Competition, financing, and performance of
newly listed firms in Europe
Master’s thesis Business Administration – Finance F.M. den Blanken
Student number: 1339818
First supervisor: Dr. H.J. von Eije Second supervisor: A.J. Meesters
Abstract
We study the performance of initial public offerings (IPOs) from ten European countries during 2001 – 2006, in relation to comparable firms that stay private and comparable firms that were already listed. Our study is different from the extant literature, because (1) we compare IPOs with both listed and with unlisted firms, (2) we compare them both before and after the equity offering, (3) we use three measures of performance: return on assets, and its constituents return on sales and capital efficiency, (4) we study whether competition makes a difference for the IPO firms in comparison to their benchmarks, and (5) we study if the relaxation of financial constraints after the IPO affects the performance of IPO firms compared to their benchmark firms. We find that the capital efficiency prior to the IPO is lower for firms going public in relation to unlisted firms. This result is stronger for firms facing high competition. Median regressions show that the reduction of leverage in the year of the IPO improves the IPO’s long-term performance. However, the positive effect of deleveraging diminishes when companies reduce their leverage ratio too much, and when companies receive too much cash from their issue. We find a positive relation between high competition and capital efficiency of firms that stay private, and a negative relation for IPO firms. Finally, median leverage of listed firms is low, and for these firms, increasing leverage has a positive effect on long-term performance.
JEL classification: G3, L1
1. Introduction
Selling equity shares to public investors is a major source of financial resources for publicly listed companies. Certo et al. (2009) show that the amount of capital raised by initial public offerings (IPOs) is large: the average U.S. initial public offering (IPO) raised more than $215 million in the period 1998 through 2007. The average European company in our sample raised €83 million from its initial offering between 2001 and 2006.1
Much research has been done on companies going public, not only because of the money involved in these IPOs, but also because IPOs provide a unique insight into fundamental firm characteristics (Poulsen & Stegemoller, 2008). Certo et al. (2009) find that the literature on IPOs encompasses several disciplines, such as human resource management, finance, and entrepreneurship.
A major branch of the financial IPO literature focuses on long-run performance.2 Studies on the long-term performance of IPOs often find that they do not perform as well as comparable firms (e.g. Ritter, 1991; Loughran & Ritter, 1995; and Schultz, 2003). However most of the long-run performance literature only compares the performance of IPOs with that of listed firms, using market returns as performance measure. It is however equally relevant to look at unlisted firms as the offering should have positive effects on a company compared to staying private. We add to the discussion by studying long-run IPO performance in relation to both listed and unlisted firms. Since the performance of private firms cannot be measured using stock prices, we use finance based accounting data. We also add to the literature by looking at both ex ante and ex post performance of IPOs relative to comparable firms.
In recent years, more attention is given to the relation between competition and performance. A theory by Povel & Raith (2004) argues that financially constrained firms show a negative relation between competition and performance, while the opposite is true for unconstrained firms. Schoubben & Van Hulle (2009) argue that unlisted firms have a higher chance of being financially constrained than listed firms. Their results indeed show that there is a negative relation between the level of competition and performance for unlisted firms in Belgium, and a positive for listed firms. However, they do not distinguish between firms that stay private, and firms that eventually issue public equity. We do make this distinction, and
1 Data on IPO proceeds are obtained from Bureau van Dijk Electronic Publishing Zephyr, a database containing information on 800,000 M&As and IPOs of European companies between 2000 - present. Zephyr gives the ‘deal equity value’ for 189 of our 262 initial IPOs
study the performance of newly listed, previously listed and unlisted firms in relation to competition.
We construct a database with which we attempt to compare the performance of newly listed firms with the performance of private and previously listed firms. We apply rigid selection criteria to collect benchmark firms. In contrast with Ritter (1991) and Loughran & Ritter (1995), we do not restrict our benchmark sample to only one company per IPO. Instead, our sample includes all companies that comply with the country, industry and size selection criteria, regardless of whether they are listed or private firms.3 This approach results in a sample that gives a good view on the level of competition faced by the IPOs, since with more similar companies within an industry and country, there will be more competition between the companies. Furthermore, since the selection criteria are not relaxed when we do not find a matching company for our IPO, we have an extremely close match between our IPOs and their control firms.4
Our major measure of performance is return on assets. However, we disentangle this variable into a profitability measure, and a capital efficiency measure. This approach allows us to study the reasons for possible lower performance of IPOs deeper, by allocating it to one of the two (or to both) measures.
We do not find lower ex ante return on assets of IPOs compared with unlisted firms. However, return on sales is higher for the IPOs, while their median capital efficiency is lower. The higher return on sales disappears in the third year after the IPO, while the lower capital efficiency persists over the years. Median regressions show a positive effect on long-term performance for IPO firms of a reduction in leverage in the year of their offering. However, this effect is diminished when the reduction in leverage is too high. Furthermore, a large increase in cash levels also has a negative relation with long-term performance of IPO firms.
The lower capital efficiency of IPOs compared with firms that stay private seems to come from IPOs facing high competition. Low-competition IPOs show higher return on sales and higher return on assets than comparable low-competition unlisted firms. Median regressions show a positive relation between competition and long-term capital efficiency for unlisted firms, but a negative relation for IPO firms. This is contrary to Schoubben & Van
3 U.S.-based empirical research on publicly listed companies has the advantage of high information availability and uniform rules and restrictions. Studies on multiple countries, such as ours, are difficult due to differences between countries. This is one of the reasons why Pagano et al. (1998) focus solely on Italy. We address this criticism by using comparables from the same country as the IPO’s.
Hulle (2009), who find a negative relation between competition and performance for unlisted firms.
The results on comparisons of IPO firms and listed firms show that ex ante capital efficiency is higher for IPOs, but the difference disappears after the offering. Furthermore, there is some evidence of long-term lower performance of IPOs compared with previously listed firms. Median regressions show a positive relation of a change in leverage and long-term performance of previously listed firms. We find a relatively low leverage ratio for previously listed firms, which could explain the positive relation: when leverage becomes too low, firm-performance decreases.
2. Literature
2.1 Why do companies go public?
To be able to understand the effects of going public, it is important to first identify the reasons why companies go public. As Pagano et al. (1998) state, “The conventional wisdom is that going public is simply a stage in the growth of a company.” However conventional wisdom may be partly wrong, as all countries, including those with highly developed stock markets such as the U.S., have a number of very large companies that are not publicly listed. Apparently, going public is also a matter of choice rather than only a phase that all companies enter at a certain time.
Pagano et al. (1998) review the main theories on costs and benefits of going public. Costs of going public include the fixed costs of the offering process, adverse selection, and loss of confidentiality. Fixed costs include costs made for the offering itself, such as underwriter and registration fees, but also annual costs of being public, e.g. auditing costs and stock exchange fees. Much of these costs do not increase much with firm size, therefore, smaller firms face higher relative costs of listing. The adverse selection theory argues that potential investors are less informed than the issuers. To persuade investors to invest in the IPO, the shares have to be sold at a discount. Chemmanur & Fulghieri (1999) argue that this discount will be higher for smaller and younger companies, because they have lower visibility and shorter track records. Pagano et al. mention two major costs due to the loss of confidentiality. First, companies may have to reveal information that is valuable to competitors. Second, they are under closer scrutiny from tax authorities, decreasing tax evasion opportunities.
now publicly available, increasing competition between lenders. Next, public shares are more easily traded than private shares, and by divesting from the company and re-investing in other companies, initial shareholders can diversify. Furthermore, by creating the danger of hostile takeovers and by allowing the market to assess management performance, management discipline can be increased. Fifth, listing on a major exchange can act as a visibility instrument for companies. Also, going public enables an owner/shareholder to arrange the change of control of his company. Finally, if companies recognize that other companies in their industry are overvalued, they may choose to go public to benefit from this overvaluation.
There is a significant amount of literature that supports these theories on the costs and benefits of going public. Ritter (1987) finds evidence that the fixed costs of going public and underpricing together account for between 21 and 32 percent of the realized market value of the IPO. Jain & Kini (1999) find that firms that go public in a stage too early in their lifecycle have a higher probability of failure, which supports the adverse selection and fixed costs theories. Mikkelson et al. (1997) find evidence that U.S. start-up companies go public to finance investments, i.e. to overcome borrowing constraints. Brau & Fawcett (2006) interview a large number of CFOs of U.S. companies on the motivation, timing and underpricing of IPOs. They find the main reason for CFOs to issue stock is to enable acquisitions. Furthermore, CFOs time the issue based on overall market conditions and apply underpricing to compensate new investors for the risk they take. In a similar survey among CFOs of companies from 12 European countries, Bancel & Mittoo (2009) find that contrary to U.S.-based CFOs, European CFOs view outside monitoring as a major benefit of going public. Pagano et al. (1998) find that the main factors that predict the probability of a company to go public are the industry to-book and the size of the company: larger and higher market-to-book ratio companies are more likely to issue stock. The large size is in line with the adverse selection and fixed cost theories, while the high industry market-to-book points to the window-of-opportunity theory and growth opportunities. Kim & Weisbach (2008) study for what purposes companies use the cash generated by issuing stock in a large sample of IPOs and seasoned equity offerings (SEOs) in 38 countries. They examine changes in a range of balance sheet and profit & loss account variables such as cash holdings, capital expenditures and R&D expenses. The main findings include that for every dollar raised in the issue, cash holdings increase with almost 39% even after four years. R&D and capital expenditures also rise significantly, which confirms the borrowing constraints hypothesis. Lerner (1994) studies the timing of biotechnology firms to go public and finds that they tend to go public during market peaks. This supports the windows of opportunity theory.
generate better information than private financiers. However, this diverse information must be obtainable at low costs, “because stock market analysts tend to duplicate each other’s efforts.” Poulsen & Stegemoller (2008) study two ways of transferring a firm from private to public ownership: issuing IPOs, or being acquired by a public company. They find that there is a higher probability of an IPO when companies have relatively high borrowing constraints and higher investment opportunities. However, they find mixed support for the information availability theory of Subrahmanyam & Titman (1999). Ritter & Welch (2002) argue that public trading in itself can increase the value of a company, “as it may inspire more faith in the firm from other investors, customers, creditors, and suppliers.” Furthermore, they state that there are first-mover advantages to going public.
In contrast with the authors mentioned above, Helwege & Packer (2009) investigate reasons for staying private. Unlike other studies, they find that neither growth opportunities nor diversification arguments decrease the probability of a firm staying private. Private benefits of control, e.g. high autonomy for top management, strongly favor companies to stay private.
2.2 IPO performance
The seminal article on long-run IPO performance by Ritter (1991) describes the average three-year buy-and-hold performance of 1,526 U.S. IPOs during 1975 to 1984. He finds that the average IPO in that period returned $ 1.34 for every dollar invested, while the average already listed firm returned $ 1.62 for every dollar. A lot of research has been done on the subject since then, confirming the persistence of IPO long-run underperformance.
returns to measure firm performance. A number of authors report a decline of ex post IPO profitability, e.g. Jain & Kini (1994) and Mikkelson et al. (1997).
Although the main literature focuses on performance of U.S. IPOs, long-run underperformance appears to be a world-wide phenomenon. Levis (1993) finds average 36-month underperformance of IPOs in the UK for the period 1980 to 1988. Keloharju (1993) finds significant underperformance in Finland. Aggarwal et al. (1993) investigate IPO activity in Brazil, Chile and Mexico, and find underperformance in all three countries. The results of Lee et al. (1996) suggest significant underperformance in Australia. Evidence of underperformance in the Chinese market is recently provided by Cai et al. (2008). Álvarez & González (2005) find a decline in return on assets for Spanish IPOs during 1985 – 1997.
There is less literature on the performance of IPOs compared with unlisted firms. Pagano et al. (1998) use a sample of 69 Italian IPOs during 1982 through 1992, and find that their profitability and investments decline after the IPO.
Not all literature concludes on negative incentives for investing in IPOs. Brav & Gompers (1997) find that underperformance is typical for both issuing and non-issuing small, low book-to-market firms. In a very recent study, Chen et al. (2010) show that adding an IPO portfolio to a benchmark portfolio increases diversification benefits: they find a 5.5% increase of the Sharpe ratio after they include IPO stocks in the tangency portfolio. Schultz (2003) finds no underperformance if he uses calendar time instead of event time returns.
2.3 Explaining IPO performance
Several theories exist for explaining IPO performance. Miller (1977) assumes that there are constraints on shorting IPOs, and that investors have heterogeneous expectations of IPO performance. Due to these shorting constraints, initial investment is skewed towards optimistic investors, which leads to abnormally high initial stock returns. In subsequent years the IPO will regress to its mean value, as shorting constraints are removed. In support of this theory, Ritter (1991) finds a negative correlation between initial abnormal returns and long-run IPO performance. Purnanandam & Swaminatham (2004) also find support for this overvaluation theory.
A number of studies find other explanations for underperformance. Teoh et al. (1998) attribute underperformance to ex ante earnings management. They find that IPOs with unusually high reported earnings in the IPO-year show poor stock returns in the three years after the IPO. They argue that managers apply window dressing prior to the IPO, to create investor over-optimism on the future performance of the IPO. This theory is supported by the research of Brau & Fawcett (2006), who find in their survey that IPO CFOs “view strong historical earnings as the most positive signal in the IPO process.” In another over-optimism theory, Heaton (2002) thinks that not only investors but also managers tend to be overoptimistic. When the issue brings in (excess) capital, they spend it on negative NPV projects. This eventually reduces the value of the company. Another finding originates from Brav & Gompers (1997), who argue that underperformance is not an IPO problem. They find that IPO underperformance is strongly skewed towards small and high-growth companies.
As stated above, a number of studies find a decline in ex post IPO performance using return on assets. One of the explanations is based on adverse selection: companies may go public if they foresee that their profitability is about to decline permanently (Leland & Pyle, 1977). This would also cause low survivorship of IPOs relative to non-issuing companies: the permanent decline in profitability can lead to termination of the company. Second, in line with Jensen & Meckling (1976), Jain & Kini (1994) suggest that a reduction in management ownership leads to a heightened conflict of interest between pre-issue owners and new shareholders. Moreover, more dispersed ownership in general (whether or not the management was the initial owner) could increase agency costs, due to lower monitoring. This could lead management to increase their private benefits to the detriment of the company’s profitability. Another approach of Pagano et al. (1998) suggests that IPOs have an incentive to give a fair or even inflated view on their profitability prior to and after the issue, while companies that stay private may attempt to understate the real value of their assets for tax evasion purposes. Since profitability is measured as earnings divided by assets, this decreases the observed profitability. Finally, the ex post decline in profitability of the IPO firm may be caused by a lack of sales growth opportunities or an ex post decline in capital expenditures, which would lead to a decline in operating performance (Jain & Kini, 1994).
2.4 Competition and performance
the major determinant of profitability. This theory predicts that companies facing heavy competition will have lower profitability.
Looking at other performance measures, a number of theories predict improved efficiency and productivity growth with higher competition, due to better managerial incentives and monitoring. Nickell (1996) reviews a number of theories that predict a positive relation between productivity and competition. The main theories assume agency cost to be a major determinant. One theory poses that more competition enables firm owners to compare managerial effort better. This gives incentives to managers to increase their effort, thereby increasing efficiency and productivity. A second theory suggests that higher competition increases the sensitivity of profits to managerial effort. Thus, owners have a “greater incentive to ensure that manager’s effort is kept high.” Karuna (2007) finds that firms have stronger incentives for management when competition is greater. Januszewski et al. (2002) also find support for the prediction of a positive relation using a sample of German manufacturing firms. Hou & Robinson (2006) find a significantly positive relation between competition and stock returns. They argue that barriers of entry “insulate firms from undiversifiable distress risk”, and financial theory dictates that lower risk means lower stock returns. A related explanation for their empirical findings is that firms in high-competitive environments are riskier because they invest more in innovation.
The literature mentioned above focuses on listed firms. There is limited literature on the performance of unlisted firms in relation to the level of competition. Povel & Raith (2004) argue in a theoretical paper that firms that have financing constraints will perform worse when facing high competition, while the opposite counts for firms with no financing constraints. Schoubben & Van Hulle (2009) find empirical support for this theory. They argue that unlisted firms are more financially constrained than listed firms due to information asymmetries. They find evidence of a negative relation between competition and profitability and productivity for firms with financing constraints. For firms without these constraints, a positive relation is found. We attempt to add to the discussion by studying to what extent high or low competition influences long-term IPO performance.
market share to the detriment of competitors that stay private. They mention a number of mechanisms leading to this increased market share, including gaining increased credibility with customers and suppliers, being able to hire higher quality employees, and being able to acquire companies in the same industry through takeovers using their own shares. However, to our knowledge, there is no literature that studies the effect of the level of competition on the performance of IPOs.
One conclusion that can be drawn from the discussion above, is that it is important to make clear what measure of firm performance is actually used: while many theories focus on profitability (e.g. Porter 1980), others look at productivity (e.g. Nickell, 1996). The next section goes into this question. Furthermore, what follows from Schoubben & Van Hulle (2009) is that going public could increase profitability and productivity in highly competitive environments, due to the reduction of financial constraints.
2.5 Measures of performance
In their review of IPO literature in management and entrepreneurship journals, Certo et al. (2009) find three common approaches to measuring long-term IPO performance: accounting-based variables; financial market variables; and survival rates. As stated in the introduction, most literature on IPO long-run performance focuses on comparing IPOs with already listed companies. This enables researchers to use data on publicly listed companies, which are more widely available.5 A second advantage of only looking at firms listed on a stock exchange is that performance can be measured using market returns. Since an important part of our study focuses on unlisted companies, we cannot use market based financial measures. Instead, we look at accounting based measures.
As is shown by Sapienza & Grimm (1997), different measures can lead to different conclusions on the performance of a company. Perhaps the most used performance measure is operating return (Murphy et al., 1996; Mikkelson et al., 1997; and Álvarez & González, 2005). Murphy et al. (1996) also use sales growth as an accounting based measure. Pagano et al. (1998) are among the few who look at firms that stay private to test post-IPO performance. They investigate the effect of going public by separately estimating regressions of a number of traditional accounting-based variables including return on assets, leverage, and growth through the years following the offering.6
5 The low availability of (financial) data on unlisted firms can be seen in our IPO sample, where we lose almost two thirds of our initial 262 IPOs due to the lack of pre-IPO data.
2.6 Benchmarking
It is important to learn which types of benchmarks are used in the relevant literature. There appear to be three main strategies of comparing long-run performance (Barber & Lyon, 1997).
First, Ritter (1991) matches IPOs with listed firms based on industry (three-digit SIC codes) and size (market value). He uses only one benchmark firm per IPO. For a significant number of companies a matching firm in the same three-digit SIC industry does not exist: only 36% of the IPOs could be matched with a firm of the same three-digit SIC industry.7 Yi (2001) uses a similar selection procedure. He is able to closely match 59% of the IPOs, the rest of the IPOs are either matched to a company with equal two-digit SIC, or, if still no benchmark company exists, with a company that has similar market cap. As in Ritter (1991), Loughran & Ritter (1995) choose one publicly listed benchmark firm per IPO. Contrastingly, Loughran & Ritter (1995) only match firms based on market capitalization. They argue that industry-matching can cause dilution of statistical analysis results, because if companies time their issue to take advantage of wide misvaluations, matching IPOs with industry-peers will reduce the ability to identify abnormal returns. Therefore, Loughran & Ritter (1995) do not apply an industry selection criterion. Mikkelson et al. (1997) use a group of matching firms by industry and size. If they do not find comparable companies within the IPO’s four-digit SIC, they loosen the industry restriction.
The second benchmarking strategy is shown in Brav & Gompers (1997), who use a group of U.S. IPOs similar to Ritter (1991) and Loughran & Ritter (1995). They pool IPOs and peers in groups of similar size and book-to-market value. Each quarter, the groups are revised based on the changes in these two variables. Such a Fama & French (1993) three factor approach is also used in more recent studies such of Eckbo & Norli (2005) and Chen et al. (2010).
Schultz (2003) uses a third benchmarking method: in his study, he subtracts the value of the CRSP value or equal-weighted index return from the IPO return. This strategy is applied by many other authors as well, although they often use different indices such as the NASDAQ index (e.g. Krigman et al., 1999; and Yi, 2001).
Barber & Lyon (1997) evaluate the performance of these three approaches. They find that the approaches using reference portfolios, i.e. the market index and three factor methods, are prone to a number of biases, e.g. a rebalancing bias (reference portfolio are usually calculated assuming periodic rebalancing, while the sample firms are not, according to Barber sales/employees); investment (capital expenditures/sales); output (real sales); employment; leverage; and dividends (cash dividends/sales).
& Lyon (1997)); and a skewness bias (long-run abnormal returns are positively skewed). The control firm approach alleviates those biases, and yields well-specified test statistics.
3. Hypotheses
Our study is different from the extant literature for various reasons. First, we do not only compare IPOs with listed firms, but also with firms that stay private. Second, we do not only compare IPOs with listed and unlisted firms after the public offering, but also before the equity issue. Third, we compare the performance of IPO firms with their benchmark not only with respect to return on assets, but also with respect to the two constituents of return on assets, namely return on sales and capital efficiency. Fourth, we are the first to study whether competition makes a difference for the IPO firms in comparison to their benchmarks. Finally, we study if the relaxation of financial constraints after the IPO affects the performance measures of IPO firms in comparison with their benchmark firms. In order to address these issues, we develop the following hypotheses.
The first set of hypotheses test whether we find differences in performance of IPOs compared with unlisted and listed firms prior to the offering. We argue that IPOs are financially constrained ex ante, so that they have to forego projects that increase performance. Therefore, we expect IPOs to show lower performance than comparable firms before their stock issue.
H1,0: The performance of European manufacturing IPOs during 2001 – 2006 before the
equity offering is not different from the median performance of comparable firms; and
H1,A: The performance of European manufacturing IPOs during 2001 – 2006 before the
equity offering is different from the median performance of comparable firms.
Theories exist that explain increasing as well as decreasing performance after an IPO. The most intuitive expectation is an increase in performance. Due to the proceeds from the equity issue, financial constraints are relaxed, enabling IPOs to invest in profitable investment opportunities. Therefore, we expect IPO performance to increase ex post. In the long run, we expect that IPO performance does not deviate from that of comparable firms. We develop the following hypotheses on ex post IPO performance.
H2,0: The performance of European manufacturing IPOs during 2001 – 2006 after the equity
H2,A: The performance of European manufacturing IPOs during 2001 – 2006 after the equity
offering is different from the median performance of comparable firms.
The theories on competition and performance discussed earlier provide hypotheses for the sign of the relation between competition and performance. Povel & Raith (2004) argue that there is a negative relation between competition and performance of financially constrained firms: these companies have to pass on projects that increase profit and efficiency. Thus, ex ante, we expect that the performance of IPOs facing high competition is lower than that of IPOs facing low competition. After the IPO, we expect that the high-competition firms are able to improve performance due to relaxation of the financial constraints. Therefore, we expect to find a positive relation between competition and long-term performance for IPOs. Due to their high access to different sources of capital, we expect the same positive relation for previously listed firms. Schoubben & Van Hulle (2009) argue that unlisted firms are more likely to be financially constrained and find a negative relation between unlisted firm competition and performance. However, in contrast with Schoubben & Van Hulle (2009) we distinguish between firms that stay private, and firms that go public within the time period we study. Therefore, the sample of firms that stay private should be less financially constrained, and the sign of the relation between competition and performance is difficult to anticipate.
H3,0: The performance of European manufacturing firms facing high competition during 2001
– 2006 is not different from comparable firms facing low competition; and
H3,A: The performance of European manufacturing firms facing high competition during 2001
– 2006 is different from comparable firms facing low competition. 4. Data & Methodology
4.1 Sample Selection
strengthening our statistical results. As far as we know, this matching procedure is unique in the literature.8
The main data are obtained from Bureau van Dijk Electronic Publishing (BvDEP) Amadeus. This database provides information on balance sheet and income statements, IPO dates, shareholder structure, etc. Datastream provides market return data on publicly quoted companies, and is used to complement the database. 9 To test pre-IPO performance, we need data on the year before an IPO took place. Since two important performance measures – return on assets and capital efficiency – are measured using assets at the end of the previous year, we require data on the two years preceding the IPO. To test long-term IPO performance, we require data on the third year after the IPO. In order to allow for this information, only IPOs between 1 January 2001 and 31 December 2006 are included in the IPO sample.
Three main restrictions are applied to the companies in our sample. First, the sample is restricted to companies from EU-15 countries.10 Furthermore, only manufacturing companies (NACE revision 2 section C; first two digits of the NACE codes between 10 and 33) are considered.11 Finally, the accounting data of small companies are typically unreliable (Pagano et al., 1998). Therefore, the sample is restricted to companies with total assets of at least €500,000 in their last available account year. These restrictions lead to a selection of 340,400 companies.
From these 340 thousand companies three main samples are constructed: an IPO sample containing IPOs between 2001 and 2006; a sample of benchmark companies that are not publicly listed; and a sample of benchmark companies that are quoted on an EU-15 stock exchange.
IPO sample. According to Amadeus, there are 366 companies with a primary issue
date between 1 Jan 2001 – 31 Dec 2006. These companies are crosschecked in Datastream, which reduces the sample to 262 real IPOs.12 There are also companies for which Amadeus gives no IPO date, but which are publicly quoted. The IPO dates of these companies are
8 Ritter (1991) mentions a ‘survivorship bias’ in constructing the benchmark: some benchmark companies delist, go bankrupt or are taken over during the after-IPO years. We do not have this problem, since we select all matching firms of an IPO. Thus, if a company disappears from our sample, the remaining firms constitute the benchmark for that IPO.
9 All data from Amadeus and Datastream are given in Euro’s. Euro values for non-Eurozone countries are calculated according to the exchange rate policies of the software.
searched in Datastream. Companies with an IPO date within one of the relevant years for this study are added to our IPO sample, this increases our sample with five companies.13
Only companies for which information on total assets is available in at least two years prior to the issue year are useful for our analysis. This reduces the number of IPOs to 95 companies. The rest of the IPOs are removed from the sample. Furthermore, one company is removed because it could not be found in Datastream under the same name and IPO date, giving a final IPO sample of 94 companies.
Benchmark samples. The benchmark companies are initially selected based on the
same industry type as the IPO14. All companies in the IPO sample are removed from the benchmark sample. Next, a size criterion is applied: only companies that have average total assets between 1/3 and 3 times that of the relevant IPO over the two years before the issue are included. This gives a benchmark of acceptable size, comprised of companies that are comparable in total asset size. This also gives our benchmark a unique feature: the number of peer firms can be different for different IPO firms: some IPOs have only one benchmark firm, while others have more than 100. The initial benchmark sample consists of 5,916 companies. The benchmark remains the same over all five relevant years of the IPO (two years pre-IPO, the issue year and two years post-IPO). In line with Ritter (1991) and Loughran & Ritter (1995), no periodical adjustment is applied.
The total benchmark sample is split into a benchmark containing publicly quoted companies and one with unlisted companies. All companies for which Amadeus gives an IPO date and which are currently still listed are transferred to the listed benchmark. This gives an initial listed benchmark containing 56 listed firms. Eleven listed companies that were IPOs themselves within one to five years before the issue of the linked IPO are removed from the sample. This ensures that the remaining listed companies have a long-term listing experience. Next, within the remaining sample is one listed company for which Amadeus does not provide an IPO date and eleven companies that do have an IPO date but are currently unlisted. The respective primary listing and delisting dates of these companies are checked in Datastream, adding eight companies to the final listed benchmark. 15 The final listed
13 The sample contains 163 companies without an IPO date that are listed according to Amadeus. Of these, six companies cannot be found in Datastream. Of the remaining firms, four companies have a date of first listing in Datastream between January 1, 2001 and December 31, 2006. Furthermore, one IPO that is currently delisted according to Amadeus is still listed according to Datastream. This company is also added to the IPO sample.
benchmark contains 53 companies. For 29 percent of our IPOs we find at least one matching firm. This is somewhat lower than Ritter (1991), who finds matching firms with the same three-digit SIC code for 36 percent of his sample.16 The remaining 5,872 companies comprise the unlisted benchmark. For 88 percent of our IPOs we find at least one similar private firm. Contrary to Ritter (1991) and Yi (2001), we do not relax our selection criteria, because we feel this would dilute our statistical results.
4.2 Descriptive Statistics
Table 1 contains a summary of data on the entire IPO sample (Panel A), the firms that stay private (Panel B), and the firms that were already listed before our period of interest (Panel C). Because the distribution of accounting variables may be skewed, and the mean is sensitive to outliers, we focus on the median in the discussion of our descriptive statistics and results.
The median IPO in our sample is nine years old, has a market capitalization of €73 million, and a market-to-book ratio of 2.5 at the end of the IPO-year. Furthermore, the IPOs report a median €42 million in assets, 15 percent return on assets (ROA), a leverage ratio of 0.63, and more than 160 employees in the year prior to the IPO. The median firm that stays private is 19 years old in the year of its corresponding IPO, has €3.3 million in assets, a ROA of 11 percent, a leverage ratio of 0.62, and over 70 employees in the pre-IPO year. The median listed firm is 21 years old, has a market cap of €80 million, and a market-to-book of 2 in the IPO-year. Furthermore, the pre-IPO year shows a median ROA of 11 percent, a leverage ratio of 0.34, €72 million in assets, and over 460 employees. Below, we compare some of the statistics of the IPOs and the two benchmarks.
Table 1 about here
Age. The median IPO is much younger than companies who stay private or who are
already listed: 9 years versus 19 and 21 respectively. This is in sharp contrast with Pagano et al. (1998), who find a median IPO age of 26 years. The median Italian IPO in our sample is 19 years, which is closer to the figure Pagano et al. find. The oldest company in our sample to go public reached the respectable age of 111 in the year of its issue.
the relevant IPO, and are therefore removed from the sample. The listed company for which Amadeus does not provide an IPO date is added.
Level of competition. As can be expected, we find much more unlisted peer firms than
listed ones. One IPO has over 1,700 unlisted firms that match its asset size and are in the same industry, which is probably due to its general industry classification (NACE code 3299: ‘Other manufacturing’). 47 out of our 94 IPOs have between one and 25 unlisted peers. There are fourteen companies with more than 100 matching unlisted firms. For 67 IPOs we do not find listed firms that comply to our rigorous selection criteria. As argued before, we maintain our uniform selection criteria and do not loosen these constraints.
Size. As can be expected, the median market cap of our newly listed firms is lower
than for the previously listed. The size difference is not reflected in the size match variable: the median listed firm had assets over the two years prior to the corresponding IPO that were 8 percent lower than the IPO’s. Again in line with expectations, the median unlisted benchmark firm is smaller than the IPO sample. The unlisted benchmark does show higher capital intensity: 1.3 versus 1.0 for the newly listed companies. The figures for the IPOs compared with the listed firms show that both sales and assets are lower for our soon-to-be listed firms versus the already-listed ones. Their capital intensity is comparable. As a final measure of size we look at the number of employees. The descriptive statistics show that the median unlisted firm has almost half of the number of employees of its corresponding IPO. Interestingly, the listed firms have almost three times as many employees as the IPOs in our sample.
Profitability. The profitability measures we use are return on assets (ROA) and return
on sales (ROS), measured as earnings before interest, depreciation and taxes (EBITDA) divided by total assets or sales of the previous year, respectively. In the pre-IPO year, both ROA and ROS are higher for the IPO firms than for the unlisted benchmark, which points to the earnings management theory of Teoh et al. (1998). However, Chemmanur et al. (2008) argue that a longer pre-IPO period should be studied to be conclusive on the presence of earnings management. Compared with listed firms the difference is less pronounced, but again ROA and ROS are higher for the IPOs.
Leverage. In the year before the offering, the median IPO has comparable leverage to
comparable unlisted firms. It is interesting to see that the median firm in our sample has much higher leverage than reported by Rajan & Zingales (1995), who investigate leverage levels among G7 countries. They find median debt-to-assets values between 0.16 and 0.27 for the European countries. Arcas & Bachiller (2008) do find leverage ratios of over 0.6 for their sample of private and privatized firms. The leverage of listed firms is much lower: 34 percent. The reason is probably their higher availability of financing alternatives for debt.
(1994) among others. As can be expected, the median cash growth for our newly listed firms in the book year of their equity issue is very high: 178 percent. Furthermore, leverage declines hugely with 11 percentage point in that year. The subsequent year shows a large decline in cash levels, as would be expected if companies use the proceeds to invest in projects, or to deleverage.
Table 2 about here
To summarize, the IPOs in our sample are younger and more profitable than both firms that stay private and firms that were already listed. Furthermore, both IPOs and firms that stay private have higher leverage ratios than listed firms. IPOs show high median growth figures in all years around the initial offering. The capital efficiency of listed firms – both newly and previously listed – is much lower than for firms that stay private.
4.3 Methods applied
We are interested in studying two main subjects: the relative performance of IPO firms, and the performance of IPOs relative to their level of competition. In line with the extant literature we use return on assets as the main measure for firm performance. However, we apply the following formula:
1 1 − −
×
=
t t t t t tAssets
Sales
Sales
EBITDA
Assets
EBITDA
, or ROAt = ROSt×CAPEFFt,
Where EBITDA is earnings before interest, taxes, debt and amortization, ROA is return on assets, ROS is return on sales, CAPEFF is capital efficiency (sales on assets), and t is the year for which the variable is measured. Return on assets is a measure for the returns providers of funds can expect for their investment. Return on sales is a measure of marginal profits companies attain. Finally, sales over assets are not a measure of profitability, but of the efficiency companies attain. By using the sub measures of return on sales and capital efficiency, we are able to delve deeper into the return on assets performance of the companies.
firms; firms with nineteen or more competitors are called high-competition firms, and those with less than nineteen competitors are low-competition firms. We apply this distinction to find if there are material differences in the performance high- and low-competition firms. A final step in our study is estimating regressions to investigate which variables can explain differences in long-term behavior of IPOs, firms that stay private, and firms that were already listed. As argued before, our data are skewed and contain large outliers, therefore we use median regressions. We use the following regression equation on the sample of IPOs combined with the sample of unlisted firms, and on the sample of IPOs compare with the sample of previously listed firms:
it t i n j itj j itj n j j
it
IPO
CVAR
CVAR
ANGLO
YR
y
=
α
+
β
⋅
+
∑
γ
⋅
+
∑
θ
⋅
+
η
⋅
+
η
+
ε
= = + 1 2 1 1 1 2 ,Where yit+2 is the performance of company i in year t+2, and CVARitj is the value of control variable j for company i in year t. The abbreviations, descriptions, and formulae for the performance measures and control variables are shown in Table 3.
Table 3 about here
Competition. Due to the way our benchmark is constructed, we have a view on the
level of competition companies deal with. We include in our sample all companies that are highly similar to our peers. The number of matching firms we find for an IPO is our measure for its level of competition: the more similar companies we find for an IPO, the higher the level of competition for that company. The control firms have the same level of competition as their corresponding IPO, with the rationale that for a control firm the IPO plus the rest of the control firms for that IPO are its competitors. Based on the theories of Povel & Raith (2004) and Schoubben & Van Hulle (2009), we expect that IPOs and firms that stay private show a negative relation between the level of competition and their profitability and efficiency. Contrastingly, we expect that listed firms feel less financially constrained, and therefore profit from the IPO. The benefit may even be higher if the firm is facing many competitors.
Change in leverage. A reduction in leverage in the IPO-year reflects the extent to
theory holds we expect to find a negative relation between the change in leverage in year 0 and company performance in year +2: deleveraging is positive for company performance.17
Change in cash. It can be expected that the level of cash grows significantly in the
year of the IPO, due to the inflow of funds from the stock issue. If the IPO is done to finance further growth and investments of a company, one would expect a positive relation between the level of cash growth and profitability. However, as literature on initial underpricing has shown, the cash proceeds from the issue can be much higher than anticipated. From an agency cost theory perspective, this would enable overly enthusiastic managers to invest in negative NPV projects, resulting in lower profitability. Another option is that nothing is done with the cash generated: it is not invested, nor is it returned to the shareholder or used for debt relief. This is simply another form of agency cost. In both cases, one would expect a negative relation between the level of cash and long-term profitability.
Size. Baumol (1959) argues that large firms have higher profitability, because they
have all the options of smaller firms, and in addition can invest in very capital intensive projects. Hall & Weiss (1967) find evidence supporting this theory. Therefore, a positive relation between size and profitability can be expected. However, Porter (1985) argues that small firms can enter niche markets and attain high profitability as well. Thus, the sign of the relation is not per se positive.
Anglo-Saxon countries. We further introduce a dummy that takes the value one if a
company is from an Anglo-Saxon country (Great Britain or Ireland). La Porta et al. (1998) show that there are fundamental differences between companies in common law countries and civil law countries. Furthermore, public markets are more highly developed in Anglo-Saxon countries. Due to the higher level of development of these markets, they are expected to be more efficient. Due to this lower level of information asymmetries we expect Anglo-Saxon listed firms to show higher performance than continental-European firms. Furthermore, due to the higher efficiency of their public markets, any discrepancies between the performance of Anglo-Saxon IPO firms and long-listed firms will disappear sooner. Thus, if any general lower performance can be found for our IPO firms compared with unlisted or previously listed firms, we expect this to be lower for Anglo-Saxon IPOs than for IPOs from continental Europe.
Year of the IPO. Finally, we use dummies for the year in which the IPO took place.
As stated before, changes in the accounting standards for all European listed firms could lead to changes in reported profitability. Beforehand we do not have expectations on the sign of the difference, but any differences as a result of the introduction of IFRS would probably be found around 2005, the year of the introduction of IFRS.
We do not introduce a control variable for the industry a firm operates in. Since our database is constructed such that all companies are in manufacturing industries, our sample is homogeneous to that extent. 18
There is a possible difference in measuring the returns and efficiencies of listed companies versus unlisted companies. Two types of bias may occur. First, the introduction of the international financial reporting standards (IFRS) in 2005 may result in differences among IPO-years. Second, there may be differences in accounting standards between listed and unlisted companies. The accounting standards for listed companies may be higher than those of unlisted companies. Indeed, Pagano et al. (1998) give stricter accounting standards as a possible explanation for their result of underperformance. To address the first measurement bias, we introduce year-of-IPO dummies. The second bias is harder to correct. Pagano et al. (1998) argue that smaller companies should have more trouble overcoming the costs of the higher accounting standards, but other mechanisms could also explain a positive relation between size and probability of an IPO.
5. Results & Analysis
5.1 IPOs and firms that stay private
First, we look at the performance of IPOs in relation to firms that stay private. Table 4 shows the medians of the performance variables from the year before the IPO to the third year after the IPO. The results show that in the pre-IPO year, the median capital efficiency of the firms that stay private is significantly higher than that of the IPO sample. Contrastingly, return on sales and return on assets of the issuing firms in that year are higher, although ROA not significantly so and ROS only at the ten percent level. Its relatively low capital efficiency in combination with the fact that the IPO goes public in the next year is evidence that the median IPO feels financially constrained: it cannot make the investments necessary to keep up with competition. This is not coincidentally noted in the literature as one of the main reasons to go public. The stock issue should render cash that, if properly invested, increases profitability and capital efficiency in the long run.
Indeed we find a significant median increase of capital efficiency of eleven percent for the IPO sample, compared with three percent for the firms that stay private.19 However, the large difference in efficiency is persistent. Either the financial constraints are still present for issuing firms, or another mechanism is at work. The higher ROS is also persistent for the newly listed firms; ROS for our IPOs do not change significantly over year -1 through year +2, while it declines with almost one percent for the firms that stay private. Interestingly, both IPO and unlisted firm ROA decline significantly, although only at the ten percent level for the IPOs. In combination with the increasing capital efficiency for both firm types, this points to the use of windows of opportunity by the IPO firms: in the years following the offering they see declining profits, which puts pressure on efficiency, i.e. they have to work more efficient. Apparently, IPOs go public at a time when the economy enables them to earn high profits. Finally, year +2 shows a lower ROA for the newly listed firms, but the difference is not significant. Thus, we cannot reconfirm lower performance of IPOs relative to unlisted firms as is found in Pagano et al. (1998).
Table 4 about here
We now take a closer look at firms facing high- and low-competition. As discussed in Section 2.4, financially constrained firms suffer from high competition, while non-constrained firms perform better when they face more competition. This theory is empirically confirmed by Schoubben & Van Hulle (2009). They argue that unlisted firms are more likely to be financially constrained, and find that these firms show a negative relation between competition and performance, while listed firms show the opposite relation.
Table 5 shows the results for the comparison of IPO firms facing high and low competition. We do not find strong results. It appears that the low-competition IPOs perform somewhat better over all years considered, but the differences are not significant. The only significant variable the results show is that ROA declines significantly for low-competition firms. This is somewhat in line with our expectation that high-competition IPOs show improving performance ex post relative to low-competition IPOs due to relaxation of financing constraints.
Stronger results are found for the sample of firms that stay private. Table 6 gives the results for the unlisted high- and low-competition firms. Again, there are no significant differences in return on sales. However, ROA is consistently higher for high-competition
firms (at ten percent significance), and capital efficiency is consistently higher for high-competition firms (one percent significant). Capital efficiency does increase for unlisted firms facing high competition, but the difference with low-competition firms persists over the years. This is in sharp contrast with the results of Schoubben and Van Hulle (2009).20 Finally, the declining profitability for unlisted firms found in the comparison with IPOs appears to be consistent for both the low- and high-competition category.
Table 5 about here
Table 6 about here
Next, we introduce two more comparisons in Table 7: high-competition IPOs versus high-competition unlisted firms, and low-competition IPOs and unlisted firms. In terms of capital efficiency, high-competition unlisted firms outperform our IPOs from year 0 to year +2, consistent with the results of our general comparison of IPO with unlisted firm in Table 4. Furthermore, although IPO profitability appears higher in the pre-offering year, it is not significantly so. In fact, year +2 shows slightly significantly higher ROA for the unlisted sample. Turning to the results on low competition in Panel B, we see that for most years, low-competition IPOs outperform firms that stay private in terms of profitability. Their ROA does seem to decline faster than for unlisted firms: 2.5 percentage point over three years, compared with 1 percentage point for their unlisted competitors. In contrast with the results on high-competition firms, we do not find significant differences for capital efficiency of low-competition IPOs and unlisted firms.
Table 7 about here
5.2 IPOs and previously listed firms
We now take a closer look at the performance of IPOs in relation to previously listed firms. As stated earlier, much literature finds long-run underperformance of IPOs versus
listed firms, both in terms of stock prices (e.g. Ritter, 1991) and profitability (e.g. Mikkelson et al., 1997). Table 8 shows the main results for the comparison. There do not seem to be large differences between IPOs and longer-listed firms, apart from the third year after the IPO. Here, we find that both ROA and ROS for the listed firms are significantly higher. Furthermore, over the years around the offering, the median IPO ROA shows a decline of 2.5 percentage points, while the ROA of already listed firms shows an increase of 2.5 percentage points. This difference is significant at the one percent level. Thus, we find evidence of long-run lower performance of IPOs compared with previously listed firms when looking at profitability, in line with Mikkelson et al. (1997). Capital efficiency declines for the listed firms, while it grows for the IPOs, however, apart from year -1, there are no significant differences between the two types of company for the efficiency variable.
As in the previous section, we are also interested in the differences between high- and low-competition firms. Theory predicts that listed firms show a positive relation between level of competition and performance. However, Table 9 does not give strong support for this prediction. Although ROA and ROS appear consistently much higher for the long-listed firms facing high-competition, this is only significant for ROS in year +1, and only at ten percent significance. There are no significant differences between the capital efficiency of high- and low-competition previously listed firms, although the high-competition firms show a significant decline in this variable.
Table 8 about here
Table 9 about here
5.3 Regression results
In this section we present test results on the long-term performance of our sample of European IPOs in relation to comparable firms that stay private or were already listed. First, we compare the performance of IPOs and unlisted firms. Table 10 shows the results for the median regressions estimated on the IPO sample. First, we find no evidence of lower performance of IPOs in general relative to comparable firms that stay private.21 Second, in line with our results from Section 5.1, we find that the level of competition has a positive
relation with CAPEFF for unlisted firms. However, this relation is negative for IPOs. Third, for unlisted firms, change in leverage squared has a significant negative relation with ROA and ROS. Apparently, when unlisted firms highly increase or decrease their leverage, this is detrimental to their long-term profitability. In line with expectations, even stronger results are found for IPOs: we find a large negative relation between linear change in leverage in the IPO-year and ROA. When IPOs decrease their leverage after the offering, they improve their long-term performance. Change in leverage also has a negative relation with CAPEFF. However, we also find a negative relation between the squared change in leverage and ROA. This means that if deleverage is very large for IPOs, this is detrimental to their long-run performance.22 Fourth, the results show a positive relation between the quadratic change in cash ratio in year 0, and ROS two years later for firms that stay private. This is in line with their relatively high leverage: if unlisted companies are somehow able to increase their cash levels, they invest it wisely and increase their long-run ROS. The change in cash ratio in the year of their offering has a negative effect on IPO firm long-term ROS and CAPEFF. There is a very large nonlinear negative effect for ROA and ROS, implying that when companies raise too much cash from their offering this is detrimental to long-term profitability. Finally, size does not seem to be a very important determinant of IPO and unlisted firm performance: we only find a small negative relation with ROS.
Next, Table 11 shows the results of median regressions on the comparison between long-term performance of IPO and previously listed firms. Again, we find no evidence of general long-term lower performance of IPOs relative to already listed firms. Second, in contrast with the results on unlisted firms and IPOs, longer listed firms show a positive relation between the change in leverage in year 0 and ROA in year +2, one percent significant for both the linear and quadratic coefficients. The positive relation can also be found for the linear variable and long-run ROS, although only at ten percent significance. For the IPOs we again find a negative relation between change in leverage and long-term ROA. We also find a negative relation between squared change in leverage and long-term ROA. Third, a highly negative relation is found between squared change in cash and CAPEFF of listed firms. If they largely increase their level of cash, this is detrimental to listed firms’ efficiency. Finally, a slightly positive relation is found between quadratic change in cash levels in year 0 and long-term CAPEFF.
Table 10 about here
Table 11 about here
6. Conclusion
The literature on long-run performance of IPOs generally finds evidence of underperformance. Although most literature uses U.S. based companies, some authors show that underperformance is a worldwide phenomenon (e.g. Levis, 1993; and Cai et al., 2008). Most authors compare the performance of the IPOs with that of listed peer firms. Moreover, performance is usually measured by looking at stock returns. We study performance, measured primarily as return on assets, of European manufacturing IPOs during 2001 – 2006 in relation to both comparable unlisted and previously listed companies.
We use a unique and rigorous selection method for our peer firms. Our method is comparable to that found in Ritter (1991) and Loughran & Ritter (1995), but we allow for multiple peers per IPO, and we do not loosen our selection criteria. This means we get a view on the level of competition IPOs deal with, as well as ensure that our peers are a very close match to our IPOs. We distinguish between firms facing high and low competition. Furthermore, we disentangle return on assets into return on sales and capital efficiency. This allows us to go deeper into the reasons for possible lower performance of IPOs compared with unlisted and listed firms.
We find that in the year before the stock issue, IPOs show lower capital efficiency than comparable unlisted firms. Possibly, IPO firms are more financially constrained, and therefore cannot make the investments needed to keep up with competition. The public equity offering should reduce these constraints, resulting in higher IPO performance. However, the difference with unlisted firms persists over the years following the issue. The difference is highest for IPO firms facing high competition.
Furthermore, we find a positive relation between high competition and capital efficiency of unlisted firms, but a negative relation for capital efficiency of IPO firms. This is contrary to the results of Schoubben & Van Hulle (2009), who find a negative relation between competition and performance for unlisted firms.
7. References
Aggarwal, Reena, and Pietra Rivoli, 1990, Fads in the initial public offering market?,
Financial Management 19(4), 45-57.
Aggarwal, Reena, Ricardo Leal, and Leonardo Hernandez, 1993, The aftermarket performance of initial public offerings in Latin America, Financial Management 22(1), 42-53.
Álvarez, Susana, and Víctor M. González, 2005, Performance of Spanish firms going public: windows of opportunity and the informative effect, Applied Financial Economics 15(18), 1283-1297.
Arcas, María J., and Patricia Bachiller, 2008, Performance and capital structure of privatized firms in Europe, Global Economic Review 37(1), 107-123.
Bancel, Franck, and Usha R. Mittoo, 2009, Why do European firms go public?, European
Financial Management 15(4), 844-884.
Barber, Brad M., and John D. Lyon, 1997, Detecting long-run abnormal stock returns: the empirical power and specification of test statistics, Journal of Financial Economics 43(3), 341-372.
Baumol, William J., 1959, Business Behavior, value, and growth, MacMillan, New York, USA.
Boulton, Thomas J., Scott B. Smart, and Chad J. Zutter, 2010, IPO underpricing and international corporate governance, Journal of International Business Studies 41(2), 206-222.
Brau, James C., and Stanley E. Fawcett, 2006, Initial public offerings: an analysis of theory and practice, Journal of Finance 61(1), 399-436.
Brav, Alon, and Paul A. Gompers, 1997, Myth or reality? The long-run underperformance of initial public offerings: evidence from venture and non venture-backed companies, Journal of
Cai, Xiaoqiong, Guy S. Liu, and Bryan Mase, 2008, The long-run performance of initial public offerings and its determinants: the case of China, Review of Quantitative Finance &
Accounting 30(4), 419-432.
Carter, Richard B., Frederick H. Dark, and Ajai K. Singh, 1998, Underwriter reputation, initial returns, and the long-run performance of IPO stocks, Journal of Finance 53(1), 285-311.
Certo, S. Trevis, Tim R. Holcomb, and R. Michael Holmes Jr., 2009, IPO research in management and entrepreneurship: moving the agenda forwards, Journal of Management 35(6), 1340-1378.
Chen, Hsuan-Chi, Keng-Yu Ho, Yu-Jen Hsiao, and Chen-Huan Wu, 2010, The diversification effects of initial public offerings, Journal of Business Finance & Accounting 37(1) & (2), 171-205.
Chemmanur, Thomas J., and Paolo Fulghieri, 1999, A theory of the going-public decision,
Review of Financial Studies 12(2), 249-279.
Chemmanur, Thomas J., and Jie He, 2009, IPO waves, product market competition, and the going-public decision: theory and evidence, Working paper, revision March 15, 2010, available at SSRN: http://ssrn.com/abstract=1270985.
Chemmanur, Thomas J., Shan He, and Debarshi Nandy, 2008, The going public decision and the product market, Working paper, revision August 25, 2009, available at SSRN: http://ssrn.com/abstract=1440930.
Dahlquist, Magnus, and Frank de Jong, 2008, Pseudo market timing: a reappraisal, Journal of
Financial & Quantitative Analysis 43(3), 547-580.
Eckbo, B. Espen, and Øyvind Norli, 2005, Liquidity risk, leverage, and long-run IPO returns,
Journal of Corporate Finance 11(1)&(2), 1-35.
Hall, Marshall, and Leonard Weiss, 1967, Firm size and profitability, Review of Economics &
Statistics 49(3), 319-331.
Heaton, J.B., 2002, Managerial optimism and corporate finance, Financial Management 31(2), 33-45.
Helwege, Jean, and Frank Packer, 2009, Private matters, Journal of Financial Intermediation 18(3), 362-383.
Hou, Kewei, and David T. Robinson, 2006, Industry concentration and average stock returns,
Journal of Finance 61(4), 1927-1956
Ibbotson, Roger G., Jody L. Sindelar, and Jay R. Ritter, 1994, The market’s problems with the pricing of initial public offerings, Journal of Applied Corporate Finance 7(1), 66-74.
Jain, Bharat A., and Omesh Kini, 1994, The post-issue operating performance of IPO firms,
Journal of Finance 49(5), 1699-1726.
Jain, Bharat A., and Omesh Kini, 1999, The lifecycle of initial public offering firms, Journal
of Business Finance & Accounting 26(9)&(10), 1281-1307.
Januszewski, Silke I., Jens Köke, and Joachim K. Winter, 2002, Product market competition, corporate governance and firm performance: an empirical analysis for Germany, Research in
Economics 56(3), 299-332.
Jensen, Michael C., and William H. Meckling, 1976, Theory of the firm, managerial behavior, agency costs and ownership structure, Journal of Financial Economics 3(4), 305-360.
Karuna, Christo, 2007, Industry product market competition and managerial incentives,
Journal of Accounting and Economics 43(2)&(3), 275-297.
Keloharju, Matti, 1993, The winner’s curse, legal liability, and the long-run price performance of initial public offerings in Finland, Journal of Financial Economics 34(2), 251-277.
