The Current State of LBOs

(1)

1

Bachelor Thesis Economics and Business

Faculty of Economics and Business

Academic year: 2019 – 2020

The Current State of LBOs

Specialization: Finance

Student Name: Duricu Vlad Marco George

Student Number: 11838418

(2)

2

Statement of Originality

This document is written by Student Vlad Marco George Duricu who declares to take full

responsibility for this document's contents. I declare that the text and the work presented in

this document are original and that no sources other than those mentioned in the text and its

references have been used in creating it. The Faculty of Economics and Business is

responsible solely for the supervision of completion of the work, not for the contents.

Abstract

This paper investigates if trends in the private equity industry and leveraged buyouts established by Stromberg (2007) prior to financial crisis still hold for the last decade. The focus of the research is on three categories sub-deals, regions and industries. It is also investigates what are the common characteristics of an attractive target for a leveraged buyout. The outcome regarding the common trends are different compared the period before the financial crisis. The main reason could be the change in the investor’s behavior (mainly PE). On the other hand, the characteristics of an attractive target for an LBO are in line with the literature. This illustrates that the behavior of the investors only changed in terms of non-financial factors.

(3)

3

Private equity is one of the most popular alternative asset classes for investors. Cambridge Associates (2019), an investment firm located in the United States, illustrates compelling research about the returns of Private Equity Index compared to the returns of the well-known equity indexes such as S&P 500, Russell 2000, Russell 3000, MSCI World Index, MSCI Europe Index, and MSCI EAFE Index. The research reveals that at the end of 2019, on different timespans ranging from 3, 5, 10, 15, 20, 25 years, the Private Equity Index has higher returns than any other equity index presented above. It is essential to mention that gains from all funds are computed net of fees. The equities overperformed the PE index only on a one-year basis in 2019.

The discussion about asset allocation is a relevant topic for investors all over the world, considering the impact of unexpected events, such as the coronavirus pandemic. Alternative asset classes such as Private Equity funds gain higher importance in the asset allocation topic. One apparent reason is the past return success of outperforming the stock markets on a global level. Another reason is related to the increased risk of the stock markets compared to the volatility of a PE fund. Furthermore, the practitioners may also wonder if the industry can find positive return investments due to the high number of competitors and a volatile world economic environment.

From general partners (the professionals in charge of a PE fund) point of view, the increased uncertainty is also an important topic on the agenda. The issues of the GPs are more related to the investor's trust, the fundraising, and, nevertheless, the process of finding positive alpha returns. A topic almost attached to the private equity world is leveraged buyout transactions. As a standalone subject, LBOs transactions gained more considerable influence. Probably, one factor that leads to increase notoriety for this topic is the large transactions where companies listed on stock markets were taken private. The leverage became a vital tool over time because it enables investors to perform a higher number of transactions and offers benefits to firms such as an increased value due to the tax shield effect. However, leverage is also a risky tool that can drive a company bankrupt, especially in uncertain periods.

The consulting firm Bain & Company (2020), which is highly specialized in PE

transactions, issues a report once a year about the state of the PE market. In the latest report, at the

beginning of 2020, Bain & Companyshows that the global deal value of buyout transactions has

been increasing since 2011. However, it has been bouncing from 2015 until 2019, with the peaked reached in 2015 and 2018 (approximately $608 billion). In 2019, the global buyout transactions

(5)

5

market dropped to $557 billion, with a total number of deals slightly below 3000. A noticeable aspect is that the global buyouts market was never able to recover to the levels seen before the global financial crisis in 2008 and 2009. The absolute top was registered in 2006 with $804 billion and more than 4000 deals.

Understanding the importance of PE market and global buyout transactions, this paper

investigates the state of the leveraged buyout transaction deals. Stromberg (2007) conducted a study about the demography of LBOs on a period of 38 years from 1970 to 2007, analyzing more than 19000 deals. An essential insight is that 63% of the transactions were performed between 2001 and 2007, and the value of the transactions corresponds to 68% of the total value of the sample, approximately $2.7 trillion measured by the enterprise value of the acquired firms. Therefore, this shows the hike in popularity for leveraged buyout deals. The transactions were divided into five different categories of sub-deals as Public to Private, Private to Private, Divisional buyouts, Financial vendors, and Financial distress.

Furthermore, the deals are also categorized into ten different regions. For instance, regions with the highest number of completed deals are the United States, Continental Europe, and Scandinavia. On the opposite side, regions with the lowest number of completed deals are the Middle East and Africa, Latin America, and Eastern Europe. Moreover, the author also divided the buyouts based on the industries in order to test the hypothesis that most of the transactions are performed in well-established and mature industries where the firms already reached the level of

having steady cash flows.

The goal of the research is to test if Stromberg's findings, especially from 2001 to 2007, are still representative of leveraged buyouts nowadays, and if there are any significant changes in the demography of the deals. The investigated categories are the sub-deals, regions and industries of the transactions. The Stromberg's study will not be replicated entirely considering that the author also tested, using multivariate regression, the characteristics of the targets that end up having a successful exit at the end of the buyout period. On the other hand, the new addition that this research tries to bring on the table compared to the Stromberg's article is related to the characteristics that can make a company, a potential target for a PE fund or a leveraged buyout. Therefore, this thesis will follow to find the standard features of the firms targeted by potential buyers.

(6)

6

This paper tests several hypotheses. The first hypothesis is related to the total number of deals per sub-category. Although the number of 2000 investigated transactions differs from the 19000 deals interpreted by Stromberg, this paper expects a similar distribution of deals sub-types that were completed in 2001-2007. Thus, 6.8% of the deals are expected to be public takeovers, 36.9% private to private, 16.8% secondary buyouts, and 3.2% financial distressed. The 36.3% difference up to 100% is coming from the lack of operationalization of the Divisional buyout variable. It is expected that 36.3% will be divided between Private to private and secondary buyouts transactions. The second hypothesis that will be further checked is the ranking of buyouts per region. The United States, Continental Europe, Scandinavia, and the United Kingdom are supposed to be the top countries by a total number of deals and deals valuation. On the contrary, Africa and the Middle East, Latin America, and Eastern Europe are regions expected to be on the lower part of this table.

The third hypothesis investigated is the well-known idea that most of the targets that are acquired through a leveraged buyout are part of the old, well-established, and mature industries. Furthermore, the targets are also categorized as mature companies, and their main characteristic is steady cash flow. Hence, the target companies of the completed transactions are expected to be part of industries such as manufacturing, retail, construction, or food and beverages. In contrast, the targets that are part of the technology industry are forecasted to be very few in this sample. Nevertheless, this paper pursues finding out the characteristics of a company that might be considered a potential target in a future LBO. The independent variables that will be used in the multivariate regression are related to the deal's sub-type, the regions and the industry of the firms. This thesis identifies the main characteristics of the LBO transactions using more than 2000

deals from Zephyr dataset on a period of nine years from 2011 to 2019.The deals are divided into

four different categories, such as Public Takeover, Private to Private, Secondary Buyouts, and Financial Distress. Public Takeover refers to all transactions where the target company is delisted from the stock exchange. Private to private are the deals where the private company is acquired through an LBO by the buyer. Secondary Buyouts is the category that includes all secondary, tertiary, and so forth buyouts. Nevertheless, Financial Distress contains all the firms that are in the process of financial restructuring and are acquired through a leveraged buyout. In addition to the initial division, the deals are also categorized per region and per industry. The regional categorization is similar to Stromberg's study; thus, the deals are split into ten distinct regions. The

(7)

7

standard industrial classification code is used to determine the industry of each transaction (USIC sector code).

In order to test the last hypothesis related to the features of the potential targets, this research uses a different sample of 1550 mergers and acquisitions and leveraged buyouts from Zephyr. The sample is equally distributed between M&As and LBOs. They are similar in terms of deals per region, period and industries. The only significant difference is in terms of deal values. The prices paid for M&As are higher compared to LBOs. The financial data for the targets is downloaded from Orbis database. They are merged though BVD id code, which is unique for each firm. Operating revenues from the last for years prior to the deals, tangible assets, total assets, equity, the amount of tax paid, the cash and equivalents on the balance sheet and the level of debt are the variables used in order to build the regression models.

The main results of this paper are different compared to the demography of LBOs prior to the financial crisis. There are more private to private deals and secondary buyouts as a proportion compared to Stromberg’s findings. Furthermore, the value of the deals is also higher compared to 2001 and 2007 period. The public takeovers are similar in terms of number and in terms of value. The financial distress deals are too few and thus, they are negligible in this analysis. The regions are similar as popularity with the highest number of deals in the US and Continental Europe but with a consistent difference in terms of transactions value. The average deal value in the US is higher compared to 2001-2007, while less money on average are spent in Continental Europe in comparison with the same period.

The old idea regarding the mature industry does not hold for this sample. The outcome is in accordance with the main findings of the anchor paper. As a value, the transactions related to services are higher than deals from manufacturing industry. Nevertheless, two regression models (logistic regression and probit) are used to test the characteristics of the most attractive targets. They are in accordance to the outcomes from the literature. There is more explanation about the models in the results and discussion part and, and the outcomes of the logit and probit can be found in table 10 in the appendix.

This paper is organized into five parts. The first part is the introduction, where the research question and the hypotheses were introduced. A summary of the common findings and consensus in academia about leveraged buyouts and private equity and the definition of several highly technical terms (some of them already used in the introduction) can be found in the second section.

(8)

8

The third component of the thesis consists of a description of the methodology used in order to perform the research and the empirical models used to perform the statistical tests. In the fourth part of the paper, the main results of this research will be explained, and the limitations will be highlighted. Further research recommendations related to leveraged buyout and PE topics will also be an element of this category. Nevertheless, the last component is represented by the appendix where tables, charts, and multivariate regression about the outcomes of the paper can be found and checked by the readers.

2. Literature review

This section starts by defining several critical elements related to the topics discussed in this paper. The literature review is organized into five different sections. The first two parts define the private equity funds and leveraged buyout terms broadly. Section three explains the economic role of a leveraged buyout. The fourth part relates to the value creation through LBOs and contains several sub-categories where different possible ways for value creation are discussed. Nevertheless, the last section illustrates the state of the private equity industry and LBOs before the financial crisis and what is the literature consensus related to the targets in a leveraged buyout.

2.1 Private Equity Funds

Kaplan and Stromberg (2009) explain that private equity firms utilize private equity funds in order to raise capital from investors. The funds raised from investors are used by the firms to invest in private companies and to cover the yearly expenses of the fund. From a legal perspective, the funds are managed by general partners (GPs) that commit a slight amount of money, usually around 1% of the total funds. In contrast, the limited partners, represented in this case by investors, are the parties that allocate the majority of the capital available for the funds. Institutional investors such as public pension funds, endowment funds, and even wealthy individuals are the common investors for a private equity fund.

The main characteristic of PE funds is the restricted life of the firm. Frequently, the life of a PE fund is, on average, ten years. Sometimes, the fund can increase the functional period by up to three years. Typically, during the first five years, the private equity funds invest in potential targets, having between five and eight years to return the profits to limited partners (initial investors).

(9)

9

Metrick and Yasuda (2010) study the rewarding structure of GPs; the authors identified three different means of compensation for general partners. The first way is illustrated by the management fees charged by the PEs. Regularly, the fee is a percentage of the total funds raised. Furthermore, when the investments are completed, the fee is replaced by another charge, which is equal with a fraction of the capital employed by the investment firm. Another means of compensation for GPs is represented by a share of profits made by the fund. For instance, most of the private equity firms charge a 20% share of the profits. Nevertheless, several funds also have a monitoring fee. Thus, the businesses in the PE portfolio, together with the current clients, must pay the GPs for monitoring.

2.2 Leveraged Buyout

A leveraged buyout is defined as a transaction where the buyer purchases the target using a substantial amount of debt. The acquired firm can be public or private. If the target company is a public company, usually it is taken private through an LBO deal. Bargeron et al. (2007) found out that the premium paid for buying a company that is listed on a stock exchange is highly volatile between 15% and 50% of the share price.

Frequently, the amount of debt used in LBOs varies between 50% and 90%. The acquirer is usually a PE fund, or a consortium of funds lead by a sponsor. A new firm is created, where all the assets of the acquired firm and the new issue debt used in the acquisition process are part of the balance sheet of the new entity. Thus, the debt is repaid using the cash flows from the target firm. Usually, the PE funds exit the transaction selling the firms in the portfolio to a strategic acquirer, to other financial sponsors, or by listing the firm at the stock exchange. (Eckbo & Thorburn 2012).

2.3 Economic Role of Leveraged Buyout

Two well-known theories are found in the existing literature related to the economic impact of a leveraged buyout transaction. Jensen (1989), argues that leveraged buyouts would become the main form of organization for the firms due to the remarkable governance. Therefore, the firms organized as an LBO are easily verified by the GPs in charge of the PE funds. Furthermore, a significant amount of debt also has a role in disciplining the managers of the firms that used a

(10)

10

leveraged buyout structure. Jensen takes a step further and even predicts that in the future, a vast majority of the firms would adopt an LBO organizational form due to the benefits presented above. On the other hand, Rappaport (1990) argues that a leveraged buyout is a "shock" for the acquired company that allows the firm to get back on track from the performance perspective. The author states that firms struggling with corporate governance adopt an LBO organizational form. Hence, after the governance issues are corrected during the leveraged buyout period, in a few years, the company would decide to go back to the previous organizational form. The firms that have the best structure are considered the public companies listed on a stock exchange.

Kaplan (1991) showed that the average time of an LBO ownership is 6.8 years. In

conclusion, both theories illustrated above and supported by Jensen and Rapport are refuted. The common idea embraced by academia is that a leveraged buyout is used for a limited amount of time as a governance structure. It usually helps the company listed on the public markets to correct the governance structure and to pay back to investors the extra amount of available cash flows. After all the firm's governance issues are solved, the firms are ready to go back to the public markets.

On the other hand, Stromberg (2007) finds evidence that might bring the academical

perception of leveraged buyouts towards Jensen's theory from 1989. About 69% of all the firms investigating this study between 1970 and 2007 still have an LBO organizational form. The number of companies that were organized as an LBO in 2007 was 14000 compared to a maximum of 5000 in 2000 and less than 2000 in 1990. The author argues that the LBO organization is more related to the long term rather than temporary since approximately 40% of all leveraged buyouts remain in this form for more than ten years from the initial transaction. Compared to Kaplan's finding that the median LBO period is between 6-7 years in 1991, Stromberg identifies that the median LBO period between 1995-1999 before the transaction is exited is nine years, a significant change only in a few years.

2.4 Value Creation through Buyout Transactions

Kaplan and Stromberg (2009) consider three possible sets of adjustments that a PE firm can undertake to make the company in the portfolios more profitable: financial adjustments, governance changes, and operational engineering adjustments. One action completed by the private equity firms immediately after the transaction was to align management interest with the

(11)

11

firm's goals and targets. Therefore, the managers of the acquired firms, especially the CEO, received a new compensation structure based more on stock options, increasing the amount of equity that management-owned before the transaction. Hence, the personal targets of the directors become aligned with the firm's goal. It is essential to mention that this was a new practice in 1990. (Jensen and Murphy, 1990). Furthermore, Kaplan (1989b) discovered that the percentages owned by the managers in the firm increased by four times if the company was taken private. The private equity companies required the management team to buy new shares in the company. Thus, directors' stake in the firn is higher. Hence their incentives and effort should also be more significant than before.

In general, for leveraged buyouts, an important factor is the amount of debt of the new entity. The leverage has a role in increasing the pressure on the manager and in disciplining the management team. The firms must pay the interest and the principal, and there is no money left for directors to engage in behavior that can hurt the value of the company. Moreover, leverage is a valuable tool for companies that are part of the mature industries that do not have an immediate growth opportunity. (Jensen 1986). The ultimate benefit of the leverage is the growth in the firm's value through the tax shield. In many regions such as North America, Continental Europe, and the UK, the interest is tax-deductible; hence the company can boost the value through debt. On the other hand, if the amount of credit is very significant, it can deteriorate the financial health of the company and can push the firm in financial distress, which will ultimately have an effect on the company's value.

Governance engineering refers to board control over the business. The boards that

supervise companies that are part of PE fund's portfolio are usually smaller compared to other private and public companies and are controlled by the PE funds. These boards are more actively involved in the business, and they meet more frequently. (Gertner and Kaplan 1996). Furthermore, the board of these firms meet around twelve times per year and do not hesitate to fire the chief executive officers, about 33% of the CEOs are fired in less than 100 days after the acquisition. Besides, 66% of the top executives are released in a period of three to four years after the deal is completed. (Archarya and Kehoe 2008)

In order to increase the chance of raising the firm's value, a vast majority of PE funds hire external consultants that help the company design and implement different business and organizational plans to improve the performance and the structure of the target firm. Therefore,

(12)

12

several opportunities that can help the firm to upgrade the current situation are related to cutting costs, increasing revenues, strategic adjustments within the company, and, nevertheless, reshaping the firm's organizational structure.

2.4.1 Operating Performance

There is a significant amount of empirical evidence that illustrates how operating performance changes in a positive manner after a leveraged buyout, in general, for the newly acquired public companies. Three ratios are strong evidence in favor of increased performance for public to private deals: 1. 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝐼𝑛𝑐𝑜𝑚𝑒 𝑡𝑜 𝑠𝑎𝑙𝑒𝑠 𝑟𝑎𝑡𝑖𝑜 = 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝐼𝑛𝑐𝑜𝑚𝑒 𝑆𝑎𝑙𝑒𝑠 2. 𝐶𝑎𝑠ℎ 𝐹𝑙𝑜𝑤 𝑅𝑎𝑡𝑖𝑜 = 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝐼𝑛𝑐𝑜𝑚𝑒−𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐸𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒 𝑆𝑎𝑙𝑒𝑠 3. 𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐸𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒 𝑡𝑜 𝑆𝑎𝑙𝑒𝑠 𝑅𝑎𝑡𝑖𝑜 = 𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝐸𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒 𝑆𝑎𝑙𝑒𝑠

Kaplan (1989b) identifies that the operating income to sales ratio rises by 10% to 20% compared to the rest of the industry in absolute and relative terms. Cash Flow Ratio improves by almost 40% while Capital Expenditure to Sales Ratio decreases.

2.4.2Employment

Employment is an important topic for LBOs. Based on the study of several leveraged buyouts from the 1980s period in the US, Kaplan (1989b) found that companies that were acquired through an LBO transaction do not lay-off personnel, rather the growth in the total number of employees is lower than the rest of the industry. Moreover, similar results were observed by Lerner and Miranda (2008) during a more extensive period from 1980 to 2005. Besides, the authors found that also before the LBO deal, these firms had a lower increase in the number of employees compared to the industry average. In Europe, the outcomes illustrate that firms in an LBO organizational form have a similar increase in employees compared to the rest of the industry. However, the salary raises are smaller than the industry average. Kaplan and Stromberg (2009) conclude that

(13)

13

employment findings are in line with operating performance efficiency; firms can increase their value by operating more efficiently.

2.4.3 Asymmetric Information

The asymmetric information is not an essential reason for the operating changes. Kaplan (1989b) studied several LBOs from the 1980s and found that public companies that were acquired after the analysts were updating the forecasts were unable to reach the expectations. Furthermore, public firms that rejected an LBO offer were not able to outperform their peers even though the board approved the offer (Ofek, 1994). The only conclusion drawn by Stromberg and Kaplan (2009) was that private equity firms were able to reach better deals because of the negotiation skills or a lack of skills from the acquired party.

2.5 Demography of Transactions& Financial Features of an LBO

Considering the potential economic value of a leveraged buyout and the structural changes that take place from an operating perspective, but also management and overall employment angle, this paper believes that understanding specific trends of LBO transactions and private equity firms is very important. Stromberg (2009) finds that only 6.7% of all LBO transactions were public to private transactions. This number of deals represent around 28% of the total value. Contrary, the author identified that secondary buyouts and divisional buyouts were gaining popularity over time.

The growth in overall activity in other countries outside the US increased during the last period of the sample, with Continental Europe being the most important region. On the other hand, if North America and Western Europe are left apart, the rest of the world performs only 13% of LBO deals and 7% of the total value between 2001 and 2007 (Stromberg 2009). Furthermore, the total number and the value of buyouts in Continental Europe and the UK was steadily increasing between 1996-2005. (Wright et al. 2006)

Stromberg (2009) believes that not only firms from old and matures industries were involved in LBO transactions, and this idea was never true. Even though the author believes that companies that are part of industries such as chemicals, machinery, and retailing are still an attractive target for current LBOs, he also finds that firms from high-growth industries such as tech, biotech are becoming more popular during the last decade.

(14)

14

Le Nadant and Perdreau (2006) inspect the financial features of the LBO targets and test if they are different compared to a firm that did not go through the leveraged buyout process. The authors expect that a growing revenue influences negatively the chance of a firm to undergo through an LBO. On the other hand, the companies that pay a higher amount of taxes as a percentage of the revenues are expected to be an interesting target for a leveraged buyout because of the potential value creation through the tax shield. A low-level of capital intensity and leverage is representative for the potential targets. Capital intensity illustrates the investment level and is predicted to be lower for the LBOs targets because the investors want already a well-established firm in terms of the assets, not a firm that heavily invests prior to the LBO. The potential acquirers are attracted by low leverage firms because the target will be highly leveraged at the end of the leveraged buyout process. Nevertheless, the authors believe that firms with a higher level of cash on the balance sheet are an attractive target for the LBOs investors.

3. Methodology

This paper aims to establish if there are any significant changes in deal's trend compared to Stromberg's (2007) findings. As was already mentioned and supported by the literature, the level of LBO transactions is still below the financial crisis peak in 2006, considering the total number of transactions, and the total value. The methodology is organized into three different parts. The first section discusses data construction and data visualization. The second part illustrates the process used in order to impute the missing data. The third section presents the hypotheses again, whereas the last part illustrates the empirical models used for statistical tests.

3.1 Data Construction

The investigated sample contains 2071 leveraged buyouts from 2011 to 2020. The total value of all the transactions, measured as the price paid by the acquirer in order to purchase the target is about $1.53 trillion. Data was downloaded from the Zephyr database, a well-known database, especially for mergers and acquisitions. For this research, only deals that were confirmed as completed and the acquired stake, where known was above 50% threshold, were included in the sample. To have a set of variables similar to Stromberg's paper, the transactions were divided into four sub-deals categories. The four categories are Private to Private, Public Takeover, Secondary

(15)

15

Buyouts, and Financial Distress. If the words "Public Takeover" are part of the description the deals were considered public takeovers. If "secondary" or "tertiary buyout" were identified in deals' description sections, those transactions fell under the secondary buyouts category. Financial distress deals had the "financial distress" description. Nevertheless, the remaining buyouts were categorized as private to private transactions.

The second sample used to test the hypothesis related to the features of the potential target contains 1550 M&As and LBOs. The sample is equally distributed between M&As and LBOs and the period, regions an industry are also similar for the transactions. The only significant difference is related to the value of the transactions, M&As are more valuable in terms of final price paid than LBOs. In order to understand the differences between these two different types of deals, the financial data are downloaded from Orbis and the datasets are merged through the BVD id code which is unique for every firm in the sample. The financial data was selected in accordance with the existing literature and the main variables used in the model were operating revenue, total and tangible assets, cash and cash equivalents, taxes, leverage and shareholders’ equity (Le Nadant and Perdreau, 2006).

3.2 Data Visualization

Private to private deals illustrate all leveraged buyout transactions where the target was a private company. This sample contains 1309 private to private deals with a total value of $651 billion (%42.40 of the total value of the sample), 63.29% of the total number of deals. Secondary buyouts are leveraged buyouts where the target was already in an LBO organizational form. Secondary, tertiary, and so forth, buyouts fall under this category. 629 transactions are part of this classification, which represents around 30.37% of the number of deals from this sample. The total amount of money spent by acquirers on secondary buyouts is close to $471.5 billion (30.69% of the total sample's value).

Public takeovers are transactions where the target company was a public firm, listed on a

stock exchange, and following the acquisition, the firm was private. 127 deals fall under this category, representing 6.13% of the total number of buyouts from this sample. The total transaction value is about $396.6 billion and illustrates 25.75% of the total sample's value. Finally, only six transactions (0.29%) out of 2071 are part of the financial distress classification. Their total value is close to $17.8 billion. The transactions and deal's value can be observed in table 1 and table 2.

(16)

16

Total number of deals per sub-category Total Value of deals per sub-category

(Table 1) (Table 2)

The distribution per region can be observed in table 3 and 4. The United States contains almost half of the sample's deals (909), which have a total value of close to $975 billion (two-thirds of the total value). As expected, Continental Europe and the United Kingdom are the second and the third in terms of deal's number and value. On the contrary, Middle East& Africa, Eastern Europe, and Latin America have the lowest number of deals. The first two regions also register the smallest value, while Latin America has a higher than expected transaction value.

Total Number of Deals per Region Total Value of Deals per Region

(17)

17

The industry of the target firms is another important topic that this paper investigates. The SIC codes are used in order to identify the industry and the sub-industries. A notable mention is that out of 2071 deals, 102 transactions did not have the industry code available. These buyouts were omitted entirely from the industry analysis. SIC codes methodology recognized ten classes of industries. Deals with codes from 01 to 09 are part of the Agriculture, Forestry, and Fishing industry, whereas transactions with codes from 10 to 14 fall under the Mining industry.

On the other hand, buyouts with a code between 15-17 are part of the Construction segment while the codes between 20 and 39 stand for the Manufacturing industry. Transportation, Communication, Electric Gas, and Sanitary Services are all represented by codes between 40 and 49, while 50-51 codes illustrate the Wholesale Trade segment. The number between 52 and 59 stands for Retail Trade, while 60 to 67 stands for Finance, Insurance, and Real Estate. Nevertheless, Services are represented by codes between 70 and 89. No Public Administration deal is part of this sample. The state of the industries can be observed in tables 6 and 7. The two largest industries as the total number and total value of transactions are Manufacturing and Services. The old theory about the attractiveness of mature industries for leveraged buyouts will be later discussed in the paper.

Total Number of Deals per Industry Total Value of Deals per Industry

(18)

18

3.3 Constructing Missing Deal Values (Heckman Regression)

One of the main issues of this research was the missing data for Deal Values, which represented the price paid by the acquirer to purchase the target. Out of 2071 buyouts, only 812 deals had deal value transactions available. Considering the continuation of the study, the deal value for all the transactions was necessary in order to be able to identify any significant shifting of the principal trends. Consulting the previous literature, which is very limited for the demography of deals, and analyzing the method used by Stromberg in his research, this paper decided to use a Heckman (1979) maximum likelihood estimation to impute the missing data. Stromberg (2009) used the same model to predict the missing value for enterprise values in his study.

There are four essential variables for Heckman regression. The sub-deal categories of transactions, the regions, the presence of the financial sponsor, and if the transaction was exited or not. Stromberg utilized almost similar variables in the constructed model. In addition to the presented variables, the author also included the type of fund in terms of private and public, the age of the fund with a cut-off at 20 years of function, and if a syndicate performed the transaction. Those variables were not available in the database; hence, they were not used to construct the model. In comparison with Stromberg's variables, the exit type was added in this paper. In the following section, the choice of each variable will be explained.

As can be observed in the regression table, the sub-deal categories have an essential role

in determining the price of a deal due to the different organization of the firms, ownership structure, and asymmetric information. It is intuitive to pay a higher price for public takeovers on average because the acquirer must deal with many shareholders that require a higher premium for selling their stake. On the other hand, a secondary buyout is cheaper because the fund wants to exit the transaction since the opportunity is considered lost. In contrast, other funds might see a profitable change. The financial distress variable is counterintuitive and is not significant in the regression since it has a p-value of 0.209. The explanation is the lack of buyouts that face financial distress in the current sample. All other variables from the sub-deals category are significant at p<0.05. The private buyouts were omitted because of the perfect collinearity issue.

A financial sponsor can influence the price in a significant manner. A financial sponsor such as a private equity fund brings on the table the knowledge and the negotiation skills, as observed in the literature, enhancing a better price for the acquirer. On the other hand, a successful exit represents a profitable transaction. Thus, the regression coefficient is positive, showing that

(19)

19

most of the exit transactions in the sample are negotiated for a higher price compared to a non-exit deal.

Nevertheless, the region plays a critical role in establishing the final price of the deal. One possible explanation is represented by the state of the economy in a particular region. For instance, in the United States, Continental Europe, Canada, or Australia, the deals are expected to be more expensive because the potential targets are more substantial than in other regions with a weaker economy, such as Eastern Europe or Middle East& Africa. In addition, the amount of money raised by the funds is higher in regions with a developed economy than regions with a less developed economy, hence a higher number of deals is expected in those regions. There are several counterintuitive coefficients in the regression. For instance, Latin America has a positive value compared to the United States (omitted variable). The coefficient of Eastern Europe has a higher value than the coefficient of Continental Europe. It is complicated to believe that these relations hold in practice. Furthermore, the coefficients of Eastern Europe, Latin America, Australia, and the Middle East& Africa are not significant. The main reason is the limited number of observations and the high spread in the sample.

Overall, the Heckman model is significant. Having a Wald chi2 of 180.60 and a chi2 value smaller than 0.001. Moreover, the Leverage Ratio test shows if the rho is significantly different from zero, it has a chi2 value of 3.01, which illustrates a p-value of 0.0826 (smaller than 0.1). Using an alpha of 10%, the Heckman Maximum Likelihood model has better results than an OLS regression. In the end, Heckman regression was used to predict all the missing data for deal value. A logarithmic variable for deal value was constructed. It was regressed as a dependent variable using the model described above. After all missing values were filled in for the logarithmic deal value, they were transformed back to deal value using the exponential function. The results of the Heckman regression can be found in the appendix of the paper.

3.4 Hypotheses

This research aims to find out the current trends in the LBO and private equity world. Thus, it determines if the trends presented by Stromberg (2009) before the financial crisis are still unchanged, or there are any significant differences between 2001- 2007 and the researched period of this paper (2011-2020). The first hypothesis relates to the proportion of sub-deals presented in the sample. There are expected proportions that are similar to numbers found by Stromberg (2009).

(20)

20

If there are significant changes, it can illustrate a major shift in popularity for a specific type of deals.

Stromberg finds that from 2001 to 2007, 6.8% of transactions were public takeovers, 36.9% private to private deals, 16.8% secondary buyouts, and 3.2% financial distress deals. The missing part of divisional buyouts represents the difference of 36.3%. This research was not able to operationalize this variable, and it expects that those percentages will be equally distributed between private to private transactions and secondary buyouts. In terms of deal's value proportion, 28.8% of the transaction value were public takeovers, 14.7% private to private, 23.5% secondary buyouts, and 3.2% financial distress deals. 31.6% is missing because of divisional buyouts, and it has the same expectation to be equally divided between private to private and secondary buyout deals

The second hypothesis is illustrated by the popularity of deals in different regions of the world. Stromberg (2009) finds that Continental Europe is above the United States considering deal values and the total number of transactions. It is expected to find Continental Europe, the United States, and the United Kingdom as top countries, while Eastern Europe, Middle East, Africa, and Latin America are on the bottom. Any different results might show a shift in the preference of investors for a specific region.

The third hypothesis tested by this paper is the old idea about the attractiveness of mature industries. Stromberg (2009) finds that this idea was never true and did not hold in this sample. Considering the literature results, it is not expected to find a preference only for mature and established industries. Nevertheless, this research brings on the table a new hypothesis compared to the other paper from the literature. It investigates what the characteristics that can make a potential target attractive for an LBO transaction are. It is expected that firms that have a high level of taxes paid related to the revenue and a large amount of cash out of the total assets on the balance sheet to have a positive effect on the acquisition likelihood. On the contrary, a negative effect over the acquisition outcome is predicted by a growing revenue, high capital intensity and high debt level compared to equity. This paper will further investigate if the presented hypotheses hold for this sample.

(21)

21

3.5 Empirical Models

This section of the research discusses the methods used to test the hypothesis mentioned above. In order to test the first hypothesis and to compare the proportion of the samples, a z- test with proportions will be used. The normality and the random sample are conditions assumed for this test. Both conditions, x*p>5 and x*(1-p)>5, are met for all sub-categories of deals. The same test is going to be used to test the second hypothesis regarding the popularity of the regions. Similar conditions are assumed, and all conditions are met in order to perform the test. If the p-value is below the alpha level of 0.05, the null hypothesis will be rejected, and a change in the sub-deal categories will be concluded.

4. 𝑧 = 𝑝̂−𝑝0

√𝑝0∗(1−𝑝0)

𝑛

In order to determine if there is a significant difference between mature and younger industries, this paper is going to use a z-test. The null hypothesis is represented by the total number of transactions and the total value of deals from a young industry, for instance, service, whereas manufacturing, a mature industry, illustrates the alternative hypothesis. If the p-value is going to be below 5%, then the test will be significant; otherwise, the alternative hypothesis will be rejected.

5. 𝑧 = 𝑥̅− µ

𝜎/√𝑛

In order to test the last hypothesis, logistic regression will be used. Stromberg applied a similar model in his paper to test the likelihood of a successful exit. This paper is going to use logistic regression in order to determine the likelihood that a firm is going to be acquired through an LBO. In addition to the logit, a probability model will be also used to test the likelihood of an acquisition. The two models will be compared in the end. Five variables are going to play a role in these regression models, the revenue growth, the tangible to total assets ratio, debt to equity ratio, tax to revenue ratio and cash to total assets ratio. An important mention is that for all the variables are used the latest financial data (usually one year prior to the acquisition) and the growth of the revenue was constructed as a compound annual growth of the last four years.

(22)

22

Dependent Variable: Type of the Deal Probit and Logit

Constant Revenue Growth Tangible Assets/ Total Assets

Debt/Equity Tax/ Revenue

Cash/Assets

(Table 7). Probit and Logit Model

6. Y= 𝛽₀+ 𝛽₁∗ 𝑋₁+ 𝛽₂∗ 𝑋₂+ 𝛽₃∗ 𝑋₃ + 𝛽₄∗ 𝑋₄+ 𝛽₅∗ 𝑋₅ Probit model:

𝑃(𝑌 = 1|𝑋₁, 𝑋₂, 𝑋₃, 𝑋₄, 𝑋₅) = 𝜙(𝛽₀+ 𝛽₁∗ 𝑋₁+ 𝛽₂∗ 𝑋₂+ 𝛽₃∗ 𝑋₃+ 𝛽₄∗ 𝑋₄+ 𝛽₅∗ 𝑋₅) Where X1,X2, X3, X4 andX5 are the independent variables from Table 7

Logit model:

𝑃(𝑌 = 1|𝑋₁, 𝑋₂, 𝑋₃, 𝑋₄, 𝑋₅) = 1

1 + 𝑒𝛽0 +𝛽1∗𝑋1+𝛽2∗𝑋2+𝛽3∗𝑋3+𝛽4∗𝑋4+ 𝛽5∗𝑋5

Where X1,X2, X3, X4 andX5 are the independent variables from Table 7

4. Results and Discussion

In this part of the paper, the main results of the statistical tests are discussed and the potential implication of the outcomes. In order to test the first hypothesis, a z-test with proportions is used. The result shows that in terms of number of deals per categories, only public takeovers follow a similar proportion to Stromberg’s paper. The statistical outcomes for other three categories reject the null hypothesis; thus, the sub-deals follow a different distribution in the sample. The p-value for Public Takeovers distribution is equal to 0.2274. The p-value for private to private, secondary buyouts and financial distress is below 0.001, which makes the rejection significant. The private to private deals and secondary buyouts have a higher number of transactions than Stromberg’s sample. The results can be found in table 9.

Considering the average prices of a private to private transaction and a secondary buyout, two important conclusions can be drawn. The private to private deals are the cheapest with a mean deal around $350 million. Therefore, the investors prefer to invest in private businesses because a

(23)

23

lower price can be achieved. About 63% of all deals in this sample are private to private. The second number of highest transactions in the sample is represented by secondary buyouts (30.37%). The average deals price is about $750 million, which is four times cheaper than the average public takeover ($3200 Million). Using the same rationale, the private equity funds can achieve a cheaper transaction and the negotiation process is less complex when only private entities are involved. This is a major shift in the LBO trends compared to Stromberg’s period. This paper concludes that is better to leave apart the financial distress deals because the number of observations is small, and the normality condition is not met.

The four sub-deals are all significant and the null hypotheses are rejected. The public takeovers have a p-value equal with 0.0113 while the other sub-categories have a p-value smaller than 0.001 (see table 10). Thus, the distribution of value in this sample is completely different from Stromberg’s paper. The reasoning is similar to the total number of deals. More opportunities are taken by the private equity funds in the private transactions. Thus, the total amount of money spent on private to private transactions and secondary buyouts is higher than in Stromberg’s study. Potential reasons are cheaper deals, and easier negotiation due to a simpler ownership structure.

In terms of ranking per regions, the United States, Continental Europe and the United Kingdom are top three countries, but the average deal value is significantly different from Stromberg’s sample. The p-value for the three regions is below 0.001. The average value per deal in the US is above $1000 million, while Stromberg finds only $389 million. For Continental Europe the mean amount paid is $351 million compared to $439 million in the Stromberg’s study. This paper believes that the difference in the results is due to the selection biased. Zephyr contains more United States transactions registered than any other region. On the other hand, a potential explanation could be that investors value more firms from the US, and they accept to pay more for a firm from this region. The hypothesis regarding the old mature industries is in line with findings made by Stromberg. The amount of money spends on transactions that are part of Services industry is significantly higher than Manufacturing. Considering services, a modern industry, the null hypothesis is rejected.

Regarding the likelihood of target to be acquired through an LBO the results are in accordance with the existing literature. The models are very similar in terms of outcome. Both probit and logit illustrate that the growth of the revenue and the capital intensity have a negative impact over the likelihood of an LBO, whereas the tax ratio has a positive impact. The described

(24)

24

variables are significant in both models at a p-value smaller than 0.05. (see table 10). On the other hand, the debt to equity ratio is very close to zero and does not have an influence over the likelihood of the transactions, but it is not significant from a statistical point of view with a p-value o equal to 0.355. On the contrary, the cash to assets ratio illustrate a negative impact in both models, but this ratio is also statistically insignificant due to the high p-values (close to 0.2). A potential explanation for this result could be the higher valuation that M&As have compared to LBOs. Thus, larger corporations were acquired through M&A and considering the common trend of high hoards of cash on the balance sheet, the cash to assets ratio is higher for M&As compared to LBOs. Both models are significant overall having a likelihood ratio of 80.42 (probit) and 80.09 (logit). The p-values for the chi2 are significant for both models (p<0.001).

This paper encountered several limitations. The small amount of data is a serious issue for this research. There are several regions and industries with below 30 observations. Therefore, the statistical tests might be questionable for such a small number of transactions. On the other hand, the selection biased illustrates another problem. Zephyr is biased especially towards the United States. Most of the registered deals are coming mainly from the United States. This might question the external validity of this research. Thus, if this sample is representative for the entire world of LBOs and private equity might be challenging assumption.

5. Conclusion

This study had two objectives. The first goal was to understand if there are any significant changes in the LBO market in terms of sub-deals, regions and industries. It can be concluded based on the performed tests that private to private transactions and leveraged buyouts are very popular compared to 2001-2007. The main reason is related to the average deal price and the ownership structure. Considering the negotiation knowledge of a PE fund, a better price can be obtained easier when there are a reduced number of shareholders. In addition, a smaller price paid for the deal means a higher potential to create value for the PE funds. This paper believes that a private company might have smaller problems related to the governance because the owners can spend more time checking the managers of the companies. This is a research recommendation for the future.

The transactions per region, still follow the same order, but the total number of deals and the total amount of funds spent on the deals are different. The US is a very popular target for

(25)

25

investors. Around one billion dollars are spent on average on LBOs in the US. It is a significant difference compared to Continental Europe where the average deal is less than 400 million dollars. The investors consider United States as a better place for their investments. This conclusion can be reached also by factoring only the price, the funds are able to pay twice as much compared to Continental Europe on average. A proposal for a future research is to investigate to what extent, the US companies provide a higher value at the end of the LBO transaction compared to the European counterparties.

Regarding the industries, this paper tested only if the mature industries are in higher demand compared to the new industries. The outcome is as expected, the maturity of an industry does not play a significant role. There are different characteristics that are not related to mature industries that can make a potential target attractive. Thus, the conclusion is that the financial data of the attractive targets are not necessarily representative only for mature industries such as manufacturing sector.

The second objective of the paper, which is the biggest contribution is represented by determining the characteristics of a successful LBO. Two models were used (probit and logit). The conclusion is that revenue growth, higher taxes relative to the revenue and capital intensity have a significant effect over the likelihood of an LBO. In correlation with the previous literature, a high revenue growth and capital intensity have a negative effect over the probability of a successful LBO transaction while a high amount of taxes has a positive impact. Nevertheless, the debt to equity ratio and the cash to assets ratio are insignificant variables for this model.

(26)

26

6. References

Bain & Company. (2020). Global Private Equity Report 2020. Retrieved from

https://www.bain.com/globalassets/noindex/2020/bain_report_private_equity_report_2020.pdf Cambridge Associates. (2019). Private Equity Index and Selected Benchmark Statistics. Retrieved from

https://www.cambridgeassociates.com/wp-content/uploads/2020/06/WEB-2019-Q4-Global-Private-Equity.pdf

Espen Eckbo, B., & Thorburn, K. (2012). Corporate restructuring. Foundations and Trends in Finance, 7(3), 159–288. https://doi.org/10.1561/0500000028

Gertner R., and Kaplan, S. (1996). The Value Maximizing Board. Available at SSRN: https://ssrn.com/abstract=10563

Heckman, J. (1979), “Sample selection bias as specification error,” Econometrica 47, 153-162

Jensen, M. (1986). Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers. The American Economic Review, 76(2), 323–329. https://doi.org/10.2307/1818789

Jensen, M. (1989). Eclipse of the Public Corporation. Harvard Business Review, 67(5), 61–74. https://doi.org/10.2139/ssrn.146149

Jensen, M., & Murphy, K. (1990). Performance Pay and Top-Management Incentives. The Journal of Political Economy, 98(2), 225–264. https://doi.org/10.1086/261677

Kaplan, S. (1989). The effects of management buyouts on operating performance and value. Journal of Financial Economics, 24(2), 217–254. https://doi.org/10.1016/0304-405X(89)90047-0

Kaplan, S. (1991). The staying power of leveraged buyouts. Journal of Financial Economics, 29(2), 287–313. https://doi.org/10.1016/0304-405X(91)90004-4

Kaplan, S., & Strömberg, P. (2009). Leveraged Buyouts and Private Equity. The Journal of Economic Perspectives, 23(1), 121–146. https://doi.org/10.1257/jep.23.1.121

Le Nadant, A., & Perdreau, F. (2006). Financial profile of leveraged buy-out targets: some French

evidence. Review of Accounting and Finance, 5(4), 370–392.

https://doi.org/10.1108/14757700610712444

Metrick, A., & Yasuda, A. (2010). The Economics of Private Equity Funds. The Review of Financial Studies, 23(6), 2303–2341. https://doi.org/10.1093/rfs/hhq020

Ofek, E. (1994). Efficiency Gains in Unsuccessful Management Buyouts. Journal of Finance, 49(2), 637–654. https://doi.org/10.1111/j.1540-6261.1994.tb05155.x

(27)

27

Rappaport, A. (1990). The Staying Power of the Public Corporation. Harvard Business Review, 68(1), 96–104. http://search.proquest.com/docview/1296479854/

Steven, D., Haltiwanger, J., Jarmin, R., Lerner, J., & Miranda, J. (2008). “Private Equity and Employment.” U.S. Census Bureau Center for Economic Studies Paper CES-WP-08-07. http://papers.ssrn.com/sol3/papers.cfm?0abstract_id_1107175

Stromberg, Per. (2007). The New Demography of Private Equity. The global impact of private equity report.

https://www.hhs.se/contentassets/662e98040ed14d6c93b1119e5a9796a4/strombergdemography2 008.pdf

Stromberg, P. (2009) The Economic and Social Impact of Private Equity in Europe: Summary of

Research Findings. Available at

SSRN: https://ssrn.com/abstract=1429322 or http://dx.doi.org/10.2139/ssrn.1429322

Viral A., & Kehoe C. (2008). “Corporate Governance and Value Creation Evidence from Private Equity.” http://www.ecgi.org/ competitions/rof/files/Acharya_Kehoe_v5.pdf

Why Do Private Acquirers Pay So Little Compared to Public Acquirers? (2007). National Bureau of Economic Research.

Wright, M., Renneboog, L., Simons, T., & Scholes, L. (2006). Leveraged Buyouts in the U.K. and Continental Europe: Retrospect and Prospect. Journal of Applied Corporate Finance, 18(3), 38– 55. https://doi.org/10.1111/j.1745-6622.2006.00097.x

(28)

28

7. Appendix

Table(8) Sub-Deals Distribution per Transactions Proportion Tests

Sub-Deals- Total Transactions Variable z-test proportion tested sample proportion Public Takeovers -1.210 0.068 0.061 (-0.227) Private to Private 24.81 0.369 0.632 (0.000)*** Secondary Buyouts 16.52 0.168 0.304 (0.000)*** Financial Distress -7.53 0.032 0.003 (0.000)*** Note: * p<0.05; ** p<0.01; *** p<0.001

Table(9) Sub-Deals Distribution per Value Proportion Tests

Sub-Deals- Total Value Variable z-test proportion tested sample proportion Public Takeovers -3.010 0.288 0.258 (0.001)** Private to Private 35.60 0.147 0.424 (0.000)*** Secondary Buyouts 7.73 0.235 0.307 (0.000)*** Financial Distress -5.17 0.032 0.012 (0.000)*** Note: * p<0.05; ** p<0.01; *** p<0.001

(29)

29

Table(10) Likelihood of a target to become LBO

Dependent variable: Type of Deal

Probit and Logit Regressions

Probit Logit

Constant 0.142 0.321

(2.60)** (2.64)**

Revenue Growth -2.94E-07

-4.89E-07 (-2.49)** (-2.39 )** TangibleAssets/Total Assets -1.142 -1.847 (-7.90)*** (-7.70)** Debt/Equity 0.003 0.005 (-0.95) 0.93 Tax/ Revenue 0.348 0.593 (2.12)** (2.05)** Cash/Assets -0.258 -0.42 (-1.29) (-1.3) LR chi2(5) 80.42 80.09 Observations 1,550 1,550 Note: * p<0.05; ** p<0.01; *** p<0.001

(30)

30

Table (11) Distribution per Region Proportion Tests

Countries- Total Number of Deals Variable z-test proportion tested sample proportion US 8.690 0.348 0.439 (0.000)*** Continental Europe 13.74 0.176 0.291 (0.000)*** UK -14.29 0.287 0.145 (0.000)*** Note: * p<0.05; ** p<0.01; *** p<0.001

(31)

31 LR test lambda sigma rho /lnsigma /athrho _cons UnitedStates 0 (omitted) EasternEurope Asia UnitedKingdom Canada LatinAmerica Scandinavia ContinentalEurope Australia MiddleEastAfrica ExitTransaction FinSponsor PrivatetoPrivate 0 (omitted) FinancialDiss SecondaryBuyout PublicTakeover select _cons UnitedStates 0 (omitted) EasternEurope Asia UnitedKingdom Canada LatinAmerica Scandinavia ContinentalEurope Australia MiddleEastAfrica ExitTransaction FinSponsor PrivatetoPrivate 0 (omitted) FinancialDiss SecondaryBuyout PublicTakeover lnDealVal lnDealVal Log likelihood = -2731.874 Wald chi2(14) = 180.60 Nonselected = 1,259 (regression Heckman selection model Iteration 4: of indep. eqns. (rho = 0): chi2(1) = 3.01 Prob > chi2 = 0.0826 -.442236 .2542477 -.9405524 .0560804 1.697106 .0572807 1.588471 1.813171 -.2605824 .1437069 -.5145602 .0354314 .5289246 .033752 15.67 0.000 .4627719 .5950772 -.2667331 .154176 -1.73 0.084 -.5689125 .0354463 -1.843789 .1569103 -11.75 0.000 -2.151327 -1.536251 .1770587 .2819165 0.63 0.530 -.3754875 .7296049 1.634995 .2118404 7.72 0.000 1.219796 2.050195 .5110046 .0882191 5.79 0.000 .3380984 .6839108 -.1335488 .226393 -0.59 0.555 -.5772709 .3101733 1.765963 .406638 4.34 0.000 .9689668 2.562958 .5418459 .1971887 2.75 0.006 .1553633 .9283286 .213532 .0718683 2.97 0.003 .0726727 .3543914 1.446897 .2266706 6.38 0.000 1.002631 1.891163 .7715755 .3833122 2.01 0.044 .0202974 1.522854 .2891405 .1617856 1.79 0.074 -.0279534 .6062345 .7122113 .201307 3.54 0.000 .3176567 1.106766 1.465269 .7051696 2.08 0.038 .083162 2.847376 .1350918 .1644642 0.82 0.411 -.1872521 .4574356 2.429011 .2196612 11.06 0.000 1.998482 2.859539 5.648561 .5416846 10.43 0.000 4.586878 6.710243 -.2339182 .6417705 -0.36 0.715 -1.491765 1.023929 -1.084829 .3151016 -3.44 0.001 -1.702416 -.4672409 -1.432603 .1917901 -7.47 0.000 -1.808505 -1.056702 1.136916 .4787989 2.37 0.018 .1984869 2.075344 .1470418 .3905477 0.38 0.707 -.6184177 .9125013 -.9559811 .3674914 -2.60 0.009 -1.676251 -.2357112 -.9154726 .156989 -5.83 0.000 -1.223165 -.6077798 -.2957856 .3453699 -0.86 0.392 -.9726981 .3811269 -.5909908 .706467 -0.84 0.403 -1.975641 .7936591 1.421192 .344208 4.13 0.000 .7465571 2.095828 -.9468246 .3911639 -2.42 0.015 -1.713492 -.1801574 .9978877 .7939306 1.26 0.209 -.5581876 2.553963 -.726842 .3447209 -2.11 0.035 -1.402483 -.0512016 1.217855 .3131703 3.89 0.000 .6040522 1.831657 Coef. Std. Err. z P>|z| [95% Conf. Interval] Prob > chi2 = 0.0000 model with sample selection) Selected = 812 Number of obs = 2,071 log likelihood = -2731.8742

Iteration 3: log likelihood = -2731.8744 Iteration 2: log likelihood = -2731.9096 Iteration 1: log likelihood = -2734.6929 Iteration 0: log likelihood = -2743.2595

Table (12) Heckman Regression- Missing Values

The Current State of LBOs

The Current State of LBOs

Table of Contents