We do this separately for domestic banks,

foreign banks, thrifts, finance companies, investment banks, and insurance companies.8

ntj

jjntmtnnnt

tj

jjtmtt

eDIcRbaR

eDIcRbaR

++∆++=

++∆++=

∑

∑

=

=

12

1,1

1

12

1,11111

γ

γ

M

where,

Rit = the observed return on firm i’s stock on day t, Rmt = the observed return on the equal-weighted CRSP market index on day t, ∆It = the change in the rate on day t for a 10-year constant-maturity Treasury (It – It-1), Dj = an indicator variable equal to 1 on the jth legislative event, and 0 otherwise, n = the number of firms, and eit = the residual error term.

Our regression model incorporates changes in interest rates (similar to Booth and Officer (1985))

as well as overall market returns. In this regard, bi measures market risk and ci captures firm i’s

interest rate risk.

6 Each of these events was covered by the Wall Street Journal. 7 We use two-day legislative event windows in the SUR estimation. The two-day windows include the days listed in Table 1 as well as the following trading day. Thus, we use a total of (6 x 2) = 12 dummy variables in the SUR. In this context, our estimate of the overall market reaction to the legislation represents the sum of the coefficients on the 12 dummy variables. 8 We analyze US and foreign banks separately throughout the analysis. We find similar results if we analyze the different types of insurance companies (life, health, and property / casualty) separately. We combine the three types into a single category of "insurance companies" because there are relatively few publicly held life and health insurance companies.

14

The weighted least squares approach used to estimate the parameters assumes that

disturbances are independently and identically distributed within each equation, but allows

variances to differ across equations. In this SUR framework, heteroscedasticity and

contemporaneously correlated disturbances are incorporated into the hypotheses tested.

5. Empirical Results

5.1 Stock price reactions to the passage of the GLBA

Table 3 contains the estimates of cumulative abnormal returns (CARs) from the SUR

estimation. The first panel provides the CARs for all six legislative events listed in Table 1; the

second panel reports abnormal returns for the third event only (the conference committee

compromise). In addition to the value of abnormal returns, in each panel we provide the

percentage of positive abnormal returns, signs tests, and tests of statistical significance. Results

are assembled for the following categories of financial companies; U.S. banks, foreign banks,

thrifts, finance companies, investment banks, insurance companies, and for the aggregate

portfolio of financial companies (i.e., all firms). The reported P-value is the level of significance

at which the reported cumulative abnormal return (CAR) would differ from zero. The z-statistic

captures whether the percent of firms with positive cumulative abnormal returns is statistically

different from the null hypothesis of 50%.

The estimated mean cumulative abnormal return of 0.06% for U.S. banks as a whole is

not statistically significant and the percentage of U.S. banks that responded positively (52.23%)

does not differ statistically from 50%. The stock prices of foreign banks decline by an average

of 6.33% and only two of the 10 foreign banks experienced a positive CAR. However, the small

sample size for foreign banks reduces the power of the statistical tests to the extent that only the

15

signs test shows statistical significance -- and then only at the 10% level. The mean CARs and

the percentage of positive CARs for thrifts are also statistically insignificant.

Turning now to nondepository firms, the 32 finance companies in the sample show a

negative effect. They experienced an average abnormal return of -5.72% and only 25% had

positive CARs. As with foreign banks, the roughly 6% decline is not quite statistically

significant at conventional levels, but the signs test indicates that the proportion of positive

abnormal returns is significantly below 50%.

Unlike finance companies, investment banks and insurance companies gained value

during the legislative process. On average, investment banks had CARs of 4.05% and 66.67% of

them showed positive returns (significantly different from 50% at the 10% level). Insurance

companies gained even more from the GLBA's passage. Their average abnormal return is 5.15%

and 68.24% had positive abnormal returns (both significant at the 1% level).

The second panel focuses on the event of October 22, 1999, because the conference

committee agreement is widely viewed as the least predictable part of the legislative process of

the GLBA.9 Early in the morning on October 22nd, the Conference Committee managed to

reconcile the differences between the bills approved by the Senate and the House. The

Conference Committee negotiated several key compromises. First, the Committee fashioned a

pattern of shared regulatory authority between the Federal Reserve and the OCC that was

supported by both Fed Chairman Alan Greenspan and Treasury Secretary Summers. Second, the

Committee won Treasury Secretary Summers’ support for a compromise on community-lending

9 The nature of these compromises prompted Senator Schumer to state “I am for the first time optimistic that we will pass, after a long time, financial services legislation. … In 19 years, we are the closest we have ever been to actually achieving a bill.” [see Yonan, Alan and Dawn Kopecki. “Lawmakers reach an agreement on financial-services reform bill,” The Wall Street Journal, October 22nd, 1999.]. Prior to the barrage of last-minute compromises on October 22nd, Gene Sperling, economic advisor to the White House told Treasury Secretary Summers that “the odds of a deal were just one in three.” [see Schroeder, Michael. “Glass-Steagall compromise is

16

requirements, which largely eliminated the threat of a presidential veto. The return patterns on

this individual date are, for the most part, similar to what we find in cumulating across all six

event-dates.10 Specifically, the October 22nd event shows significant positive gains, supported by

the signs tests, for both investment banks and insurance companies and weaker evidence of

losses by finance companies. The abnormal returns of the remaining categories of financial

companies are not significant at conventional levels.

Overall, the SUR analysis shows that some sectors of the financial industry reacted in a

positive manner to the passage of the GLBA, whereas some sectors declined in value. The final

line in Table 3 shows the corresponding figures when all categories of financial companies (552

total firms) are aggregated into a single portfolio for the SUR estimation. The overall result is an

average abnormal return of 0.17% when cumulated across all event dates and 0.94% on October

22nd. We also test whether the cumulative abnormal returns over all announcements for all

portfolios jointly equal zero or jointly equal each other. The values of the F-tests are 3.28 and

3.92, with p-values of .0033 and .0015, respectively. The joint hypothesis tests confirm that the

returns are not jointly equal to zero and that significant cross-sectional variation exists in the

sample. In the following section, we use a multivariate cross-sectional analysis to investigate

additional factors that might have influenced subsample price reactions.

5.2 Cross-sectional tests

Table 4 presents a multivariate analysis of abnormal return estimates that controls for

additional influences on the stock price reactions of financial companies. The control variables

include size, whether the company has a unitary thrift, and whether the company has a section 20

reached – lawmakers poised to pass banking-law overhaul after last-minute deals.” The Wall Street Journal, October 25th, 1999.] 10 An analysis of the abnormal return correlations of industry portfolios across each of the event dates reveals that the abnormal returns for event periods 2, 4, 5, and 6 are significantly positively correlated with the 3rd event period

17

subsidiary. The endogenous variables in these tests are the CARs across all six of the legislative

events whose subsample averages are reported in the first panel of Table 3. All cross-sectional

models are highly significant, with p-values less than 0.001 and adjusted multiple correlation

coefficients ranging from 15.06% to 16.60%.

We base our cross-sectional return analysis on Table 3’s first panel for at least two

reasons. First, the probability of passage does not reach 100 percent until the bill is signed.

Second, at each stage of the legislative process, additions or deletions may be made that may

result in material changes in each sector’s stake in the bills going forward. Therefore, at each

legislative event, there exists new information about both the features of the bill and the

probability of its ultimate passage. The only way to capture the ebb and flow of information is to

cumulate over all relevant legislative events, ending with the signing of the bill (see Appendix A

for further discussion).11

Each cross-sectional regression model includes a dummy variable that identifies each

category of financial companies (with the exception of small (< $10 billion) domestic banks,

which are represented by the intercept). In this model, sectoral coefficients measure the extent to

which the abnormal return for each sector differs from the abnormal return for small domestic

banks without a unitary thrift or section 20 subsidiary. Categorical variables for company size

represent the extent to which the abnormal returns for firms with greater than $10 billion in

assets differ from the abnormal returns of firms in the same sector with assets less than $10

billion.12

(October 22nd, 1999) returns. 11 Given the amount of news coverage surrounding October 22, we also tested whether the conclusions we draw from our cross-sectional tests also hold on this individual date. The primary conclusions of our study do not change if we use only the third event period. 12 Note that the coefficients and statistical significance of the indicator variables for each category of financial service companies in Table 4 are similar, but are not directly comparable to the univariate estimates in Table 3 because the multivariate analysis includes measures of size and other factors that may influence stock price

18

The CAR regressions indicate that, when controlling for size, three categories of financial

companies (foreign banks, thrifts, and finance companies) experienced significantly lower

abnormal returns than did U.S. banks. Across both models, the coefficients on the foreign

banks’ indicator variable are significantly negative at the 1% level, ranging from -6.627% in

model 2 to -8.292% in model 1. The most likely reason for the significantly lower stock price

reaction for foreign banks than domestic banks has to do with the GLBA's capitalization

requirements for conducting business as a financial holding company in the United States. For a

foreign bank to be eligible for financial holding company status in the U.S., it must be

considered "well-capitalized" by U.S. standards. In determining capitalization status, U.S.

regulators must assess the entire organization, including operations outside of the United States.

Because many foreign countries impose lower capital requirements than the U.S., this eligibility

requirement imposes new costs on foreign institutions that want to do business in the United

States as a financial holding company.

Coefficients for the thrift sector are nearly the same in both models: –1.453% (model 1)

and –1.765% (model 2), although the statistical significance varies. The significantly lower

stock price reaction of thrifts relative to banks is to be expected given the GLBA's reduction of

the thrift charter's advantages. Prior to the GLBA, the thrift charter not only provided a method

of combining financial and nonfinancial enterprises through a unitary thrift holding company,

but also provided a loophole mechanism for combining depository services, insurance, and

brokerage. The GLBA closed the unitary thrift holding company loophole that permitted

nonbanking companies (including insurance companies) to establish a single thrift subsidiary,

and established the financial holding company structure as the primary structure for combining

banks, insurance, and brokerage firms (for further discussion, see Broome and Markham (2000)).

reactions, such as having a unitary thrift charter or a section 20 subsidiary.

19

The coefficients for the finance companies are also close: –6.055% and –7.054%. These

values are statistically lower than the corresponding returns of U.S. banks at the 1% level. As

did the univariate SUR results, this suggests that this sector was harmed by the GLBA's passage.

Again, this is a predicted reaction. The GLBA called for a greater separation between financial

and nonfinancial companies and most large finance companies are affiliated with nonfinancial

firms. GLBA simultaneously limits the ability of finance companies to expand into other

financial activities and expands opportunities for other sectors to compete with them.

Two sectors experienced significantly more positive abnormal returns than U.S. banks.

Both models indicate that the returns to investment banks exceed those for banks by nearly 3% to

4%. Similar results occur for insurance companies. The gains experienced by investment banks

and insurance companies are likely attributable to two, possibly reinforcing factors. The first is

that the GLBA defeated efforts to charter a new umbrella financial regulatory body. The GLBA

specifically limits the extent of the Federal Reserve’s authority over entities that are functionally

regulated by other regulators, and it strengthens the hand of the SEC and insurance regulators by

assigning them equal deference in disagreements that end up in court. Without these limitations

and assignments, nonbanks would experience an additional layer of regulation, and the ability of

the SEC, insurance regulators, and bank regulators to promote sectoral interests would be

curtailed. The second factor favoring these sectors is the market's presumption that many of

their members are apt to be to be targeted by banks for acquisition in the post-GLBA period.13

Berger, Demsetz, and Strahan (1999) extensively review the bank merger and acquisition

literature and suggest that we may be entering a new wave of mergers and acquisitions that will

encompass large banking organizations and financial service providers worldwide. Many studies

13 Most news commentary suggests that investment banks and insurance companies are less likely to acquire banks. For a discussion of the advantages and disadvantages of an insurance company acquiring a bank, see Broome and

20

show that the majority of the gains from acquisitions accrue to the shareholders of target

companies (e.g., Jensen and Ruback (1983) and Jarrell, Brickley, and Netter (1988)). Such a

result is reinforced by prior studies of the effects of deregulation in the banking industry. For

example, Carow and Heron (1998) find evidence that likely takeover targets experienced

significantly higher returns at the passage of the Interstate Banking and Branching Efficiency

Act of 1994.

Coefficients benchmarking the unitary-thrift effects are not significant. The GLBA

grandfathers these companies, so that they can continue to use this organizational form to offer a

variety of financial products. We may infer that these companies fare better than other thrifts

because the coefficient for all thrifts (of which 40 percent are chartered as unitary thrift holding

companies) is significantly negative.

We also find no significant influence from the existence of a section 20 subsidiary. It

appears that firms that were using section 20 subsidiaries to underwrite securities prior to the

GLBA will not retain important first-mover advantages.

Both the financial press and past academic literature suggest that large financial

institutions may gain more from deregulation than smaller firms (see for example, Houston and

Ryngaert (1994), Kane (2000), Delong (2001), and Houston, James, and Ryngaert (2001)). To

test this hypothesis, we include binary measures that identify as large firms in each sector those

whose assets exceed $10 billion. The size dummy in model 1 ignores the sectoral nature of large

firms. The significantly positive coefficient of 3.779% confirms the conjecture that the largest

gains from the deregulation are likely to accrue to large companies.

Model 2 further refines the analysis by specifying separate size dummies for all but one

sectoral category (foreign banks). This allows us to investigate whether and how the size effect

Markham (2000).

21

varies across financial subsectors. Although the coefficients on all sectoral size dummies are

positive, their magnitude is not statistically significant for banks and thrifts. Given that large

banks had repeatedly experienced larger stock price responses to pre-1999 signals of the

impending deregulation of their financial powers, and in light of the concessions banks had to

make to other institutions to elicit their support for the bill, the lack of a significant size effect in

the stock returns of depository institutions is not unreasonable.14 Stock markets had already

capitalized the major gains associated with expanded banking powers and what is being

evaluated here is the market’s assessment of the horse-trading the industry engaged in to seal the

deal. Regulatory decisions and court rulings had already enabled large banks to engage in most

of the services being authorized by the GLBA.15 As Kane (1996) points out after the passage of

the Interstate Banking and Branching Efficiency Act in 1994, “History suggests that for the US

financial-services industry, before an exclusionary statute comes to be formally rescinded, most

of the effects targeted by the rescission will have already been tolerated by the enforcement

system for years.” Although many of the benefits of deregulation were already included in the

stock prices of the affected firms, the legislation would eliminate some of the unnecessarily high

costs of doing things in a circumventive fashion. Our results suggest that the value of these cost

savings to domestic banks and thrifts approximately equaled the opportunity cost of increased

competition from other industry sectors. It is important to note that despite the GLBA’s benefits,

many restrictions still exist. For example, the GLBA restricts banks in choosing between a

14 The GLBA also addresses some of the concerns of small depositories through the creation of Community Financial Institutions. Depositories with less than $500 million in assets can obtain advances for small business and agricultural loans by pledging these assets as collateral for Federal Home Loan Bank Advances. While an indicator variable for depository institutions with assets of less then $500 million has a positive coefficient, it is not significantly different from zero. Excluding this variable does not affect any of our conclusions. 15 See our earlier discussion in section 2 regarding how banks could sell insurance products and how large banks could underwrite securities long before the GLBA’s passage. Gaetano (1998) argues that of the basic powers approved in GLBA, prior regulations had been relaxed to the point where banks were only restricted from controlling open-end mutual funds and underwriting insurance. Even the restrictions on underwriting insurance

22

subsidiary or affiliate structure as well as its ability to share customer data with nonaffiliated

firms.

The statistically significant positive coefficient on size for finance companies does not

suggest that large finance companies benefited from the Act at the expense of smaller finance

companies. Rather, the net effect for large finance companies is determined by adding three

coefficients: the intercept, the coefficient on the finance companies indicator, and the coefficient

on the finance companies size dummy. This sum equals -0.667% (0.034% - 7.054% + 6.353%).

Along with the industry-wide estimate reported in Table 3, we infer only that large finance

companies were not harmed by the threat of increased competition to the same extent as small

finance companies were.

In contrast, the significant positive coefficients on the size dummies for investment banks

(8.308%) and insurance companies (7.871%) have the same sign as the sectoral indicators

(3.349% and 3.012%, respectively). This implies that the GLBA increased the values of large

investment banks and large insurance companies by more than what was experienced by their

smaller competitors. This large increase shows that the dismantling of restrictions that prevented

large nonbanks from acquiring banks and that complicated bank acquisitions of nonbanks has

created expectations that profitable combinations of large investment banks, insurance

companies, and banks may lie in store. In this regard, Cybo-Ottone and Murgia (1998) show that

mergers between banks and insurers in Europe led to significant increases in shareholder wealth.

The size effect is also broadly consistent with Houston and Ryngaert (1994), Kane (2000),

Delong (2001), and Houston, James, and Ryngaert (2001), who find that in the banking industry,

mergers between larger organizations and mergers that increase concentration and market power

produce larger shareholder gains. Hoenig (1999) and Kane (2000) also argue that the gains to

were being undermined. Five states allowed state chartered banks to underwrite insurance (Carow (2001b)).

23

large investment banks and insurance companies may also reflect an implicit extension of “too-

big-to-fail” guarantees. If FHC firewalls are not strong enough to prevent the “too-big-to-fail”

guarantee from being shared with affiliates, an investment bank or insurance company that

acquires or is acquired by a large bank can gain from these implicit guarantees through lower

funding costs or an improved credit standing.16

6. Summary and Conclusions

After spending two decades on the Congressional agenda, financial services

modernization was enacted on November 12, 1999. Financial institutions may now establish

financial holding companies to consolidate bank and nonbank financial companies. An analysis

of the stock price reactions of 552 financial companies to the legislative progress of what became

the Gramm-Leach-Bliley Act (GLBA) reveals a pattern of both gains and losses across the

different segments of the financial industry.

Although financial companies and even commercial companies had a host of exceptions

and loopholes that allowed them to cross-sell financial services prior to the GLBA, the

patchwork of laws in place importantly restricted the creation of financial conglomerates. As the

GLBA moved through the legislative process, it became increasingly likely the costs of effecting

cross-sectoral combinations of financial firms would be greatly reduced. Capital markets may

have bid up the stock prices of likely acquisition targets by predicting potential deals. Our

results suggest that the largest returns to the GLBA's passage were realized by large investment

banks and large insurance companies. The stock prices of banks, both small and large, were

unaffected by the legislation. We interpret this result to mean that the concessions banks made

to pass the legislation produced a deal that left them little average net incremental benefits. The

16 Kane and Yu (1994) and Kroszner and Strahan (1996) show how subsidies implicit in the government safety net may distort a financial institution’s actions.

24

major benefits from product-line diversification at banks appear to have already been impounded

into bank stock prices. Finally, thrifts, finance companies, and foreign banks lost value due to

the passage of the GLBA.

25

References Ang, J. and T. Richardson, 1994. The underwriting experience of commercial bank affiliates prior to the Glass-Steagall Act: A reexamination of evidence for passage of the act, Journal of Banking and Finance 18, 351-396. Baumol, W., 1982. Contestable markets: An uprising in the theory of industry structure, American Economic Review 72, 1-15. Berger, A., R. Demsetz, and P. Strahan, 1999. The consolidation of the financial services industry: Causes, consequences, and implications for the future, Journal of Banking and Finance 23, 135-194. Bhargava, R. and D. Fraser, 1998. On the wealth and risk effects of commercial bank expansion into securities and underwriting: An analysis of Section 20 subsidiaries, Journal of Banking and Finance 22, 447-465. Billingsley, R. and R. Lamy, 1992. Regional reciprocal interstate banking: The Supreme Court and the resolution of uncertainty, Journal of Banking and Finance 16, 665-686. Booth, J. and D. Officer, 1985. Expectations, interest rates, and commercial bank stocks, Journal of Financial Research 8, 51-58. Broome, L. and J. Markham, 2000. Banking and insurance: Before and after the Gramm-Leach-Bliley Act, The Journal of Corporation Law 25, 723-785. Brown, S. and J. Warner, 1985. Using daily stock returns: The case of event studies, Journal of Financial Economics 14, 3-31. Carow, K., 2001a. The Citicorp – Travelers merger: Challenging barriers between banking and insurance, Journal of Banking and Finance, forthcoming Carow, K., 2001b. The wealth effects of allowing bank entry into the insurance industry, Journal of Risk and Insurance, forthcoming. Carow, K. and R. Heron, 1998. The Interstate Banking and Branching Efficiency Act of 1994: A wealth event for acquisition targets, Journal of Banking and Finance 22, 175-196. Cornett, M. and H. Tehranian, 1989. Stock market reactions to the Depository Institutions Deregulation and Monetary Control Act of 1980, Journal of Banking and Finance 13, 81-100. Cybo-Ottone, A., and M. Murgia, 2000. Mergers and shareholder wealth in European banking, Journal of Banking and Finance 24, 831-859. Delong, G., 2001. Stockholder gains from focusing versus diversifying bank mergers, Journal of Financial Economics 59, 221-252.

26

Fama, E., F. Fisher, M. Jensen and R. Roll, 1969. The adjustment of stock prices to new information, International Economic Review 10, 1-21. Felgren, S., 1985. Banks as insurance agents: Legal constraints and competitive advances, New England Economic Review, 34-49. Gaetano, S., 1998. An overview of financial service reform 1998, Connecticut Insurance Law Journal. Gande, A., M. Puri, and A. Saunders, 1999. Bank entry, competition, and the market for corporate securities underwriting, Journal of Financial Economics 54, 165-195. Gande, A., M. Puri, A. Saunders, I. Walter, 1997. Bank underwriting of debt securities: Modern evidence, Review of Financial Studies 10, 1175-1202. Herring, R. and A. Santomero, 1990. The corporate structure of financial conglomerates, Journal of Financial Services Research 4, 471-498. Hoenig, T., 1999. Financial industry megamergers and policy challenges, Economic Review (Federal Reserve Bank of Kansas City) 84, 7-13. Houston, J, C. James, and M. Ryngaert, 2001. Where do merger gains come from? Bank mergers from the perspective of insiders and outsiders, Journal of Financial Economics, forthcoming. Houston, J, and M. Ryngaert, 1994. The overall gains from large bank mergers, Journal of Banking and Finance 18, 1155-1176. Hughes, J., W. Lang, L. Mester, and C. Moon, 1999. The dollars and sense of bank consolidation, Journal of Banking and Finance 23, 291-324. Jarrell, G., J. Brickley, and J. Netter, 1988. The market for corporate control: The empirical evidence since 1980, Journal of Economic Perspectives 2, 49-68. Jensen, M. and R. Ruback, 1983. The market for corporate control: The scientific evidence, Journal of Financial Economics 11, 5-52. Johnston, J. and J. Madura, 2000. Valuing the potential transformation of banks into financial service conglomerates: Evidence from the Citigroup merger, The Financial Review 35, 17-36. Kane, E., 2000. Incentives for banking megamergers: What motives might regulators infer from event-study evidence? Journal of Money, Credit, and Banking 32, 671-701. Kane, E., 1999. Implications of superhero metaphors for the issue of banking powers, Journal of Banking and Finance 23, 663-673.

27

Kane, E., 1996. De jure interstate banking: Why only now? Journal of Money, Credit, and Banking 28, 141-161. Kane, E., 1984. Technological and regulatory forces in the developing fusion of financial-services competition, Journal of Finance 39, 759-773. Kane, E. and M. Yu, 1994. How much did capital forbearance add to the tab for the FSLIC mess? Conference on Bank Structure and Competition (Federal Reserve Bank of Chicago). Kroszner, R. and R. Rajan, 1994. Is Glass-Steagall Act justified? A study of the U.S. experience with universal banking before 1933, American Economic Review 84, 810-832. Kroszner, R. and P. Strahan, 1996. Regulatory incentives and the thrift crisis: Dividends, mutual-to-stock conversions, and financial distress, Journal of Finance 51, 1285-1320. Macey, J., 2001. The Business of Banking: Before and After Gramm-Leach-Bliley, Journal of Corporation Law, forthcoming. Narayanan, R., N. Rangan, and S. Sundaram, 2001. Relaxing Glass-Steagall: Welfare implications of bank entry into securities underwriting, Quarterly Review of Economics and Finance, current issue. OTS, 1997. Holding Companies in the Thrift Industry. Background Paper, April, Washington, DC. Puri, M., 1994. The long-term performance of bank underwritten security issues, Journal of Banking and Finance 18, 397-419. Puri, M., 1996. Commercial banks in investment banking: Conflict of interest or certification role? Journal of Financial Economics 40, 373-402. Puri, M., 1999. Commercial banks as underwriters: Implications for the going public process, Journal of Financial Economics 54, 133-163. Rose, P., 1989. Diversification of the banking firm, The Financial Review 24, 251-280. Saunders, A. and S. Smirlock, 1987. Intra- and interindustry effect of bank securities market activities: The case of discount brokerage, Journal of Financial and Quantitative Analysis 22, 467-482. Saunders, A. and I. Walters, eds., 1994. Universal Banking in the United States: What Could We Gain? What Could We Lose? New York, NY: Oxford University Press. Schipper, K., and R. Thompson, 1985. The impact of merger-related regulations using exact distributions of test statistics, Journal of Accounting Research 23, 408-415.

28

Schroeder, M., 1999. House commerce panel approves bill to revamp financial-services industry, The Wall Street Journal, June 11, 1999. Schroeder, M., 1999. Glass-Steagall compromise is reached – lawmakers poised to pass banking-law overhaul after last-minute deals, The Wall Street Journal, October 25, 1999. Sundaram, S., N. Rangan, and W. Davidson III, 1992. The market valuation effects of the Financial Institutions Reform, Recovery, and Enforcement Act of 1989, Journal of Banking and Finance 16, 1097-1122. Szego, G., 1986. Bank asset management and financial insurance, Journal of Banking and Finance 10, 295-307. Tannenbaum, I., 1997. Despite legislative cloud, federal thrift charter popularity remains strong, Banking Policy Report 16, December 15. Secura Group, Washington, pp. 1, 14-17. Yonan, A. and D. Kopecki, 1999. Lawmakers reach an agreement on financial-services reform bill, The Wall Street Journal, October 22, 1999.

29

Table 1: Event Dates Surrounding the Passage of GLBA

Event Date Action May 6, 1999 Senate approves Financial Services Modernization Act of 1999

(S.900) 55-44. President Clinton threatens a veto.

July 1, 1999 House of Representatives approves H.R.10 343-86.

October 22, 1999 Conference Committee agrees on a compromise version. The early morning negotiations significantly reduced the threat of a presidential veto.

November 2, 1999 Conference report signed by majority of conferees, clearing the way for votes in the House and Senate

November 4, 1999 Gramm-Leach-Bliley Act passes the Senate 90-8 and the House 362-57

November 12, 1999 President Clinton signs Gramm-Leach-Bliley Act

30

Table 2 Descriptive Statistics – Sample Composition The Table provides a sectoral breakdown of the 552 firm sample of financial companies.

US Banks

Foreign Banks

(SIC6029)

Thrifts (SIC603)

Finance Companies

(SIC61)

Investment Banks

(SIC62)

Life

Insurance Companies (SIC631)

Health

Insurance Companies (SIC632)

Property-Casualty Insurance Companies (SIC633)

Sample Size 247 10 145 32 33 18 12 55 Proportion of firms whose structure includes a unitary thrift holding company 6.88% 40.00% 6.25% 12.12% 11.11% 8.33% Proportion of firms whose structure includes a section 20 subsidiary 7.69% 30.00% Average market value of assets (millions) $14,252 $170,992 $4,237 $35,529 $20,025 $33,964 $29,054 $16,381 Proportion of firms with assets > $10 billion 17.41% 80.00% 6.90% 21.88% 9.09% 44.44% 41.67% 18.18%

31

Table 3 Abnormal Returns Using the SUR Model Cumulative Returns for all 6 Events Cumulative Returns for Event 3 Category of Financial Company

N

Cumulative Abnormal Return

P-value

% Positive

Z-statistica for signs test

Cumulative Abnormal Return

P-value

% Positive

Z-statistica for signs test

US Banks 247 0.06% 0.966 52.23% 0.70 0.72% 0.218 59.92% 3.12 *** Foreign Banks 10 -6.33% 0.150 20.00% -1.90 * -0.54% 0.756 30.00% -1.26 Thrifts 145 -1.68% 0.258 44.14% -1.41 0.59% 0.314 58.62% 2.08 ** Finance Companies 32 -5.72% 0.109 25.00% -2.83 *** -1.80% 0.205 34.38% -1.77 * Investment Banks 33 4.05% 0.229 66.67% 1.91 * 3.09% ** 0.025 78.79% 3.31 *** All Insurance Companies

85 5.15% *** 0.003 68.24% 3.36 *** 2.55% *** 0.001 71.77% 4.01 ***

All Financial Companies

552 0.17% 0.889 51.27% 0.60 0.94% * 0.064 60.51% 4.94 ***

Estimated equation: it

jjjitimtiiit eDIcRbaR ++∆++= ∑

=

12

1,γ . Each legislative window spans two days: the days listed in Table 1 and the

following trading day. Thus, we incorporate a total of 12 (6 x 2) dummy variables in the SUR. Within this framework, for each firm, the estimation of the market's overall reaction to the legislation represents the sum of the coefficients on the 12 dummy variables. The reported cumulative abnormal return figures in this Table are the average across each category of firms. a The z-statistic is determined as, )P1(N/)NG( pp −− , where G is the number of positive parameter estimates, N is the total number of parameter estimates, and P = .50 (the probability of a positive estimate). *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels.

32

Table 4 Cross Sectional Regressions Seeking to Explain Abnormal Returns The endogenous variable is the abnormal return from the SUR analysis. The remaining variables are dummy variables that identify designated categories of financial companies. *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels.

Model 1 Model 2 Coefficient p-value Coefficient p-value

Intercept (Domestic Banks SIC=602 excludes 6029) -0.416 (.396) 0.034 (.947)

Foreign banks (SIC=6029) -8.292*** (.001) -6.627*** (.009)

Thrifts (SIC=603) -1.453* (.082) -1.765** (.040)

Finance companies (SIC=61) -6.055*** (.001) -7.054*** (.001)

Investment banks (SIC=62) 4.175*** (.002) 3.340** (.015)

Insurance companies (SIC=63) 4.547*** (.001) 3.063*** (.003)

Unitary thrift holding company 0.177 (.852) -0.346 (.714)

Section 20 Subsidiaries -2.172 (.211) -0.108 (.953)

Size Dummiesa

All Large Firms 3.779*** (.001)

Large Banks (SIC=602) 0.736 (.598)

Large Thrifts (SIC=603) 3.732 (.112)

Large Finance Cos. (SIC=61) 6.353** (.040)

Large Investment banks (SIC=62) 8.310* (.056)

Large Insurance Cos. (SIC=63) 7.871*** (.001)

R-squared 16.30% 18.42% Adjusted R-squared 15.06% 16.60% F-statistic 13.215 10.14 P-value <.001 <.001

a In each industry sector, a firm is defined as large if its assets > $10 billion.

33

Appendix A: A separate analysis of the event dates does not fully account for the dynamic nature of the legislative process. In order to capture the dynamics of a legislative process, one should cumulate over all relevant legislative events, ending with the signing of the bill – at which point the features of the bill are fully known and the probability of its passage has by definition, reached 100 percent. Consider the following example. Example: Suppose that at the time the ABC bill (a fictitious bill that was for the purpose of this example, introduced in 1999) started in the legislative process, the proposed bill had 3 features (1) it would remove all remaining restrictions between financial companies, (2) it would remove any privacy concerns, and (3) it would allow banks to further expand their activities into more risky, commercial activities. Also suppose that we know somehow that if the ABC bill passed its proposed form, that it would lead to a 3% increase in the value of banks with 1.5% of the wealth effect attributable to the removal of restrictions between financial companies, 1.00% attributable to the allowing financial companies to share data as they see fit, and 0.5% attributable to possible expansion into commercial activities. Now, consider a simple scenario that includes 4 legislative dates: Scenario: Event 1: Bill is introduced: Regulators are skeptical about allowing banks to expand into commercial activities. The market knows this and realizes that without compromise, the bill has a 40% chance of emerging as is from the legislative process. Stock market reaction to legislative event = .40 * 3% = 1.2% Event 2: Bill passes the House, remains intact. It is known at this time that compromise is still likely, in particular, with regard to the proposed commercial activities. However, the probability of the bill’s passage as is climbs to 70%. Stock market reaction to the legislative event = (.70 - .40) * 3% = .9% Event 3: Bill passes the Senate, but only after compromise led to the removal of the provision that would have allowed banks to expand their levels of commercial activities. The president concurrently indicates that he will likely (with probability of 95%) sign the bill when it crosses his desk. Stock market reaction to the legislative event = (.95 - .70) * (1.5% + 1.00%) - (.70 * 0.5%) = 0.625% - 0.35% = 0.275% Event 4: President comes through and signs the bill (i.e., probability of passage reaches 100%) Stock market reaction to the legislative event = (1.0 - .95) * (1.50% + 1.00%) = .125%

34

Sum of abnormal returns across all legislative events: 1.2% + 0.9% +0.275% + 0.125% = 2.5% Notice that 2.5% is exactly equal to what we knew in advance to be the collective effect if the legislation were to pass with 100% probability removing the separation between financial companies (1.50%) and not include provisions limiting the use of personal bank data (1.00%). Although deliberately simplistic, this example captures our motivation for summing across all legislative dates in order to determine the overall wealth effects of the legislation. It also shows the potential problems with an analysis of individual dates. If one were to analyze our hypothetical event 3 separately, they could easily make the wrong conclusion. Given the prior information, we knew that eliminating the provision that allows banks to expand into commercial activities would result in a reduction in value; however, note that the stock price reaction on the 3rd event is a positive 0.275%. Because both the features and the probability of passage are simultaneously changing in a dynamic legislative process, a summation across legislative events is necessary to correctly ascertain the underlying wealth effects. In our study, it is clear that the most important date in the legislative process occurred when the Conference Committee agreed to a compromise version of the GLBA. This date enhanced the likelihood that the bill would ultimately pass (i.e., at this point the probability of the bill’s passage became very high). We include the remaining dates to acknowledge the possibility of additional compromise and/or failure.