impact of mergers - world banksiteresources.worldbank.org/dec/resources/impact_of_mergers_on... ·...
TRANSCRIPT
1
Impact of Mergers on Customer Retention and Acquisition
In this paper, I use customer-level data from the underwriting industry to test
the belief that mergers result in customer defection. I focus on mergers between
commercial banks and investment banks between 1997and 2001 following the
relaxation of regulatory restrictions on commercial bank underwriting. I find that
acquired investment banks lose more underwriting customers and gain fewer new ones
in the first three years after the mergers, compared to their own performance prior to
the mergers, and compared to the performance of un-acquired investment banks. The
results appear consistent with the organizational economics literature on synergy-
related costs of integration.
1
1This paper is based on chapter 4 of my dissertation at Harvard University. I am grateful to my advisors, Jeremy Stein, Richard Caves, Fritz Foley and Bharat Anand, as well as to Ariel Pakes and George Baker. Seminar participants at Harvard Business School and Harvard University provided useful input. All remaining errors are mine.
2
1. Introduction
Anecdotal information and practitioner evidence both suggest that there is a
high risk of customer defection after mergers. However, this issue has not received
much attention in the empirical literature. In this paper, I empirically examine the
impact of mergers on customer retention and acquisition.
There have been some studies examining the impact of mergers on the ability
to gain and retain customers using market share data (see for instance, Mueller
(1985)). However, using such data makes it difficult to effectively control for other
factors that can also impact customer retention and acquisition. Using customer-level
data would make it easier to control for selection issues, and correctly identify the
impact of the merger. Such data is comprehensively available in the financial industry.
Hence, I use data from the financial industry to evaluate the impact of mergers on
customer retention and acquisition.
My sample consists of a set of mergers between commercial banks and
investment banks in the late 1990s. These mergers were in response to a regulatory
change in December 1996, relaxing restrictions on commercial bank participation in
the underwriting market, as described in Section 2. The commercial banks sought to
leverage the customer base and capabilities of the acquired investment banks to
expand into the underwriting business. I use detailed data on customers of each bank
over the period 1990-2003, to examine the impact of these mergers on customer
retention and acquisition in the underwriting business. I do so by comparing the ability
of merged banks to retain and gain customers with their performance prior to the
3
mergers, and also to the performance of comparable non-merging investment banks
(henceforth referred to as “benchmark banks”). Section 3 describes the data used in
the paper and selection of benchmark banks.
Section 4 focuses on the retention of underwriting customers. I study whether
mergers increase the probability that an issuing company switches away from its prior
underwriting relationship, using a logistic regression. I find that across all mergers,
customers of the acquired bank are more likely to leave. This is relative to their
switching rates away from these banks pre-merger, as well as switching rates away
from benchmark banks. I control for any deliberate attempt by the merged bank to
realign the customer base of the acquired bank, either by dropping small customers (as
suggested by the commercial banking literature on mergers) or by focusing on their
lending clientele. I find that the negative impact of mergers still holds. Moreover,
customers who switch away from any of the merged banks are more likely to switch to
benchmark banks, rather than to any other merged bank.
In section 5, I focus on acquisition of new underwriting customers. I use a
conditional logit model of underwriter choice to compare the relative abilities of
merged banks and benchmark banks to win new underwriting deals over the period
1998-2003, after the deregulation. I control for the pre-deregulation ability of each
bank to win a particular deal, by including controls for the past expertise of the bank in
the product and industry segment corresponding to the deal. I find that most merged
banks are significantly less likely to win deals from new clients relative to benchmark
banks. The results are not driven by exogenous changes in the mix of underwriting
deals and differences in bank capabilities. Moreover, the loss of customers occurs
4
immediately after the mergers, and, at a time when the underwriting market is
undergoing a boom period. This makes it unlikely that the failure to gain customers is
a result of the merged banks reducing focus on the underwriting business. Finally, in
this section, I examine changes in market shares for different types of banks over time.
This analysis confirms that after the mergers, merged banks lose out to benchmark
banks in overall market share, even though the acquired investment banks competed
effectively with benchmark banks just before the mergers.
Andrade, Mitchell and Stafford (2001) show companies in various industries
have undergone mergers in the last decade to take advantage of new investment
opportunities created by deregulation. I focus on one such industry in this paper and
find significant customer defection after the mergers. These findings have important
strategic and competitive implications. In addition, customer defection after mergers
can also directly impact measures of overall merger performance, such as profitability
and value. (see for instance, Homburg and Buceruis (2005)).
The results in Sections 4 and 5 strongly show that mergers harmed the ability
of banks to retain existing and obtain new underwriting customers. It would be useful
to identify the mechanisms through which mergers result in such customer defection.
In section 6, I relate my empirical results on cross-sectional differences among
mergers and merger impact to qualitative information obtained from analysts’ reports.
I consider three phenomena associated with mergers that may result in customer
defection. These include employee turnover, customer uncertainty and dissatisfaction
due to internal focus on merger integration (see for instance, Homburg and Bucerius
(2005)), and synergy-related costs of integration (Dessein et al (2005) and Hart and
5
Holmstrom (2002)). I find suggestive evidence in favor of synergy-related costs of
integration; some of the differences across mergers are also consistent with the other
two explanations. Identifying precisely why and how mergers result in customer
defection is an important topic for future research. Case studies of specific mergers
may be the most effective way to shed further light on these questions. Section 7
concludes the paper.
2. Background
In this section I briefly review the circumstances under which the mergers took
place. Commercial banks in the United States were prohibited from engaging in the
underwriting of stocks, bonds, and other securities under the Glass-Steagall Act of
1933. The Act was motivated by fears that commercial banks’ participation in the
underwriting market would create conflicts of interest, harming investors. Specialized
intermediaries carried out investment banking and commercial banking activities until
the 1990s. Thereafter, the Federal Reserve Bank began to gradually relax restrictions
on the activities of commercial banks. In January 1989, it allowed commercial banks
to underwrite corporate debt, and shortly thereafter corporate equity. Commercial
banks were required to house these underwriting activities under independent units, or
Section 20 subsidiaries, to be approved by the Fed on a case-by-case basis. The
commercial banks were subject to revenue limitations on overall underwriting
revenue. They were also required to maintain several operating limitations or
“firewalls” between the commercial banking units of these banks and the Section 20
subsidiaries.
6
Many commercial banks entered the underwriting market between 1990 and
1996 by setting up Section 20 subsidiaries (see Chaplinsky and Erwin (2005) and Puri
(1996)). In this period, their presence was felt mostly in the debt markets and they did
not have much success in the equity underwriting market (see Table 1). Over this
period, the equity underwriting market was also experiencing a boom both in terms of
number of issues and issue value, benefiting many mid-sized investment banks.
Meanwhile, academic research using data from the 1990-96 period as well as
from the period before 1933, did not find empirical evidence supporting the theory of
conflicts of interest. Instead, research in the nineties emphasized information-based
economies of scope between lending and underwriting. This research suggested that
younger, smaller, and less well-known firms would benefit from the entry of
commercial banks into the underwriting market, by obtaining greater access to the
public markets at lower cost.
In December 1996, the Fed eased many of the regulations on commercial bank
underwriting. It raised the revenue cap on commercial banks in the underwriting
market to 25 percent and eliminated certain firewalls between lending and
underwriting. This allowed for larger levels of underwriting activity, information
sharing and joint operations between the Section 20s and their parent commercial
banks. In response to the deregulation, more than 25 commercial banks entered into
mergers with investment banks between 1997 and 2002. The mergers were explicitly
motivated by the desire of commercial banks to offer their clients the full range of
services, particularly equity underwriting (see Chaplinsky and Erwin (2005)). Many of
the acquired banks were mid-sized investment banks that catered to middle market
7
clients and had a significant presence in the equity underwriting market. On the
commercial banking side, the acquiring commercial banks included large domestic
commercial banks, mid-sized regional ones, and foreign banks. Many investment
banks, including 7 large investment banks, and a large number of smaller investment
banks, remained independent.
3. Data and sample selection
3.1 Data Sources
Data on the mergers between investment banks and commercial banks is
obtained primarily from the Securities Industries Association’s Securities Industries
Fact Book. Table 1 contains a list of the investment banks that underwent mergers,
along with the effective date of the merger.
8
17 Wassertein Perella Dresdner Kleinworth Benson Germany January-01
Table 1 : Mergers between commercial banks and investment banks - 1997 to 2002
Acquired bank Acquiring bank Country of acquiring
bank Effective date
of merger*
1 Salomon Brothers Smith Barney (Travellers Group) December-97Salomon Smith Barney Citigroup USA October-98Wertheim Schroder Citigroup USA May-00 Robinson Humphrey sold USA July-01
2 Hambrecht & Qvist Chase Manhattan Bank USA December-99 JP Morgan Chase Manhattan Bank USA December-00
3 Dillon Read Swiss Banking Corporation (SG Warburg) Switzerland September-97 UBS Securities (Union Bank of Switzerland) Swiss Banking Corporation Switzerland June-98 JC Bradford Paine Webber USA June-00 Paine Webber (including JC Bradford) UBS Switzerland November-00
4 Alex Brown Bankers Trust Corp USA September-97 Alex-Brown - Bankers Trust Deutsche Bank Germany June-99
5 Montgomery Securities Nations Bank USA October-97
Robertson Stephens Bank of America USA October-97NationsbancMontgomery Securities (Nations Bank) Bank of America uSA September-98 Robertson Stephens sold USA September-98
6 Robertson Stephens Bank of Boston USA September-98 Banc Boston Robertson Stephens Fleet Bank USA October-99Robertson Stephens -closed in July, 2002
7 Oppenheimer CIBC Canada November-97 8 Piper Jaffray US Bancorp USA May-98 9 Cowen Societe General France September-98
10 Dain Bosworth Rauscher Pierce Refnes USA January-97 Wessels, Arnold & Henderson Dain Rauscher USA March-98Dain Rauscher Wessels Royal bank of Canada Canada January-01 Tucker Anthony Sutro Royal bank of Canada Canada March-02
11 Wheat First Butcher Singer First Union USA January-98 Kemper Securities Everen USA January-95 Principal Securities Everen USA January-98 Everen Securities First Union USA October-99First Albany Securities First Union USA August-00 Interstate Johnson Lane Wachovia Bank USA April-99 Wachovia Securities (Wachovia Bank) First Union USA September-01
12 Chicago Corp ABN Amro Netherlands January-97 Furman Selz ING Barings UK September-97 ING Barings LLC ABN Amro Netherlands April-01
13 Mc Donald Investments Key Corp USA October-98
14 Morgan Keegan Regions Financial USA March-01 15 Equitable Securities Sun Trust Bank USA September-97
Robinson Humphrey Sun Trust Bank USA July-01 16 Van Kasper First Security Corp-Wells Fargo USA June-99
This table lists all the mergers between investment banks and commercial banks over the period 1997-2002, in response to the relaxation of restrictions on commercial bank underwriting in 1997. Information on mergers is obtained from the Securities Industries Association’s Securities Industries Fact Book for the years 1997-2002. For the purposes of my analysis, mergers are considered effective from the first day of the month following the effective date of the merger.
9
The data on customers of the banks comes from the SDC Platinum’s new issues
database. I use data on public underwritten debt and equity issues from the industrials
segment for the period 1990 to 2003. I exclude issues from financial firms (sic 6000-
6999), government agencies (sic 9000-9999) and utilities (sic 4900-4999). I focus on
the lead underwriter(s) in each deal, as it is the lead underwriter who has the most
interaction with the client and receives most of the fees.
I match SDC data to Compustat to obtain company information on sales,
assets, etc. I also use Dealscan, a database of commercial loans, to identify pre-merger
lending relationships of the commercial banks, since lending relationships have been
found to be a significant factor in determining underwriter choice. There have been a
significant number of mergers among the commercial banks themselves, which must
be accounted for while defining lending relationships. I use Federal Depository
Insurance Corporation’s Institution Directory and Industry Analysis databases, and the
Federal Reserve’s National Information Center database for obtaining data on mergers
within commercial banking.
I also use IBES, a database containing data from analyst reports, to obtain
information on research coverage provided by each bank. I also form a measure of the
quality of an underwriter’s analyst coverage, by using information on annual analyst
rankings from Institutional Investor.
3.2 Selection of benchmark banks
Table 2 shows the distribution of issues and share of the acquired banks prior
to the deregulation, among large and mid-sized customers. It can be seen that, with the
10
exception of Salomon Smith Barney, the acquired banks catered primarily to mid-
sized customers, and had significant market shares in this segment. For this purpose, I
define mid-sized companies as those with assets less than $1,500 million for fiscal
year 1998 or fiscal year 1997.
Table 2 : Market shares of banks in underwriting issues over 1990-1996 (by number of issues)
Number of distinct issuers 560 3,821 Number of issues 3,225 5,833 Average number of issues per issuer 5.76 1.53
Product-wise-distribution of issues
Equity issue - IPO 119 4% 2,442 42% Equity issue - SEO 416 13% 1,890 32% Debt issue - CD 150 5% 244 4% Debt issue - NCD 2,540 79% 1,257 22% Total issues 3,225 5,833
Share of underwriters in number of issues by bank type
Equity issues
Investment banks that get acquired after 1996 Salomon Smith Barney 68 13% 293 7% All others 52 10% 1,283 30%
Large independent investment banks 373 70% 1,077 25% Small independent investment banks 31 6% 1,609 37% Commercial banks 11 2% 70 2%
Total equity issues 535 4,332 Debt issues
Investment banks that get acquired after 1996 Salomon Smith Barney 294 11% 173 12% All others 51 2% 105 7%
Large independent investment banks 1,910 71% 891 59% Small independent investment banks 73 3% 97 6% Commercial banks 360 13% 235 16%
Total debt issues 2,688 1,501
Large companies
Midsized companies
This table compares the market shares of the acquired investment banks by number in underwriting issues of midsized and large companies over the period 1990-1996 period, prior to the deregulation and the mergers. Mid-sized companies are defined as those with assets lower than $1,500 million for fiscal year 1998 or fiscal year 1997. If this information is missing in Compustat, a company is considered mid-sized if it issues securities on less than 10 occasions between 1990- 2003.
Market shares of investment banks that will be acquired subsequently are shown separately for Salomon Smith Barney and other banks. Large independent investment banks or benchmark banks include Goldman Sachs, Merrill Lynch, Morgan Stanley, CSFB, DLJ, Lehman Brothers and Bear Stearns. All other investment banks that remain independent are classified as small investment banks. Commercial banks are Sec20 subsidiaries, permitted to offer underwriting services on a limited basis over this period.
11
Hence, I evaluate the impact of mergers by focusing on underwriting issues
made by mid-sized customers before and after the deregulation.
The mergers between commercial banks and investment banks took place
within a relatively short period, after the deregulation. There was often a delay
between the announcement date and the effective date of each merger, making it
difficult to identify bank-months that reflect the impact of the merger and bank-
months that do not. Hence, I evaluate the impact of mergers on customer retention
and acquisition, by comparing the performance of the merged banks against
comparable banks that remain independent throughout the sample period.
Table 3 shows the market shares of different banks by number of issues over
1994-1996 among mid-sized customers. Overall there are 116 investment banks that
underwrite at least 5 issues at any time between 1990 and 2003. The comparison of
market shares in the table shows that that several of the acquired investment banks
have market shares comparable to the 7 largest independent investment banks,
particularly in equity underwriting. Moreover, the mergers were expected to enable the
acquired banks to build further on their shares in both debt and equity underwriting, by
leveraging the lending relationships and capital base of acquiring commercial banks.
Hence, I use the 7 large investment banks as a benchmark for evaluating impact of the
merger on customer performance. These banks are referred to as “benchmark banks”
throughout the paper.
12
12 Chicago Corp 5Furman Selz 10 1ING Barings LLC 4 1ABN Amro
13 Equitable Securities 6Robinson Humphrey 12Sun Trust Bank
14 Morgan Keegan 15Regions Financial
15 Van Kasper 9 1First Security Corp
16 Mc Donald Investments 6Trident Securities
Key Corp
17 Wassertein Perella 2Dresdner Kleinworth Benson 2
Table 3 - Identification of benchmark independent investment banks
Large independent investment banks Equity Debt Equity Debt Equity Debt
1 Salomon Brothers 66 50 Merrill Lynch 117 84 WM Blair 29 0 Smith Barney 97 28 Goldman Sachs 106 87 Needham 24 0 Wertheim Schroder 16 3 Morgan Stanley 106 64 Cruttenden-Roth 21 0 Citigroup 12 DLJ 77 65 Rodman 15 12
Lehman Brothers 76 44 Jefferies 14 0 2 Alex Brown 116 8 CSFB 48 59 Raymond 13 1
Bankers Trust Corp 4 30 Bear Stearns 42 34 Steichen 12 0 Deutsche Bank 6 2 Dean Witter 11 1
Gerard Klauer 11 0 3 Hambrecht & Quist 95 2 Paulson 11 0
JP Morgan 18 41 Unterberg 11 0 Chase/ Chemical Securities 3 37 Volpe 11 0
Janney Montgomery 9 0 4 Dillon Read 27 13 Kinnard 9 0
Paine Webber including JC Bradford 65 7 Allen 8 0 UBS 15 4 Baird 7 0 SBC Warburg 7 3 Commonwealth 7 0
Josephthal 7 0 5 Montgomery Securities 90 3 Vector 7 3
Nations Bank 8 Lazard 6 0 Bank of America 2 Advest 4 0
Hanifen 4 0 6 Robertson Stephens 73 2 AG Edwards 3 0
Bank of Boston Freidman 3 0 Fleet Bank Miller-Johnson 3 0
Stephens 3 0 7 Dain Bosworth, Wessels & Rauscher 31 Adams-Harkness 2 0
Tucker Anthony Sutro 7 Fahenstock 2 0 Royal bank of Canada 1 Ladenburg 2 2
Roney 2 0 8 Oppenheimer 34 1 First Michigan 1 0
CIBC 2 5 9 Cowen 36
Societe General 10 Wheat First Butcher Singer 11 1
Everen (with Kemper and Principal) 10 1 First Albany Securities 2 Interstate Johnson Lane 4 Wachovia Securities (Wachovia Bank) 1 First Union
11 Piper Jaffray 22 1 US Bancorp
This table compares the number of equity and debt issues over the period 1994-1996 made by midsized companies, underwritten by different investment banks just prior to the deregulation. The investment banks listed here include banks that get acquired subsequently and independent investment banks that have underwritten at least 5 issues over the sample period 1990-2003, and remain in operation throughout this period.
Investment banks that remain independent Investment banks that get acquired after 1996 Small independent investment banks
13
In the next section, I evaluate the impact of mergers on retention of pre-merger
underwriting customers.
4. Impact on retention
In this section I examine whether existing clients are more likely to switch
relationships if their underwriter undergoes a merger with a commercial bank. I
compare the retention rate for existing relationships at the time of the merger and the
relationships formed by merged banks, to the retention rate of benchmark banks after
the deregulation. I also compare these retention rates of different banks to their own
retention rates prior to the deregulation.
The underwriting data for defining relationships comes from 2,004 issuing
companies that issue securities on at least 2 occasions, and that have a relationship
with at least one of the 116 investment banks under consideration. A relationship is
defined to start when a bank underwrites an issue for a client, and is defined to break if
and when that client chooses another bank for a subsequent issue. Both debt and
equity issues are used in this analysis. The results are substantially similar if I focus
only on companies where both initial and follow-on issues are equity deals, and are
left unreported.
The total number of observations with follow-on issues, where the relationship
under consideration can be either continued or broken, is 3,958. A merger occurs
between two successive issues for 318 of these observations. I investigate the impact
of a merger occurring between two issues on the continuance or severance of a
relationship.
14
Table 4 provides a comparison of retention rates of benchmark banks,
investment banks that subsequently undergo a merger, and the merged commercial-
investment banks over the period 1990-2003. During the period 1990-1996, before the
relaxation of the Glass Steagall Act, there is very little difference in the retention rates
of benchmark banks and the investment banks that subsequently undergo a merger.
Both these types of banks retain about 60% of their relationships. However,
immediately after the mergers, the retention rate of acquired banks drops significantly.
Only 34% of relationships are retained when a merger occurs between an initial issue
and a follow-on issue. The retention rate for new relationships formed by merged
banks after 1996 is higher at 51%, but compares unfavorably with that of benchmark
banks at 57%.
Table 4: Relationship retention rates of different types of banks over the period 1990-2003
Bank type Total
relationships
Number of relationships
retained Number of
relationships lost
Retention %
Total relationships
Number of relationships
retained Number of
relationships lost
Retention %
Benchmark banks 743 439 304 59% 1,161 675 486 58% Investment bank that gets acquired subsequently 624 374 250 60%
Investment bank that is acquired between initial and follow-on issue 305 108 197 35%
Merged commercial-investment bank 259 132 127 51%
Other banks 442 195 247 44% 593 231 361 38%
Total 1,809 1,008 801 56% 1,725 915 810 53%
Follow-on issue during 1990-1996 Follow-on issue during 1997-2003
This table compares retention rates of different types of banks before and after the deregulation. Panel A considers relationships where the follow-on issue occurs between 1990 and 1996, prior to the deregulation, while Panel B only considers relationships where the follow-on issue occurs after 1996. In Panel A, the categories of banks considered are the 7 largest independent investment bank or benchmark banks, investment banks that subsequently undergo a merger and other banks. In Panel B, the bank categories are 7 largest independent investment bank or benchmark banks, investment banks that undergo a merger between the initial and follow-on issue, merged commercial-investment bank and other banks.
Panel A Panel B
15
Table 5 provides a comparison of retention rates across different mergers. It
can be seen that the retention rate suffers in almost all the mergers, relative to
retention rates before the merger. These results are consistent with the findings in a
contemporaneous paper on the market share impact of these mergers (Chaplinsky and
Erwin (2005)).
16
Robinson Humphrey Suntrust Bank 2 0 0% 24 10 42%Equitable Securities 2 2 100%
Morgan Keegan Regions Bank 1 0 0% 7 3 43%
Total 305 108 35% 782 456 58%
Table 5 : Impact of different mergers on existing relationships This table compares the retention rate of acquired banks at the point of merger to their retention rate before the deregulation.
Investment bank that undergo mergers Acquiring commercial bank
Relationships with merger
between issues Relationships
retained
Retention % at
merger Relationships
total Relationships
retained Retention
rate
Salomon Brothers Citibank 26 8 31% 157 92 59% Smith Barney Citibank 17 6 35% 75 40 53% Salomon Smith Barney Citibank 17 9 53% Wertheim Schroder Citibank 2 0 0% 11 5 45% Alex Brown Bankers Trust/ Deutsche Bank 35 15 43% 89 61 69% Bankers Trust Alex Brown Deutsche Bank 17 6 35% Montgomery Securities Nations Bank 36 11 31% 87 64 74% Nations-Montgomery Bank Of America 8 2 25% Chemical Bank Chase Manhattan 2 1 50% Hambrecht-Quist Chase Manhattan 23 5 22% 49 34 69% JP Morgan Chase Manhattan 13 6 46% Dillon Read Swiss Banking Corporation 17 9 53% 34 19 56% Swiss Bank-Warburg Dillon Read UBS 5 1 20% 7 2 29% Paine Webber UBS 6 1 17% 73 43 59% Bradford UBS 1 1 100% 13 7 54% Robertson Stephens Bank of America/ Bank of Boston / Fleet 19 8 42% 51 32 63% Bank of America-Robertson Stephens Bank of Boston/ Fleet 11 5 45% Robertson Stephens - Bank of Boston Fleet 9 5 56% Oppenheimer CIBC 10 4 40% 19 7 37% Cowen Societe General 8 1 13% 9 6 67% Piper Jaffrey US Bank 5 1 20% 13 6 46% Dain Bosworth Rauscher Royal Bank of Canada 4 0 0% 11 3 27% Tucker Anthony Sutro Royal Bank of Canada 12 6 50% Chicago ABN Amro 2 0 0% 0 Furman Selz ING/ ABN Amro 3 2 67% 13 6 46% First Union Securities Wachovia 2 1 50% Everen First Union 2 0 0% 7 3 43% Wheat FBS First Union 1 0 0% 11 4 36% First Albany First Union 2 0 0%
Mc Donald Key Bank 1 1 100% 3 1 33% Van Kasper Wells Fargo 1 0 0% Wassertein Perella Dresdner Bank 1 0 0% 3 0 0%
After merger Prior to merger
17
Thus, Table 4 and Table 5 suggest that mergers are associated with greater
customer defection. To confirm whether this result holds up in a multivariate analysis,
I perform a logistic regression to examine whether an issuer stays with or switches
away from its underwriter for a follow-on issue.
The logistic model is displayed in Table 6. The dependent variable is a binary
variable that takes value 1 if a relationship is broken. The independent variables of
interest are the dummy variables corresponding to the different type of banks. The
dummy for the benchmark banks is omitted. Panel A of Table 6 shows the results for
the entire sample of follow-on issues from 1990-2003, while Panel B shows the results
for relationships where the follow-on issues occur during the period 1997-2003.
Literature on commercial banking mergers suggest that as merged banks become
bigger, they tend to reduce focus on small customers. To ensure that the results are not
driven by any such change in focus, in Column B, I exclude small issues. For this
purpose, small issues are defined as those with issue-sizes smaller than US$50 million
in inflation-adjusted 1990 dollars. The regression also includes controls for several
other factors that have been shown to impact retention rates. These include controls
for the strength of the relationship, the rating of the issuer, and prior lending
relationships of the underwriter with the issuer, the availability and quality of analyst
research coverage, and time between initial and follow-on issue. I also include
industry dummies for the 35 industries that the issues have been categorized into
(based on Fama-French classification).
18
Dependent variable:Binary variable x, x=1 if relationship is broken, x=0 if relationship is retained
Underwriter type variables
PREMERGERANDBEFORE1996 1.360 *** 1.430(2.96) (1.98)
MERGER 1.923 *** 2.03 *** 2.13 *** 2.14 ***(4.47) (4.67) (3.39) (3.35)
MERGEDCBIB 1.708 *** 1.68 *** 1.53 ** 1.41(3.36) (3.41) (2.25) (1.75)
OTHERS 2.319 *** 2.23 *** 2.27 *** 2.07 ***(8.72) (6.81) (5.56) (4.27)
Relationship variables
EXCLUSIVEREL 0.571 *** 0.567 *** 0.465 *** 0.400 ***-(4.50) -(3.63) -(4.38) -(4.25)
NOTEXCLUSIVEREL x NUMBEROFINTERACTIONS 0.815 *** 0.803 *** 0.736 *** 0.698 ***-(3.06) -(2.85) -(3.64) -(3.64)
PRIORLENDINGREL 0.979 0.960 1.140 1.142-(0.10) (0.18) (0.57) (0.55)
Issuer and issue variables
YEARSBETWEENISSUES 1.460 *** 1.420 *** 1.394 *** 1.360 ***(14.75) (11.64) (7.94) (6.71)
UNRATED/JUNKRATEDDUMMY 0.486 *** 0.528 *** 0.471 *** 0.489 ***-(6.43) -(4.02) -(4.61) -(3.05)
SIZEUPGRADE 0.960 1.000-(0.43) (0.03)
SMALLIIB x SIZEUPGRADE 1.732 ** 3.245 ***(2.09) (2.79)
IPODUMMY 0.723 *** 0.718 ** 1.028 0.718 **-(3.05) -(2.70) (0.12) -(2.70)
CHANGEINPRODUCTTYPE 1.707 *** 1.628 *** 1.541 *** 1.628 ***(6.88) (4.88) (3.95) (4.88)
Analyst Research variables
ALLSTARCOVERAGE 0.688 ** 0.785 0.602 ** 0.690-(2.30) -(1.13) -(2.21) -(1.31)
NORESEARCHCOVERAGE 1.150 1.130 1.011 0.949(1.55) (0.99) (0.10) -(0.35)
Industry dummies present
Total observations 3,955 2,310 1,924 1,290 N (1/0) 2074/1883 1146/1172 1086/837 712/578Pseudo R-square 0.12 0.13 0.08 0.11
Follow-on issues during
1997-2003
Follow-on issues during
1990-2003
Follow-on issues during
1990-2003
Follow-on issues during 1997-
2003
Column A: All issues Column B: Midsized and large issues
Table 6: Logistic regression model of relationship retention (Odds ratios)
NOTEXCLUSIVEREL x RELINTERACTIONNO - number of transactions between underwriter and issuer in current relationship for non-exclusive relationships PRIORLENDINREL - Dummy variable that takes the value 1 if the underwriter has underwritten any loans for the issuer between during 1990-1996
YEARSBETWEENISSUES - Number of years between initial and follow-on issue, 1-12 months implies 1 year, 12-24 months implies 2 years and so on UNRATED/JUNKRATED DUMMY - dummy variable for unrated/junk-rated issuer based on S&P long-tem issuer rating or SDC's initial issue rating
IPODUMMY - dummy variable that takes value 1 if initial issue is an IPO CHANGEINPRODUCTTYPE - dummy for change in product type from debt to equity or equity to debt between initial and final issue
NORESEARCHCOVERAGE - 1 if issuer does not receive research coverage from any bank during the 12 months prior to follow-on issue
This table shows the results of a logistic model that examines the impact of mergers on relationships retention. The dependent variable is a binary variable that takes the value 1 if the relationship is broken, and is 0 if the relationship gets retained. The independent variables of interest are: the variable MERGER which takes the value 1 if the bank undergoes a merger between the initial and follow-on issue, and the variable MERGEDCBIB which takes the value 1 if the initial issue is underwritten by a merged commercial-investment bank. Another variable of interest is PREMERGERANDBEFORE1996, this variable takes the value 1 when both the intial and follow-on issues take place between 1990-1996 and the underwriter is an investment bank that undergoes a merger after 1996. The omitted variable for underwriter type is large independent investment bank or benchmark bank. Thus coefficients on MERGER and MERGEDCBIB are relative to the benchmark banks. Column A shows results when the initial issue can be of any size, while in Column B, I exclude small issues. For this purpose, large issues are those whose issuesize is greater than US$ 150 mn in inflation-adjusted 1990 dollars. Medium-sized issues are those with issue-sizes between US$ 50 and 150 mn in 1990 dollars,
OTHER - underwriter is any other type of bank including small independent investment bank, commercial bank or when the follow-on issue occurs after 1996, even an investment bank that has not yet undergone a merger
ALLSTARCOVERAGE - 1 if underwriter's analyst covers the issuer during the 12 months prior to the issue and is ranked ALL-STAR by Institutional Investor in the year
before the follow-on issue
The model is estimated with clustered standard errors for issues by the same company. Results are shown in odds ratios, with figures in parentheses showing standard errors. ***, ** and * indicate significance at the 1%, 5% and 10% levels
while small issue are those with issue-sizes below US $50 mn in 1990 dollars. Inflation numbers are obtained from Bureau of Labor Statistics's CPI-U historical index. The regression also includes several control variables as described below:
SMALLIIB x SIZEUPGRADE - interaction between a dummy variable for small independent investment bank and issue size changing from small to medium or big, or medium to big
EXCLUSIVEREL - dummy variable with value 1 if the issuer does not have a relationship with any other underwriter between 1980 and date of initial issue
19
The results are shown in odds ratios or exponentiated coefficients. The model is
estimated with clustered standard errors for the issuer, because relationships of the
same issuer may be related.
Table 6 shows that other factors being equal, the odds of a relationship being
broken are twice as high when the bank undergoes a merger, as the odds of the
relationships being broken when the bank is a benchmark bank. Retention improves
somewhat for relationships formed by merged banks, but holding other variables
constant, these odds of these relationships being broken are still 1.7 times as high as
the odds for relationships with benchmark banks. A comparison of results in Columns
A and B shows that the negative impact of mergers on retention holds even after
excluding small issues. Lending relationships do not significantly improve the odds of
a relationships being retained. This is probably because there are very few matches
across lending and underwriting relationships. Similarly, the analyst research variables
are not very significant in influencing relationship retention. Other variables impact
retention in predictable ways.
Thus mergers have a significant negative impact on retention. Further, in
unreported results I find that when clients switch away from their relationship bank
that is undergoing or has undergone a merger, they do so to non-merging banks nearly
70% of the time. Merged banks, as a group manage to obtain only 35% of customers
that switch away from benchmark and other banks. Thus merging and merged banks
lose more of their existing clients compared to benchmark banks, and seem to be less
successful than benchmark banks in gaining clients who switch.
20
In the next section, I compare the ability of merged banks and benchmark
banks to win underwriting deals from customers who switch as well as from new
customers.
5. Mergers and customer acquisition
In this section, I compare the relative ability of merged banks and benchmark
banks to win underwriting clients over 1998-2003. It is not sufficient to examine this
by looking only at changes in market share. This is because market shares are also
impacted by exogenous changes in the composition of demand, for instance in the mix
of issues by industry, and bank-specific capabilities. Further, if banks anticipated any
such changes, then it may affect their decision to undergo a merger; if so, any
measured negative impact of mergers on market share could reflect selection bias.
Hence, I compare the relative abilities of merged banks and benchmark banks to win
new clients, on a deal-by-deal basis, controlling for the pre-deregulation ability of the
bank to win the deal.
In this section, I restrict my attention to equity deals, since acquired investment
banks, (with the exception of Salomon Smith Barney) primarily contributed expertise
in the equity deals only. I use a conditional logit model of underwriter choice,
analyzed separately for IPOs and SEOs, to evaluate the ability of banks to win
individual deals. I study the impact of the mergers over two time periods, 1998-2000
and 2001-2003. This is because 1) There is substantial variation in the underwriting
market over these two periods in the number and mix of issues. 2) I want to examine if
the banks that merge early improve their performance over time.
21
A unit of observation in the model is an underwriting deal. I consider
underwriting issues of a particular product by a company on a given day as one deal.
For each deal, the issuing company is allowed to choose one of 116 banks as lead
underwriter for the deal. The independent variable of interest is a dummy variable
corresponding to bank type, specifically whether it is a merged bank, a benchmark
bank or other type of bank. Banks that undergo a merger are classified as a merged
bank if the effective date of the merger is before the issue date.
I further classify merged banks into those that merge before between 1997 and
1999 (EARLYMERGEDCBIB) and those merge later than 1999
(LATERMERGEDCBIB). I include controls for the pre-deregulation expertise of each
bank in the industry and product type of the deal. For this purpose, I calculate the
market share of each bank for the product (industry) under consideration relative to the
market leader in that product (industry) prior to the deregulation. In addition, I also
include controls for existing underwriting and lending relationships of each bank with
the company and the research strength of each bank in the industry of the issuer.
Table 7 shows the results of the conditional logit model for the IPO market.
The omitted category for bank-type is the category corresponding to benchmark
banks. In column A, I do not differentiate between issues of different sizes, while in
column B, I include an interaction term between issue size and the dummy variable
indicating a merged bank. It can be seen that most merged banks perform much worse
than benchmark banks over the 1998-2000 period. Further, the negative impact of the
merger is greater for medium and large size issues. This argues against the possibility
that the results are driven by merged banks seeking to reduce focus on small issues.
22
Lead IPO underwriter choice(Odds ratios )
Underwriter type variables
EARLYMERGEDCBIB 0.457 *** 0.32 *** 0.467*** 0.297 ***-(6.62) -(7.35) -(3.34) -(4.17)
EARLYMERGEDCBIB * SMALLISSUE 1.926*** 2.73***(3.81) (2.64)
RS_BANKOFBOSTONFLEETDUMMY 3.974 *** 4.001 ***(8.77) (8.83)
LATERMERGEDCBIB 0.280 *** 0.130 ***-(2.93) -(2.73)
LATERMERGEDCBIB * SMALLISSUE 4.280(1.67)
CSFBDUMMY 3.775 *** 3.775 *** 1.04 1.04(8.75) (8.75) (0.88) (0.88)
OTHERS 0.278 *** 0.278 *** 0.190 *** 0.190 ***
-(7.18) -(7.18) -(4.69) -(4.69)Deal expertise variables
PREDEREG_PRODUCTEXPERTISE 1.016 *** 1.016 *** 1.006 * 1.006 *(8.02) (8.02) (1.79) (1.84)
PREDEREG_INDUSTYEXPERTISE 1.011 *** 1.011 *** 1.011 1.011 ***(6.83) (6.83) (4.34) (4.22)
Relationship variables
PREDEREG_LENDINGRELDUMMY 4.801 *** 4.801 *** 0.657 0.733(3.70) (3.70) -(0.53) -(0.39)
Analyst Research variables
INDUSTRYCOVERAGE 2.177 ** 2.177 ** 2.660 *** 2.170 ***(4.62) (4.62) (2.84) (4.61)
ALLSTARINDUSTRYCOVERAGE 0.966 0.966 0.657 0.960-(0.28) -(0.28) -(0.53) (0.39)
Total observations 702 702 157 157 Pseudo R-square 0.27 0.27 0.23 0.24
Column A Column B
2001-2003
Column A Column B
1998-2000
Table 7: Logit model for Customer acquisition rates of merging and benchmark banks IPO market - (Odds ratios)
The model is estimated with clustered standard errors for issues by the same company. Results are shown in odds ratios, with figures in parentheses showing standard errors. ***, ** and * indicate significance at the 1%, 5% and 10% levels
This table shows the results of a conditional logit model of lead underwriter choice in the IPO (initial public offering) market. The choice set includes the 116 banks under consideration in this paper. The independent variables of interest are: the variable EARLYMERGEDCBIB which takes the value 1 if the bank is a merged commercial-investment bank that has undergone a merger between 1997 and 1999, and LATERMERGEDCBIB, a dummy variable that is equal to 1 if the bank is a merged commercial-investment bank that has undergone a merger after 1999. In addition, there are dummy variables corresponding to specific mergers whose performance is significantly different from the overall group of mergers. The omitted category for underwriter type is the dummy corresponding to the 7 benchmark large independent investment banks. However, even among these 7 banks, CSFB does significantly better, thus I also include a dummy corresponding to CSFB.
Finally, OTHERS corresponds to all other types of banks, including small independent investment banks, commercial banks and investment banks that have not yet undergone a merger, but will do so. Column A does not consider differences by issue size. In Column B, I include interactions between the merged bank variables, EARLYMERGEDCBIB and LATERMERGEDCBIB, and SMALLISSUE, a dummy variable corresponding to issuesizes below $50 mn in inflation-adjusted 1990 dollars. The model includes variables corresponding to the pre- deregulation expertise of each bank in underwriting all IPOs and in underwriting equity issues from the particular industry. These are defined as:
Finally, the model also includes controls for pre-deregulation lending relationships of each bank with the issuer and analysts' coverage of the
ALLSTARINDUSTRYCOVERAGE - Bank covers at least one of the issuing companies in the industry of the issuer in the 12 months prior to issue date and the analyst is rated All-Star by Institutional Investor in the previous year
PREDEREG_INDUSTRYEXPERTISE - 100 * Ratio of number of equity issues underwritten by bank in the industry of the issuer over the period 1993-1997 relative to bank underwriting maximum number of issues in the same industry over the period 1993-1997 PREDEREG_PRODUCTEXPERTISE - 100 * Ratio of number of IPO issues underwritten by bank over the period 1995-1997 relative to bank underwriting maximum number of IPOs over the same period
INDUSTRYRESEARCHCOVERAGE - Dummy variable that takes value 1 if bank covers at least one of the issuing companies in the industry of the issuer in the 12 months prior to issue date
PREDEREG_LENDINGRELATIONSHIPDUMMY- dummy variable that takes value one if the bank had acted as lead arranger for any loans to the issuer over the period 1990-1997
23
The decline in performance of merged banks that merge before 1999 persists
over 2001-2003. However, it is difficult to draw any conclusions about the long-term
performance of the mergers from this. This is because the IPO market shrinks
significantly between 2001 and 2003, and this may have contributed partially to the
poor performance of the merged banks over 2001-2003.
Table 8 shows the results of a similar analysis for the SEO market. Merged
banks, in general perform poorly relative to benchmark banks over the 1998-2000
period, though there are some exceptions. In unreported results, I find that the negative
effect of mergers on performance is even stronger, if I exclude small issues. Column A
shows the results for all SEO issues over the period 1998-2003. Column B contains
results for new SEO issuers that are companies that have not made any issues over
1990-1997, and hence do not have any pre-deregulation period relationships. The
results are substantively similar in columns A and B. This suggests that merged banks
do not do as well as benchmark banks even among companies that are issuing for the
first time.
The table further shows that some banks that undergo a merger between 1997
and 1999 improve their performance over the period 2001-2003. But for many banks,
the effects of the decline in performance after mergers persist over the longer term.
24
Conditional logit model of lead SEO underwriter choice
Conditional logit model of lead SEO underwriter choice
(Odds ratios) (Odds ratios)
Underwriter type variables Underwriter type variables
EARLYMERGEDCBIB 0.469 *** 0.45 *** EARLYMERGEDCBIB 0.422 *** 0.52 ***
-(5.28) (4.67) -(5.76) -(3.31)
RS_BANKOFBOSTONFLEETDUMMY 3.992 *** 5.364 *** UBSDUMMY 2.72 *** 5.364 ***
(5.65) (4.81) (4.85) (4.81)
OPPENHEIMER_CIBCDUMMY 4.143 *** 4.688 *** COWEN_SGDUMMY 3.06 *** 4.688 ***(5.37) (4.08) (3.02) (4.08)
DR_SBCDUMMY 8.021 *** 6.376 PIPER_USBANKDUMMY 2.37 ** 0.92(3.83) (1.76) (2.18) -(0.11)
FS_INGDUMMY 2.831 *** 2.243(1.01) (1.36)
LATERMERGEDCBIB LATERMERGEDCBIB 0.390 *** 0.58-(3.52) -(1.63)
CSFBDUMMY 2.163 *** 2.338 *** CSFBDUMMY 0.94 1.07 ***
(4.25) (3.29) (0.75) (0.75)
OTHERS 0.454 *** 0.49 *** OTHERS 0.600 ** 0.78-(5.20) -(2.98) -(2.39) -(0.38)
Deal expertise variables Deal expertise variables
PREDEREG_PRODUCTEXPERTISE 1.009 *** 1.010 *** PREDEREG_PRODUCTEXPERTISE 1.010 *** 1.008 ***
(5.23) (4.10) (4.95) (3.33)
PREDEREG_INDUSTYEXPERTISE 1.009 *** 1.008 *** PREDEREG_INDUSTYEXPERTISE 1.005 *** 1.007 ***
(5.28) (3.33) (2.90) (3.44)Relationship variables Relationship variables
PREDEREG_LENDINGRELDUMMY 1.850 ** 2.968 ** PREDEREG_LENDINGRELDUMMY 2.240 ** 2.990 **
(2.09) (2.08) (2.32) (1.94)
PREDEREG_UNDRELDUMMY 19.172 *** PREDEREG_UNDRELDUMMY 4.998 ***(17.93) (2.53)
POSTDEREG_UNDRELDUMMY 45.879 *** 82.290 *** POSTDEREG_UNDRELDUMMY 16.363 *** 19.430 ***
(22.73) (20.32) (12.08) (13.81)Analyst Research variables Analyst Research variables
INDUSTRYCOVERAGE 5.830 ** 6.598 *** INDUSTRYCOVERAGE 7.008 * 6.120 ***
(7.64) (5.46) (6.86) (5.50)
ALLSTARINDUSTRYCOVERAGE 1.100 1.145 ALLSTARINDUSTRYCOVERAGE 1.320 ** 1.490 *
(0.46) (0.73) (2.02) (2.23)
Total observations 693 331 Total observations 375 229Pseudo R-square 0.41 0.46 Pseudo R-square 0.35 0.33
Column A - All issuers
Column B - New issuers
Column A - All issuers
Column B - New issuers
1998-2000 2001-2003
Table 8: Comparison of customer acquisition rates of merging and benchmark independent banks - SEO market (Odds ratios)
PREDEREG_LENDINGRELATIONSHIPDUMMY- dummy variable that is one if the bank acted as lead arranger for any loans to the issuer over the period 1990-1997 PREDEREG_UNDRELDUMMY- dummy variable that is one if the bank acted as lead underwriter for any equity issue made by the issuer over the period 1990-1997
ALLSTARINDUSTRYCOVERAGE - Bank covers at least one of the issuing companies in the industry of the issuer in the 12 months prior to issue date and the analyst is rated All-Star by Institutional Investor in the previous year The model is estimated with clustered standard errors for issues by the same company. Results are shown in odds ratios, with figures in parentheses showing standard errors. ***, ** and * indicate significance at the 1%, 5% and 10% levels
POSTDEREG_LENDINGRELATIONSHIPDUMMY- dummy variable that takes value one if the bank acts as lead underwriter for any previous equity issue by the issuer between 1998 and the date of the issue
Finally, the model also includes controls for pre-deregulation lending relationships of each bank with the issuer and analysts' coverage of the industry the issuer belongs to:
INDUSTRYRESEARCHCOVERAGE - Dummy variable that takes value 1 if bank covers at least one of the issuing companies in the industry of the issuer in the 12 months prior to issue date
This table shows the results of a conditional logit model of lead underwriter choice in the SEO (seasoned equity offerings) market. The choice set includes the 116 banks under consideration in this paper. Column A includes SEO issues by all issuers, while Column B only considers SEO issues by new issuers. New issuers are those that have not made any equity issues in the 1990-1997 period and therefore do not have any pre-deregulation underwriting relationships. The independent variables of interest are: the variable EARLYMERGEDCBIB which takes the value 1 if the bank is a merged commercial-investment bank that has undergone a merger between 1997 and 1999, and LATERMERGEDCBIB, a dummy variable that is equal to 1 if the bank is a merged commercial-investment bank that has undergone a merger after 1999. In addition, there are dummy variables corresponding to specific mergers whose performance is significantly different from the overall group of mergers. The regression also includes merger-specific dummy variables. The omitted category for underwriter type is the dummy corresponding to the 7 benchmark large independent investment banks. However, even among these 7 banks, CSFB does significantly better, thus I also include a dummy corresponding to CSFB. Finally, OTHERS corresponds to all other types of banks, including small independent investment banks, commercial banks and investment banks that have not yet undergone a merger, but will do so. The model also includes variables corresponding to the pre-deregulation expertise of each bank in underwriting all SEOs and in underwriting equity issues from the particular industry of the issuer. These are defined as: PREDEREG_INDUSTRYEXPERTISE - 100 * Ratio of number of equity issues underwritten by bank in the industry of the issuer over the period 1993-1997 relative to bank underwriting maximum number of issues in the same industry over the period 1993-1997 PREDEREG_PRODUCTEXPERTISE - 100 * Ratio of number of SEO issues underwritten by bank over the period 1995-1997 relative to bank underwriting maximum number of SEOs over the same period
25
Finally, in Table 9, I examine changes in market shares for different types of
banks over time. I also relate these changes to overall composition of issues by size.
The table shows that the share of small issues falls sharply over time throughout the
entire sample period. This change is accompanied by a steady fall in the share of small
investment banks over time. The investment banks that subsequently undergo a
merger gain in market share at the expense of small investment banks, before the
deregulation and the mergers, that is, until 1996. However, immediately after the
mergers, almost all the gains in market share go to benchmark banks. This suggests
that the acquired investment banks competed effectively with large investment banks
until just before the mergers; however, immediately after the mergers they lose out to
the benchmarks banks in terms of relative market share.
Overall, the results in sections 4 and 5 strongly show that mergers resulted in
significant customer defection. It would be useful if we were able to identify why such
defection occurs. In the next section, I analyze cross-sectional differences in the
performance of merged banks, and attempt to relate it to the characteristics of the
merging banks and the merger.
26
Table 9 - Evolution of market share over the period 1990-2003 (by number of issues)
IPO market
Year Merging and merged banks
% of total
Benchmark banks
% of total
Small investment banks
% of total
Commercial banks
% of total Total Large
issues Medium sized issues
Small issues
1991-92 182 32% 139 24% 251 44% 0 0% 572 3% 12% 85% 1993-94 255 31% 170 21% 395 48% 5 1% 825 3% 13% 85% 1995-96 337 36% 225 24% 350 37% 28 3% 940 3% 18% 79% 1997-98 167 30% 171 30% 201 36% 23 4% 562 4% 20% 75% 1999-2000 201 34% 325 54% 63 11% 9 2% 598 6% 45% 48% 2001-03 57 32% 88 50% 27 15% 4 2% 176 10% 43% 47%
SEO market
Year Merging and merged banks
% of total
Benchmark banks
% of total
Small investment banks
% of total
Commercia l banks
% of total Total Large
issues Medium sized issues
Small issues
1991-92 188 37% 140 28% 170 34% 4 1% 502 3% 19% 78% 1993-94 204 36% 162 29% 191 34% 7 1% 564 4% 23% 73% 1995-96 344 47% 199 27% 169 23% 22 3% 734 5% 34% 60% 1997-98 242 42% 210 37% 91 16% 31 5% 574 7% 40% 53%
1999-2000 184 35% 267 50% 68 13% 11 2% 530 24% 47% 29% 2001-03 221 34% 307 47% 105 16% 15 2% 648 8% 41% 50%
This table shows the evolution of market shares of different types of banks over time. The period 1990-1996 corresponds to the period before the deregulation and mergers. The bank types include merging/merged banks, benchmark large independent investment banks, small independent investment banks and commercial banks that do not undergo mergers. The table also shows change in the composition of issues by issue-size over time. For this purpose, large issues are those where issuesize is greater than US$ 150 mn in inflation-adjusted 1990 dollars. Medium-sized issues are those with issue-sizes between US$ 50 and 150 mn in 1990 dollars, while small issue are those with issue-sized below US $50 mn in 1990 dollars. Inflation numbers are obtained from Bureau of Labor Statistics's CPI-U historical index.
Share of issues by issue-size Number of issues by type of bank
Number of issues by type of bank Share of issues by issue-size
27
6. Discussion
Existing research has identified several mechanisms through which mergers
may adversely impact customer retention and acquisition. These include:
Costs of integration in synergy-driven mergers - Within the organizational economics
literature, Dessein et al (2005) and Hart and Holmstrom (2002) emphasize costs of
integration in synergy-driven mergers. These papers highlight the trade-off involved in
preserving incentives for effort and adjusting incentives to achieve co-ordination
across businesses. If incentives are indeed weakened, and synergies between lending
and underwriting take time to achieve or are not as large as envisioned, this may be
expected to have a negative impact on customer retention and acquisition, at least in
the short term.
Employee resistance and turnover, resulting in loss of capabilities of acquired
banks - There is a large literature within the organizational theory field that highlights
the potential for negative individual and collective employee reactions to mergers,
particularly employees of the acquired firms. This employee resistance and in some
cases, employee turnover can be expected to have a negative impact on customer in a
relationship-intensive business like underwriting.
Internal focus on merger integration and customer uncertainty - Homburg and
Bucerius (2005) discuss the potential for a strong internal focus during the integration
phase, as merging firms integrate people, systems, and processes. They argue that this
can result in reduced attention paid to customer issues, and considerable customer
28
uncertainty with respect to contact persons, quality, etc. Such customer uncertainty
may lead to customer defection.
Detailed information about the post-merger integration process and internal
organization would be required to precisely identify which of these mechanisms drives
the results in this paper. Such information would perhaps be available only through
case studies of the mergers. I do not attempt such an ambitious exercise. Instead, I
draw some preliminary conclusions on these mechanisms by using qualitative
information about the mergers obtained from analyst reports. I relate this information
to cross-sectional differences in merger impact, in the empirical results.
It can be seen from Tables 7 and 8 that there are important cross-sectional
differences in the impact of mergers on customers. The results for the 1998-2000
period in Table 8 show that four merged banks (CIBC - Oppenheimer, SBC Warburg
– Dillon Read, ING Barings – Furman Selz and Robertson Stephens - Bank of Boston
- Fleet Bank) improve their ability to gain new customers after the merger. In three of
these mergers, the commercial banks were international banks that wished to build an
investment banking presence in the US. It seems likely that these mergers were less
subject to the incentive-related costs of integration, relative to acquisitions by
domestic commercial banks, who sought to integrate the lending and underwriting
businesses. Indeed, there is only one merger involving a domestic commercial bank,
where the merged bank improves upon its capability to attract customers. In this
merger between Robertson Stephens and Bank of Boston - Fleet Bank, the investment
bank and commercial banks were located on opposite coasts. Analyst reports suggest
29
that Robertson Stephens was allowed to operate separately from the beginning of the
merger, with no attempt to achieve cross-selling synergies.
An examination of the results in Table 8 also shows that UBS, US Bank-Piper
and Societe Generale-Cowen improve upon their ability to gain customers over the
2001-2003 period. Two of these mergers were unwound in 2003 and 2004, with
analysts’ reports suggesting that the commercial banks did not perceive synergies
between lending and underwriting to be significant. This suggests that the
improvement in performance occurred concurrently with the banks giving up on
efforts to achieve synergies. This is consistent with the argument in the costs of
integration literature that it may be better to operate businesses independently, if
synergies between them are not sufficiently high.
With respect to the employee turnover mechanism, analyst reports suggest that
two mergers were particularly impacted by large-scale personnel turnover - the merger
between Alex Brown- Deutsche Bank and the merger between Montgomery Securities
and Nations Bank-Bank of America. The results from Table 7 and 8 show that like
other merged banks, these two banks also experienced considerable customer
defection during 1998-2000 after the merger. This happened even though issues by
technology companies dominated the underwriting market, and these two banks
specialized in such issues. These results show that employee turnover does
significantly impact customer relationships. However, the negative impact of mergers
on customer retention and acquisition is not limited to these two mergers.
Performance also suffers significantly in mergers, where integration of the
businesses of the acquired bank and acquiring bank is particularly complicated. These
30
include the mergers between Citigroup-Salomon Smith Barney, JP Morgan -Chase and
UBS-Swiss Banking Corporation. This is consistent both with intuition and with the
notion that companies are focused internally after mergers. On the other hand, the
three mergers that enjoyed early success (CIBC - Oppenheimer, SBC Warburg –
Dillon Read, ING Barings – Furman Selz) were smaller and possibly less complicated.
Analyst reports also suggest that in these three mergers, the acquiring banks already
had an organizational structure in place for co-coordinating their lending and
underwriting businesses (partly on account of previous investment bank acquisitions),
making the integration task simpler.
Overall, it appears as if all these factors together may have played a role in
impacting customer retention and acquisition after mergers.
7. Conclusion
This paper examines the impact of mergers on the ability to retain and gain
customers in the acquired business, using a sample of mergers between investment
banks and commercial banks over the period 1997-2001. Using detailed customer-
level data for the period 1990-2003, I find that mergers result in a significant fall in
customer retention and acquisition rates, relative to non-merging investment banks, as
well as performance of the acquired investment banks prior to the mergers. The results
are not driven by a deliberate attempt on the part of the merged banks to drop some
customers or selection. Cross-sectional differences in mergers suggest that synergy-
related costs of integration, employee turnover and internal integration issues may
have all played a role in causing customer defection. Of these factors, the trade-off
31
between motivation and coordination has received the least attention in research on
post-merger integration. This theory suggests that synergy-related gains may be
difficult to achieve in settings where preserving motivation is important. Future
research of other mergers with similar trade-offs would be very interesting and
valuable.
32
References
1. Aghion, P. and J. Tirole, Formal and real authority in organizations. Journal of Political Economy, 1997. 105: p. 1-29.
2. Andrade, G., M. Mitchell, and E. Stafford, New Evidence and Perspectives on Mergers. Journal of Economic Perspectives, 2001. 15: p. 103-120.
3. Ang, J. and T. Richardson, The underwriting experience of commercial bank affiliates prior to the Glass-Steagall Act: A re-examination of evidence for passage of the act. Journal of Banking and Finance, 1994. 18: p. 351-395.
4. Asker, J. and A. Ljungqvist, Sharing Investment Bankers, in Working paper. 2006, New York University.
5. Baker, G., R. Gibbons, and K. Murphy, Informal Authority in Organizations. Journal of Law, Economics and Organization, 1999. 15: p. 56-73.
6. Baker, G., R. Gibbons, and K. Murphy, Relational Contracts and the Theory of the Firm. Quarterly Journal of Economics, 2002. CXV.
7. Benston, G., The Separation of Commercial and Investment Banking. 1990, New York: Oxford University Press.
8. Benveniste, L.M., et al., Evidence of Information Spillovers in the Production of Investment Banking Services. Journal of Finance, 2003. 58: p. 557-608.
9. Benzoni, L. and C. Schenone, Conflict of interest or certification? Evidence from IPOs underwritten by the firm’s relationship bank, in Working Paper. 2005, SSRN.
10. Berger, A.N., et al., The effects of bank mergers and acquisitions on small business lending. Journal of Financial Economics 1998. 50: p. 187–229.
11. Bharath, S., et al., So what do I get? The bank’s view of lending relationships, Working Paper, 2004, New York University.
12. Boot, A., T.M. Todd, and A. Thakor, Expansion of Scope and Scale in Banking: Don’t banks know the value of focus? SSRN Working Paper Series, 1998.
13. Booth, J. and R. Smith, Capital raising, underwriting and the certification hypothesis. Journal of Financial Economics, 1986. 15: p. 261-281.
33
14. Calomiris, C.W. and J. Karceski, Is the Bank merger Wave of the 1990s Efficient? Mergers and Productivity, ed. S.N. Kaplan. 2000: University of Chicago Press/NBER
15. Carter, R. and S. Manaster, Initial public offerings and underwriter reputation. Journal of Finance, 1990. 45: p. 1045-1067.
16. Caves, R., Mergers, Takeovers and economic efficiency. International Journal of Industrial Organization, 1992. 7: p. 151-174.
17. Chaplinksy, S. and D. Erwin, Great expectations: Banks as Equity Underwriters in Working paper series 2005, SSRN.
18. Corwin, S. and P. Schultz, The role of IPO underwriting syndicates: Pricing, information production, and underwriter competition. Journal of Finance, 2005. 60: p. 443-486.
19. Crane, R. and D. Eccles, Customer relationships in the 1990s, in Financial Services: Perspectives and Challenges, S. Hayes, Editor. 1993, Harvard Business School Press: Boston.
20. Datta, S., Organizational fit and merger performance: Effects of post-acquisition integration. Strategic Management Journal, 1991. 12: p. 287-297.
21. Delong, G., Stockholder gains from focusing versus diversifying mergers. Journal of Financial Economics, 2001. 59: p. 221-252.
22. Dessein, W., Authority and Communication in Organizations. Review of Economic Studies, 2002. 69: p. 811-838.
23. Dessein, W., L. Garicano, and R. Gertner, Organizing for Synergies, in Working paper. 2005, University of Chicago.
24. Drucker, S., Information asymmetries, cross-product banking mergers, and the effects on corporate borrowers, in SSRN working paper series. 2005, SSRN.
25. Drucker, S. and M. Puri, On the Benefits of Concurrent Lending and Underwriting. Journal of Finance, 2005. 60: p. 2763-99.
26. Drucker, S. and M. Puri, Banks in Capital Markets: A Survey, in Handbook of Corporate Finance: Empirical Corporate Finance, B.E. Eckbo, Editor. 2006, Elsevier/North-Holland.
27. Dunbar, C., Factors affecting investment bank initial public offering market share. Journal of Financial Economics, 2000. 55: p. 3-41.
34
28. Eccles, R. and R. Crane, Doing Deals: Investment Banks at Work. Harvard Business School Press. 1988.
29. Fama, E. and K. French, Industry costs of equity. Journal of Financial Economics, 1997. 43: p. 153-194.
30. Fang, L., Investment bank reputation and the price and quality of underwriting services. Journal of Finance, 2005. 60: p. 2729-2761.
31. Fernando, C.S., V.A. Gatchev, and P.A. Spindt, Wanna dance? How firms and underwriters choose each other. Journal of Finance, 2005. 60 : p. 2437-2469
32. Gande, A., M. Puri, and A. Saunders, Bank Entry, Competition, and the Market for Corporate Securities Underwriting. Journal of Financial Economics. 54: p. 165-195.
33. Gande, A., et al., Bank Underwriting of Debt Securities: Modern Evidence. Review of Financial Studies. 10: p. 1175-1202.
34. Gulati, R., Does familiarity breed trust? The implications of Repeated Toes for Contractual Choices in Alliances. Academy of Management Journal, 1995. 38: p. 85-112.
35. Hart, O. and B. Holmstrom, A Theory of Firm Scope. 2003 in Working paper, SSRN.
36. Hart, O. and J. Moore, Property Rights and the Nature of the Firm. Journal of Political Economy, 1990. 98: p. 1119-1158.
37. Hart, O. and J. Moore, On the Design of Hierarchies: Coordination Versus Specialization, in mimeo. 2000, Harvard University.
38. Healy, P., K. Palepu, and R. Ruback, Do mergers improve corporate performance? Journal of Financial Economics, 1992. 31: p. 135-176.
39. Heckman, J. and R. Robb, Alternative method for solving the problem of selection bias in evaluating the impact of treatments on outcomes, in Drawing Inferences from Self-Selected Samples H. Wainer, Editor. 1986, Springer-Verla: New York.
40. Holmstrom, B., The Firm as a sub economy Journal of Law, Economics and Organization, 1999. 15: p. 74-102.
41. Holmstrom, B. and J. Roberts, Boundaries of the Firm Revisited. Journal of Economic Perspectives, 1998. 12: p. 73-94.
35
42. Homburg, C. and M. Bucerius, A Marketing Perspective on Mergers and Acquisitions: How Marketing Integration Affects Postmerger Performance. Journal of Marketing 2005. 65: p. 95-113.
43. Houston, J. and C. James, Bank information monopolies and the mix of private and public debt claims. Journal of Finance, 1996. 51: p. 1863-1889.
44. Houston, J.F., C.M. James, and M.D. Ryngaert, Where do merger gains come from? Bank mergers from the perspective of insiders and outsiders. Journal of Financial Economics, 2001. 19: p. 285-331.
45. James, C., Relationship-specific assets and the pricing of underwriter services. Journal of Finance, 1992. 47: p. 1865-1885.
46. James, C.M. and W. Peter, Borrowing Relationships, Intermediation and the Cost of Issuing Public Securities. Journal of Financial Economics, 1990.
47. Kanatas, G. and J. Qi, Underwriting by commercial banks: Incentive conflicts, scope economies, and project quality. Journal of Money, Credit, and Banking, 1998. 30: p. 119-133.
48. Kanatas, G. and J. Qi, Integration of Lending and Underwriting: Implications of Scope Economies. Journal of Finance, 2003. 58: p. 1167-1191.
49. Kaplan, S.N., A Clinical Exploration of Value Creation and Destruction in Acquisitions: Organizational Design, Incentives, and Internal Capital Markets, in Mergers and Productivity, S.N. Kaplan, Editor. 2000, University of Chicago Press / NBER: Chicago.
50. Kaplan, S.N., Introduction, in Mergers and Productivity, S.N. Kaplan, Editor. 2000, University of Chicago Press / NBER: Chicago.
51. Krigman, L., W. Shaw, and K. Womack, Why do firms switch underwriters? Journal of Financial Economics, 2001. 60: p. 245-284.
52. Kroszner, R. and R. Rajan, Is the Glass-Steagall Act justified? A study of the U.S. experience with universal banking before 1933. American Economic Review, 1994. 84: p. 810-832.
53. Larsson, R. and S. Finkelstein, Integrating strategic, organizational, and human resource perspectives on mergers: Case Survey of Synergy realization. Organization Science, 1999. 10: p. 1-23.
54. Liedkta, J., Synergy revisited: How a screwball buzzword can be good for the bottom-line. Business Strategy Review, 1998. 9: p. 45-55.
36
55. Ljungqvist, A., F. Marston, and W.J. Wilhelm, Scaling the hierarchy :how and why banks compete for syndicate co-management appointments, in SSRN working paper series. 2006a, SSRN.
56. Ljungqvist, A., F. Marston, and W.J. Wilhelm, Competing for Securities Underwriting Mandates: Banking Relationships and Analyst Recommendations. Journal of Finance, 2006b. 61: p. 301-340.
57. Maddala, G.S., Limited Dependent and Qualitative Variables in Econometrics. 1983, New York: Cambridge University Press.
58. Mailath, G.J., V. Nocke, and A. Postlewaite, The Disincentive Effects of Internalizing Externalities, in mimeo. 2002, University of Pennsylvania.
59. McFadden, D., Conditional logit analysis of qualitative choice behavior, in Frontiers in Econometrics, P. Zarembka, Editor. 1973, Academic Press: New York.
60. McGuckin, R. and S. Nguyen, On Productivity and Plant Ownership Change: New Evidence from the Longitudinal Research Database. Rand Journal of Economics, 1995. 26: p. 257- 276.
61. Mueller, D., Mergers and market share. Review of Economics and Statistics, 1985. 67: p. 259-267.
62. Mulherin, H.J. and A.L. Boone, Comparing Acquisitions and Divestitures. Journal of Corporate Finance, 2000. 6: p. 117-39.
63. Nanda, V. and V.A. Warther, Price of Loyalty: An Empirical Analysis of Underwriting Relationships and Fees, in SSRN Working Paper Series. 1998, SSRN.
64. Narayanan, R., K. Rangan, and N. Rangan, The Role of Syndicate Structure in Bank Underwriting. Journal of Financial Economics, 2004. 72: p. 555-580.
65. Peek, J. and E.S. Rosengren, Small business credit availability: How important is size of lender?, in Financial System Design: The Case for Universal Banking .I. Walters, Editor. 1996, Irwin Publishing: Homewood, IL.
66. Peek, J. and E.S. Rosengren, Bank consolidation and small business lending: It’s not just bank size that matters. Journal of Banking and Finance, 1998. 22: p. 799-820.
67. Puri, M., The long term default performance of bank underwritten security issues. Journal of Banking and Finance, 1994. 18: p. 397-418.
37
68. Puri, M., Commercial banks in investment banking: Conflict of interest or certification role? Journal of Financial Economics, 1996. 40: p. 373-401.
69. Puri, M., Commercial banks as underwriters: Implications for the going public process. Journal of Financial Economics, 1999. 54: p. 133-163.
70. Rajan, R., Commercial Bank Entry into the Securities Business: A Survey of Theories and Evidence, in Universal Banking: Financial System Design Reconsidered, A. Saunders and I. Walter, Editors. 1996, Irwin: Chicago.
71. Ravenscraft, D.J. and F.M. Scherer, The Profitability of Mergers. Journal of Industrial Economics, 1989. 7: p. 101-16.
72. Sapienza, P., The effects of banking mergers on loan contracts. Journal of Finance, 2002. 57: p. 329-367.
73. Saunders, A., Conflicts of Interest: An Economic View, in Deregulating Wall Street: Commercial Bank Penetration of the Corporate Securities Market I. Walter, Editor. 1985, John Wiley and Sons: New York.
74. Saunders, A. and I. Walter, Universal Banking in the United States. What Could We Gain? What Could We Lose? 1994, New York: Oxford University Press.
75. Schenone, C., The Effect of Banking Relationships on the Firm’s IPO Underpricing. Journal of Finance, 2004. 59: p. 2903-2958.
76. Schoar, A., Effects of Corporate Diversification on Productivity. Journal of Finance, 2002. LVII: p. 2379-2403.
77. Singer, J. and J.B. Willett, Applied Longitudinal Data Analysis. 2003: Oxford University Press.
78. Sufi, A., Does Joint Production of Lending and Underwriting Help or Hurt Firms? A Fixed Effects Approach, in Unpublished manuscript, 2004.
79. Sufi, A., Information asymmetry and financing arrangements: Evidence from syndicated loans. Journal of Finance, 2007. 62 : p. 629-668
80. Teece, D., et al., Understanding Corporate Coherence, theory and evidence. Journal of Economics Behavior and Organization. 23: p. 1-30.
81. Vayanos, D., Optimal Decentralization of Information Processing in the Presence of Synergies, in Review of Economic Studies. 2002.
82. Yasuda, A., Do Bank Relationships Affect the Firm's Underwriter Choice in the Corporate-Bond Underwriting Market? Journal of Finance, 2005. 60: p. 1259-1292.