University of Amsterdam
Refinancing in the EU:
Private Equity vs. Non-Private Equity Owned Firms
Master Thesis
Master in International Finance Prepared by: Joseba Ancín (10679448)
Supervisor: Drs. Jens Martin August 2016
University of Amsterdam
2 Abstract
This paper investigates the leverage lending scenario in the EU 28 with specific focus on financing of debt by private equity (PE) firms in contrast with non-private equity companies, such as public limited companies (PLCs) or any other private company whose core activity does not involve buyout investments via any funds. We included in the research a time series from 1999 to 2015, consisting of a Euro Loan Pipeline to finance different purpose deals across Europe. From the all the different purpose deals included in our sample, we focus most deeply on the refinancing and recap deals. The findings of the analysis suggest that private equity firms use a different approach in refinancing practices compared to non-private equity firms. Private equity firms show more proactive management practices when dealing with refinancing debt. The results of the empirical research also disclose that private equity firms have more predominant volume for refinancing deals in the years leading up to the credit crunch crisis than public companies and non-private equity firms. Private equity firms changed their behavior once the credit crunch crisis subsided, by decreasing substantially the volume of refinancing deals.
Keywords
3
1. Introduction ... 4
2. Literature Review ... 5
3 Methodology and hypothesis ... 10
3.1 Different approaches to debt ... 10
3.2 Propensity to refinancing ... 12
4. Data and descriptive statistics ... 15
4.1 Euro Loan Pipeline ... 15
4.2 Variable definitions ... 16
4.3 EURIBOR and GDP ... 17
4.4 Descriptive Statistics ... 17
5. Results ... 19
5.1 Testing the hypothesis ... 19
5.2 Growth in refinancing’s ... 23
5.3 Maturity Terms for refinancing ... 24
5.4 Euro deal size ... 25
5.5 Correlation Table: EUR deal size and maturity ... 26
5.6 Descriptive statistics and further t-test ... 27
5.7 Further empirical analysis and robustness check ... 27
5.8 Spread: are the spreads for non PE owned firms going up after the crises? .... 28
5.9 Refinancing volume: Euro Equivalent Deal Size (EDS) ... 29
5.10 Pearson goodness-of-fit test ... 30
5.11 Additional variables analysis ... 31
5.11.1 Structure ... 31
5.11.2 Terms to maturity ... 33
5.11.3 Spreads ... 34
5.11.4 Debt to EBITDA ratio ... 36
5.11.5 GDP EU28 and EURIBOR ... 37
5.11.6 Industry allocation ... 38
5.11.7 Country distribution ... 39
6. Conclusion ... 40
7. References ... 44
4
1. Introduction
The role of private equity firms in debt financing has been crucial in the last decade. Private equity investments have become popular among institutional investors since these types of investments have proven to outperform public markets on average. The private equity industry has matured greatly during the last thirty years. High yield return results and stable performance over the years have made private equity a more predominant industry, attracting the interest of investors and contributing to the overall importance of the economy.
The statement that private equity outperforms public markets has been supported by a vast amount of empirical research. Harris, Jenkinson, Kaplan and Stucke (2015) argue that buyout funds have outperformed public markets. In our paper, we focus on the 28 member countries of the European Union and the leverage lending scenario under Private Equity Firms and other non-Private Equity Firms, such as publicly listed companies or any other public company.
Our research analyses the activity of private equity firms and non-private equity firms through a time series comprising years 1999 to 2015. Factors like volume and the number of deals for private equity and public markets can be observed as they change over time. In line with this, a more in-depth analysis is carried out for the most predominant types of deals, such as acquisitions, LBOs, refinancing and recaps. Furthermore, we will closely examine the credit crisis period of 2008 to 2009 in order to observe in greater detail the debt financing practices of private equity firms versus non-private equity firms or public companies. The aim of the last part of the research is to
5 assess whether private equity firms have an approach that contrasts with non-private equity firms. For example, when it comes to issuing debt, how do these two groups behave in relation to downward and upward economic and industry cycles.
Following the introduction, the paper is structured as follows. In section 2 we discuss the literature related to the topic. Section 3 describes the methodology and hypothesis. Section 4 provides information about the data and the variables utilized during the empirical research. Section 5 includes the results with a presentation of the key findings of the research. Lastly, section 6 presents the conclusion.
2. Literature Review
A large body of literature has adumbrated the value-added role of private equity firms to the companies they invest in over time. Several studies prove that enormous growth has taken place in investor allocations to private equity funds since the late 1990s. Moreover, studies suggest that buyout funds have consistently outperformed public markets for some time. Harris, Jenkinson and Kaplan (2015) state that buyout funds have outperformed public markets, particularly the S&P 500, net of fees and carried interest, in the 1980s, 1990s and 2000s. Their estimates imply that each dollar invested in the average fund returned at least 20% more than a dollar invested in the S&P 500. This works out to an outperformance of at least 3% per year. Phalippou (2013) found that the average buyout fund outperforms the S&P 500 by about 5.7% per annum. European private equity markets are also outperforming public markets. Harris, Jenkinson and Kaplan (2015) also found that average buy out performance appears
6 quite similar between Europe and North America. On both continents, buyout funds have historically outperformed public market returns by similar magnitudes.
Conventional wisdom suggests that investors in private equity funds may be more likely to invest in partnerships that have performed well in the past. According to Harris, Jenkinson, Kaplan and Stucke (2014), there is a general belief that performance in private equity persists across funds of the same partnership. In their research they found persistence for buyout funds that started prior to 2001. Since 2000, however, they did not found enough evidence of persistence. This fact could raise the possibility that the buyout business has changed, as operations and active management factors have become increasingly important. The decline in persistence nurtures some doubt about the industry’s preference for investing in funds that were previously top performers. This suggests there may be other relevant factors to consider, along with past performance, when assessing partnerships that have performed well in the past.
Puche, Braun and Achleitner (2014) conclude that value creation was achieved, in part, by taking financial risk through levered financing of transactions. Another large percentage of value creation was generated through operating improvements within the portfolio companies, which could be the result of excellent management teams or general economic growth. Another factor that contributes to value creation is increasing the transaction-multiple from entry to exit, which may indicate the general partner’s negotiating ability or luck in market timing.
In another recent study on secondary buy outs, Degeorge, Martin and Phalippou (2015) examine when private equity firms can add value by selling companies to each
7 other in secondary buy outs (SBO). The study discovered that complementary skills between the buyer and the seller can result in better performance for secondary buy outs, creating value for investors. This can occur when the secondary buy out takes place between a private equity firm focusing on margin growth and a private equity firm focusing on sales growth, or between two private equity firms in which the general partners have different educational backgrounds or career paths.
A very surprising finding in the research of Puche, Braun and Achleitner (2014) was the extent to which value creation declined over time. Transactions exited before the year 2000 were able to achieve higher returns, 4.3x their investment, than recent transactions exited after 2008 with only 2.8x their investment. Along with the decline in value creation, the use of leverage also decreased in absolute and relative terms. Leverage decrease was also reflected in the sample data utilized to carry out our research.
Our data is constructed on a time series from 1999 to 2015. The main research question is how private equity firms and non-private equity firms use leverage through different periods of the time series. In addition, another important contribution this research focuses on is finding different strategies for debt refinancing during economic cycles along the time series. In light of this, the crisis period of 2008, known as the credit crunch crisis, plays a crucial role as a specific event in this paper.
While closely examining the credit crunch crisis, one of the topics this thesis will focus on is, how private equity firms and leveraged buyouts could weather the crisis period, retain a solid balance sheet, and outperform the market rate. Ivashina and
8 Scharfstein (2009) investigated bank lending during the financial crisis. Their research found that new lending declined substantially during the financial crisis across all types of loans. Additionally, their research showed that new loans to large borrowers fell by 47% during the peak period of the financial crisis (fourth quarter of 2008) relative to the prior quarter, and fell by 79% relative to the peak of the credit boom (second quarter of 2007). New lending for real investment (such as working capital and capital expenditures) fell by only 14% in the last quarter of 2008, but contracted nearly as much as new lending for restructuring (LBOs, M&As, share repurchases) relative to the peak of the credit boom. These findings are very much in line with the results obtained by our research, after carrying out an in-depth analysis of refinancing deals for the whole time series. In the time series analysis, presented later in this paper, we found that the volume of refinancing deals by private equity firms decreased substantially in 2008, coinciding with the peak of the credit crunch crisis.
Wright, Cressy, Wilson and Farag (2014) investigated the role of financial restructuring in private equity backed buyouts in the UK before and after the financial crisis of 2007. They showed that private equity–backed buyouts already contained provisions to optimize recovery rates under insolvency, which raised their recovery rates significantly relative to a control group of public limited companies, both before and after the crash. This active monitoring and management of debt in leveraged buy out structures could be a determinant factor in the success of private equity funds during crisis periods. They also found that private equity buyouts rapidly adjusted the capital structures of new deals in response to changes in the financial and economic climate
9 from 2007 onwards. They discovered that high leverage in pre-existing deals increased the insolvency risk of firms unable to adjust their capital structure prior to or during the downturn. Nevertheless, there was a greater adjustment over time in the leverage of buyouts than for other private companies and matched listed corporations.
In this research, we will investigate whether private equity firms are using best practices when actively managing debt in leveraged buyout (LBO) transactions, and how they managed to outperform public markets before and after the crisis. Moreover, specific attention will be given to the role of financial restructuring before and after the financial crisis of 2007–2008. Existing findings indicate that active involvement by private equity firms helps portfolio companies better deal with trading difficulties in a rapidly deteriorating financial climate. Wright, Cressy, Wilson and Farag (2014) showed that there has been greater adjustment over time in the leverage and cash position of private equity buyouts than for other private companies and matched public limited companies (PLCs). This resulted in a sharply declining rate of bankruptcy for private equity firms over the 12-year time series studied. It also indicated a lower rate of bankruptcy for private equity buyouts than for public limited companies post-2007. Active involvement by PE firms in helping portfolio companies deal with trading difficulties plays an important role when assessing bankruptcy and profitability. Also, the governance mechanisms in private equity buyouts resulted in greater preservation of value when a portfolio firm enters formal bankruptcy. The secured debt recovery rates for private equity buyouts are actually double those of PLCs, like for like.
10 The main contribution of this paper is to provide some insights on the different approaches of PE and non-PE firms when dealing with debt. This research will dive deeper into the credit crunch crisis and examine changes in the behavior patterns of PE and non-PE firms related to refinancing during the crisis. The following section of this paper describes the methodology used and formulated hypothesis.
3 Methodology and hypothesis
In this section we discuss the methodology and hypothesis deployed in this research.
3.1 Different approaches to debt
Among the vast literature regarding how PE can bring added value, research found that PE firms use debt to boost returns. Ceteris paribus, using more leverage (i.e. debt) means that the PE firms will earn a higher return on its investment. Colla, Ippolito and Wagner (2012) found that leverage in LBOs is positively related to profitability. De Maeseneire and Brinkhuis (2012) presented that LBOs are characterized by their very intensive use of debt financing. The leverage ratio is an important aspect of our research as well, and according to the literature, it is a common measurement denominator among the studies of PE firms.
It seems obvious, as mentioned in the expanded literature on LBOs and PE firms, that PE firms will be more inclined to have a higher leverage ratio than non-PE firms. In their research, De Maeseneire and Brinkhuis (2012) stated that when credit conditions loosen, LBOs use relatively more debt. They also found that the reputation of the PE player involved plays a role in the financing choice regarding LBOs. Higher reputation of
11 the PE involved is usually also associated to better financing conditions and more leverage.
As we are examining the different approach of PE and non-PE firms when dealing with debt and refinancing practices, we utilize a sample database consisting of deals pertaining to the Euro Loan Pipeline time series volume from 1999 to 2015. This database will help us analyze characteristics behind the refinancing deals in the time series and determine whether they belong to PE or non-PE firms. In this way, we will be able to examine the main variances when it comes to debt and refinancing practices.
Another important focus of this paper is to look for evidence of behavioral changes in debt and refinancing practices as a result of economic events during the time series 1999–2015. For instance, the private equity activity was meteoric during the years 2006 to 2007, but the activity declined vertiginously in 2008 after the credit crunch crisis hit the economy.
De Maeseneire and Brinkhuis (2012) found that debt market conditions do not impact public peer leverage, but they are significantly related to LBO leverage. Thus, as suggested by practitioners, the capital structure choice with respect to LBOs is heavily influenced by prevailing conditions in the debt market. As indicated in the literature, when credit conditions are loose, LBOs use relatively more debt. This assumption also suggests that credit conditions are often loosened during boom periods, as was the case from 2006 to 2007.
In light of this, and to further investigate changes in refinancing and debt practices, we distinguish three main periods in our time series. The first period ranges from the
12 start of our sample in 1999 to 2007, which end in the year right before of the credit crunch crisis hit the economy. The subsequent period, from 2008 to 2009, represents the period of the financial crisis, starting with the collapse of Lehman Brothers in August 2008. The following period, 2010 to 2015, contain the most recent transactions after the financial crisis. Hence, we analyse the time series into three periods of time, (1999– 2007), (2008–2009), and (2010–2015). The purpose of this period split is to gain a more holistic view, comprehending the periods prior, during and after the credit crunch crisis of (2008–2009). Several stages of economic cycles are represented in three periods, namely the economic boom before 2007, followed by the credit crunch crisis from 2008 until 2009, and the post-crisis periods of 2010 to 2015.
In the following section we present the hypothesis we will use to ascertain changes in the refinancing practices of PE and non-PE firms through the three defined periods in the time series.
3.2 Propensity to refinancing
The hypothesis defined to convey our research is as follows.
Hypothesis: Private equity firms are more likely to refinance during the credit crunch
crisis (2008 and 2009) than non-private equity firms.
To test the hypothesis we conduct a probit regression in which the dependent variable is a dummy variable (RFN) equal to one (RFN=1) if the purpose is Refinancing and zero for any other means of financing. We construct three independent variables: i) dummy variable (PEF) equal to one (PEF=1) if it is related to Private Equity Firms, and
13 zero otherwise; and ii) variable representing the Euro Equivalent Deal Size (EDS), we run a ln on this variable to normalize the distribution, ln(EuroEquivalentDealSize); and iii) variable representing the spread (SPR), we run a ln on this variable to normalize the distribution, ln(YieldAveragespread).
In this probit regression we estimate the likelihood of Private Equity owned firms and Non Private Equity owned firms to refinance during the credit crunch crisis period (2008 – 2009). A dummy variable is created to define the crisis year rage (2008 – 2009). Crisis dummy: values 1 if years 2008 and 2009 included, and 0 for other years. The data is set as panel data. The time series includes the years 1999 to mid-2015. Interaction to analyse the results is created between the dummy variable Private Equity Firms (PEF) and crisis dummy variable (2008 – 2009) which values 1 if years 2008 and 2009 included, and 0 for other years.
This broader perspective will allow us to test propensity for refinancing among private equity and non-private equity firms for each period in the time series. The model that we wish to fit is as follows:
𝐏𝐫(𝑹𝑭𝑵 = 𝟏) = 𝜶𝟎+ 𝜶𝟏𝑷𝑬𝒊,𝒕+ 𝜶𝟐𝑪𝒓𝒊𝒔𝒊𝒔𝒕+ 𝜶𝟑(𝑷𝑬 × 𝑪𝒓𝒊𝒔𝒊𝒔)𝒊,𝒕+ 𝜶𝟒𝒍𝒏𝑬𝑫𝑺𝒊,𝒕+ 𝜶𝟓𝒍𝒏𝑺𝑷𝑹𝒊,𝒕+ 𝜺𝒊,𝒕 This model will allow us to determine if Private Equity owned firms have a greater propensity to refinance during the credit crunch crisis (2008 – 2009) than non-Private Equity owned firms. Moreover, it will show when a specific event, such as the credit crunch crisis, has a direct impact on PE and non-PE refinancing practices. A negative sign in the coefficient results when testing the hypothesis will indicate that there is lower
14 probability of refinancing for PE firms during the crisis in 2008–2009 and greater propensity of refinancing for non-PE firms.
The aim of the analysis is to find evidence of different approaches when dealing with debt by PE and non-PE firms, as well as proof of changes in their behavior during the credit crunch crisis and the different period splits of the time series, prior to the crisis (1999–2007), during crisis (2008–2009), and after the crisis (2010–2015).
We perform a robustness check for our model using a probit regression with the same variables but in this case we compare the behavior during the whole period comprehending the start the crisis to present time (2008–2015) and the period prior to the crisis (1999—2007) to the rest of the year’s period.
We will also implement a probit regression using the variable spread (SPR) to test the increase of the spreads for non PE owned firms refinancing activity during the crisis period (2008—2015) and after the crisis period (2008—2015) compared to the rest of the year’s period.
A third probit regression will be run using the variable Euro Equivalent Deal Size (EDS) to test the change of refinancing volume in million Euros for PE and non-PE owned firms refinancing activity during the crisis period (2008—2015) and after the crisis period (2008—2015) compared to the rest of the year’s period.
In addition, we also run a further empirical test, a goodness-of-fit test, classification statistics and a marginal effects calculation for the total time series 1999–2015.
Next to testing the hypothesis, we also perform an analysis of other important control variables, such as structure, terms to maturity, spreads, Debt to EBITDA ratio,
15 GDP, EURIBOR, industry allocation and country distribution. By observing these variables, we expect to discover how their behavior differs from PE to non-PE firms and how they undergo alteration during time series events from 1999 to 2015.
4. Data and descriptive statistics
In the following section we present the data utilized in this research, including source, variable definitions and descriptive statistics.
4.1 Euro Loan Pipeline
As the primary aim of this study is to assess debt financing practices in the Euro Zone, we use a Euro loan pipeline database, a time series database from 1999 to 2015, which contains all the details, characteristics and different purpose deals of the total Euro loan volume issued within the EU28 countries.
Capital IQ is the source for our data sample from which we obtain information on the transaction details, such as the Purpose, Primary Sponsor, Broad Category (LBO / Non-LBO), Currency, Deal Size, Terms, Tranche structure, Spreads and Financial Information such as EBITDA. The complete purpose deals compilation included in the database are mentioned forthwith: Acquisition, Capital Spending/Expansion, Corp Purpose, Exit Financing/Exit Financing, Expansion/ Capex, LBO, Leveraged Buildup, MBO, Merger, Project Financing/General, Project Financing/Manufacturing Plant, Project Financing/Utility, Recaps, Refinancing and Spinoff Super-Senior/Super-Senior. Many professionals and institutions in the financial sector use these information details to analyze market trends and support decision-making.
16 4.2 Variable definitions
We define refinancing as a transaction in which a company’s new debt replaces existing debt, and only the debt portion of the capital structure is affected. Since we are focusing on refinancing purpose deals on this research, we created a group consisting of all the refinancing purpose deals from the Euro loan pipeline sample data.
Of all the deal purposes comprising the database, the following were included in the refinancing group: Recap/Dividend, Recap/Equity Infusion, Recap/General Recap, Recap/IPO, Recap/IPO/IDS, Recap/Stock Repurchase and Refinancing.
The Variables Definition Table in the Appendix defines variables we use in the analysis, including RC (Revolving Credit), TLA (Term Loan A), TLB (Term Loan B), TLC (Term Loan C), 2nd Lien (Second Lien), Spread, Terms to Maturity and Debt to EBITDA ratio.
As mentioned before the main variables to run our regression model are defined as follows: to test the hypothesis we conduct a probit regression in which the dependent variable is a dummy variable (RFN) equal to one (RFN=1) if the purpose is Refinancing and zero for any other means of financing. We construct three independent variables: i) dummy variable (PEF) equal to one (PEF=1) if it is related to Private Equity Firms, and zero otherwise; and ii) variable representing the Euro Equivalent Deal Size (EDS), we run a ln on this variable to normalize the distribution, ln(EuroEquivalentDealSize); and iii) variable representing the spread (SPR), we run a ln on this variable to normalize the distribution, ln(YieldAveragespread).
17 4.3 EURIBOR and GDP
We also gather information about EURIBOR interest rates, as well as other macro variables like the Growth Domestic Product (GDP) for European Union (28 countries). We obtained the GDP from the Eurostat databases. For interest rates we use EURIBOR rates obtained from BLOOMBERG databases.
We examine the total size of the data, including all different purpose deals over the time series 1999 to 2015. Once we have a clear picture of the overall total data, we will narrow our focus to the refinancing purpose deals, which comprise the main research topic of this paper.
4.4 Descriptive Statistics
The total value of the Euro loan pipeline amounts to 1,080,317 million Euro, encompassing a total number of 2,821 deals in the EU28 countries comprising the time series from 1999 to mid–2015. After isolating the refinancing purpose deals category group, the total value amounts to 352,155 million Euro and 1,015 deals, of which 197,646 million Euro belong to PE firms and 154,523 million Euro to non-PE firms. Table 1 shows these statistics as well as other additional features like value in million Euro per tranche (RC, TLA, TLB, TLC, and 2nd Lien), spread, Debt/EBITDA ratio, term to maturity per tranche and deal count. The summary statistics in Table 1 are based on the totals covering the years 1999–2015, except for panels E and F, which contain only the pre-crisis period 1999–2007. The panels included in Table 1 are described as follows: Panel A: Total Euro Loan Pipeline; Panel B: Refinancing deals; Panel C: Refinancing PE
18 Firms; Panel D: Refinancing no PE Firms; Panel E: Refinancing PE Firms pre-crisis period 1999 to 2007; Panel F: Refinancing no PE Firms pre-crisis period 1999 to 2007.
Table 2 discloses the refinancing volume in million Euro and number of deals for PE and non-PE firms along the time series 1999 to 2015. Totals are recorded as volume in million Euro and deal count. As presented in the Table 2, the volume of refinancing for PE firms will be higher for the years 2005–2007 than during the credit crunch crisis of 2008–2009. In contrast, non-PE firms show higher refinancing activity during the credit crunch crisis of 2008–2009, both proportionally and by volume.
Table 2 forms the basis for Figure 2, which shows a significant difference between PE firms and non-PE firms in the volume of refinancing deals issued during the years 2005– 2007. The total volume for PE firms is 87,998 million Euro, which is higher than the 26,213 million Euro of non-PE firms in the same period of 2005–2007. These differences are large, suggesting that PE firms were more proactive in the field of refinancing during the years prior to the crisis (2005–2007). Wright, Cressy, Wilson and Farag (2014) found that in the UK, much of the refinancing for PE buyouts occurred before the recession hit in 2007. After the crisis, refinancing activity was very low for PE firms, which might suggest that PE firms were more actively following economic trends and reacting quickly to stimulus from the markets. On the other hand, non-PE firms issued refinancing deals during and after the credit crunch crisis.
Table 3 presents descriptive statistics on refinancing deals and Total Euro Loan Pipeline. Panel A reports refinancing for PE firms and Panel B reports refinancing for non-PE firms, offering descriptive statistics of volume in million Euro per year. The
19 following data features are presented: Observations, Median, Mean, Std. Dev., Min, Max and Total Value.
In Table 3, Panel C reports descriptive statistics for the total Euro Loan Pipeline volume, including refinancing and non-refinancing, in million Euro per year. Panel D offers the same data as Panel C, but the time series is split into four periods of vintage years: 1999 to 2007, 2008 to 2009, 2010 and 2011 to 2014.
In our regression we define the crisis period as (2008 and 2009) and we compare the differences in behavior from PE owned firms versus non-PE owned firms during the rest of the periods of the time series, meaning pre crisis and post crisis. The time series is set in panel data comprehending the years 1999 to 2015. The aim of this analysis is to better capture the change in behavior of refinancing practices by PE and non-PE firms through the specific economic moments that those periods represent.
5. Results
5.1 Testing the hypothesis
In this section we examine the predictions of the private equity firms’ propensity to refinancing through a probit regression using the time series data.
Hypothesis: Private equity firms are more likely to refinance during the credit crunch
crisis (2008 and 2009) than non-private equity firms.
The results in Table 4 show that the coefficient on (PEF) private equity owned firms is negative and statistically significant throughout our specifications during the crisis period (2008—2009). The results are very conclusive when looking specifically into the credit crunch crisis period, 2008 to 2009, where PE owned firms have the lowest
20 propensity to refinancing from all the groups with a negative coefficient (-2.3148). Non-PE owned firms test result is not statistically significant throughout our specifications during the crisis period (2008—2009) and show a coefficient of (-0.5285) which is much higher coefficient than PE owned firms for the same period.
These results indicate that the propensity to refinancing is lower for private equity owned firms than for non-private equity owned firms during the credit crunch crisis, 2008 to 2009. Contrary to PE owned firms, non-PE firms have higher refinancing activity during the crisis period 2008 to 2009. This can also be observed in Table 5, Panel B, where refinancing during 2009 for non-PE firms reached 91% of total volume.
Outside the crisis period (1999—2007) and (2010—2015), the regression results reveal statistically significant and higher propensity to refinancing among PE owned firms with a coefficient (-0.4839). Non-PE owned firms outside the crisis results are not statistically significant and show a coefficient (-0.8635). This evidence proves that private equity firms’ likelihood to refinance during the credit crunch crisis, 2008 and 2009, is very low compared to the years prior and after this period. This test provides suggestive evidence that PE firms are more likely to refinance during periods outside the credit crisis. This finding is consistent with the results of Wright, Cressy, Wilson and Farag (2014), which found that much of the refinancing in the UK occurred before the recession hit in 2007. Colla, Ippolito and Wagner (2012) found that senior tranches become cheaper during hot markets. This evidence suggests that banks and institutional investors systematically loosen lending conditions during active buyout markets. They
21 also stated in their article that senior investors tend to relax lending standards during hot buyout markets and provide cheaper financing.
The results of our hypothesis test on this paper suggest that PE firms showed a tendency to refinance more actively when lending conditions were better (2005–2007), as this period coincided with the booming markets.
Table 5 shows EUR equivalent deal volume for PE owned firms and non-PE owned firms, panels A and B, respectively. It can be seen that non-PE owned firms have proportionally higher percentages of refinancing deals than PE owned firms along the whole time series, with the exception of years 2005 and 2014. By doing an in-depth analysis into the crisis period, refinancing activity was found to be much higher for non-PE owned firms than for non-PE owned firms. For instance, refinancing volume for non-non-PE owned firms during 2008 and 2009 is EUR 1.25 Billion and EUR 9.49 Billion, respectively (Table 5, Panel A). However, for PE owned firms there is almost no substantial refinancing volume in the crisis period (2008–2009), with only EUR 0.237 Billion in 2008. This trend continued in the years after the credit crunch crisis passed. From 2010 to 2013, the total volume of refinancing deals for non-PE owned firms was also higher in relative terms compared to PE owned firms (Table 5, Panels A and B).
In this paper we suggest PE owned firms have a balanced refinancing strategy through the time series, whereas non-PE firms commit most of the refinancing volume (i.e. percentage wise 91% in 2009 and 89% in 2010) right after the credit crunch crisis hit the economy. The same case applies to a lesser extent to the years 2011–2013. This trend shows that non-PE owned firms are refinancing during the financial crisis more
22 actively than PE owned firms in relative terms, as shown in Table 5, Panel A and B. Demiroglu and James (2015) find that the frequency of reputable private equity groups (PEGs) participation in LBO transactions is negatively related to credit risk spreads and lending standards. This finding is consistent with reputable PEGs capitalizing more on favorable credit market conditions. Demiroglu and James (2015) also state that buyouts sponsored by high reputation PEGs are less likely to experience financial distress during the 5 years after the transaction. Our results are in accordance with their research and suggest that PE firms have better and more stable refinancing strategies than non-PE firms. Overall, our findings prove that PE firms were refinancing their debt more actively prior to the credit crunch crisis and leveraging during better lending and market conditions.
As stated in the analysis and the results, we believe that active refinancing during the crisis period (2008–2009), as non-private equity owned firms had the propensity to do, can result in a more costly and tedious process than refinancing in the years prior to the crisis. For instance, credit risk spreads and lending standards were more lax prior to the crisis. It is also important to consider that the propensity of non-PE owned firms to refinance during the years right after the crisis (2010–2013) could imply a certain level of financial distress and bankruptcy.
Wright, Cressy, Wilson and Farag (2014) found that PE buyouts rapidly adjusted the capital structures of new deals in response to the changes in the financial and economic climate from 2007 onwards. Their study also shows that PE-backed buyouts already contained provisions to optimize recovery rates under insolvency. This raised their
23 recovery rates significantly relative to a control group of PLCs both before and after the crash. High leverage of preexisting deals can also increase the insolvency risk of firms unable to adjust their capital structure prior to or during the downturn. Nevertheless, the literature evidences that there was a greater adjustment over time in the leverage of buyouts by PE firms than for other private companies and matched listed corporations.
It would be prudent to mention that some articles in the literature about the crisis also adumbrate that availability of cheap debt financing contributes to "booms" (i.e., overheating) in buyout markets. Axelson, Jenkinson, Stromberg and Weisbach (2009) stated that availability of financing impacts booms and busts in the private equity market.
According to their study, in periods of economic development, LBO funds finance all projects (good and bad). This overconfidence in the market could also lead to increased risk-taking behavior, which could result in default and bankruptcy in some cases. The likelihood of default and bankruptcy is beyond the scope of this paper and it was not included in the methodology.
5.2 Growth in refinancing’s
We aim to have an additional way to our probit model to show the difference in behavior between these two groups, therefore we present a table to show the growth in refinancing’s. Taking the variables PE owned and non-PE owned firms, we run a t-test between the growth in a given year and whether the PE owned companies differ significantly from the non-PE owned firms. The results are shown in the table 4.A, which
24 present a t-test for the difference of average Growth rate Euro Equivalent Deal Size, for the refinancing firms. As shown in Table 4.A, the difference in the average growth rate of the deal size is solely significant for the years 1999-2000, 2004 and 2011. For example, the difference in mean growth in 1999 is -1.88, meaning that the mean growth for the PE sample (0.48) is smaller than for Non-PE sample (2.36). This difference is significant at the 5% level since the accompanying p-value 0.031, is smaller than 5%. In same way the difference in growth for the year 2000 is significant, but for this case the p-value of 0.001 is smaller than 1% level and so significant at the 1% level.
As it can be observed in the table, the growth in refinancing is always higher for the non-PE owned companies in almost every year. During the crisis period of 2008 there is a slight higher growth rate for PE owned firms with a very small difference of 0.04 and in 2009 the growth rate is again much higher for non-PE owned firms with a difference of -13.55 for PE owned firms.
5.3 Maturity Terms for refinancing
We present in table 4.B the t-test for refinancing deals in which we calculate the difference of the maturity terms for PE and non-PE owned firms for each year of the time series. The maturity terms are calculated as an average of maturity in number of years for each deal, consisting of the average number of years for all the tranches conforming each deal: RC, Tla, TLb, TLc and Second Lien. Taking the variables PE owned and non-PE owned firms, we run a t-test for the maturity terms in a given year and whether the PE owned companies differ significantly from the non-PE owned firms. As shown in Table 4.B, the difference in the means is significant for the years 2001 to 2007
25 and 2011 to 2015. For years 2008 and 2009 there are not enough observations to obtain a significance level. In the results for the years that are significant, it can be observed that the means of the maturity terms for PE owned firms are higher for almost every year of the time series. For instance when looking the LBO booming period of 2005, 2006 and 2007, it can be derived from the results that difference in means are 1.71, 1.26, 1.39 higher for PE owned firms respectively for each year with a significance at 1% level. The years prior and after the crisis show consistency with these results proving that PE firms will obtain higher terms to maturity than non-PE owned firms.
The following tables 4.B.1, 4.B.2, 4.B.3 are an extension of table 4.B. On these tables we perform a t-test but we use the tranches separately instead of an average of all tranches for each deal. The results are consistent with the general table 4.B results, and they will show the difference on the means for the tranches, RC, TLA and TLB respectively, the tables also include the t-test outcome. We did not include the TLC and Second Lien tranches due to the lack of significance.
5.4 Euro deal size
We present in table 4.C the t-test for refinancing deals in which we calculate the difference in the average of Euro deal size for PE and non-PE owned firms for each year of the time series. The Euro deal size is calculated as an average of volume in Euro for each deal, consisting of the average including all the tranches conforming each deal: RC, Tla, TLb, TLc and Second Lien. Taking the variables PE owned and non-PE owned firms, we run a t-test for the Euro deal size in a given year and whether the PE owned companies differ significantly from the non-PE owned firms. As shown in Table 4.C, the
26 difference in the means is significant for the years 1999, 2001, 2004, 2006, 2007, 2010 and 2013. For years 2008 and 2009 there are not enough observations to obtain a significance level. In the results for the years that are significant, it can be observed that the means of EUR deal size for PE owned firms are lower for almost every year of the time series. For instance when looking the LBO booming period in 2007, difference in means is -540.23 for PE owned firms with significance at 1% level. The years prior and after the crisis show consistency with these results proving that PE firms will have lower EUR size deals than non-PE owned firms.
The following tables 4.C.1, 4.C.2, 4.C.3 are an extension of table 4.C. On these tables we perform a t-test but we use the tranches separately instead of an average of all tranches for each deal. The results are consistent with the general table 4.C results, and they will show the difference on the means for the tranches, RC, TLA and TLB respectively, the tables also include the t-test outcome. We did not include the TLC and Second Lien tranches due to the lack of significance.
5.5 Correlation Table: EUR deal size and maturity
We present in table 4.D a correlation table for EUR deal size and the terms of maturity. The observations are related to the total level, including the total volume in EUR deal size for refinancing and non-refinancing deals, as well as both PE and non-PE owned deals. For the Euro deal size all the tranches are taken into account: RC, Tla, TLb, TLc and Second Lien, as well as the total EUR deal size for all tranches together, i.e. Euro equivalent deal size. The maturity terms will be calculated as an average of the terms of all the tranches (RC, Tla, TLb, TLc and Second Lien).
27 As it can be observed in table 4.D, there is a negative correlation of -0.129 with significance level at 1% between the Euro Equivalent Deal Size and the Average number of years. These means, the higher the Euro Equivalent Deal Size is, the lower the Average number of years to maturity.
5.6 Descriptive statistics and further t-test
In the appendix 4.E we present the outcome of several descriptive statistics and t-test output for the data and variables included in our regression model(s). These
descriptive statistics can be used to have a better understanding and interpretation of the regression results and robustness checks deployed in the following sections.
5.7 Further empirical analysis and robustness check
In order to run a robustness check on our model we perform a probit regression to analyze the behavior during the period comprehending the start the crisis to actual time (2008–2015). In this probit regression we estimate the likelihood of Private Equity owned firms and Non Private Equity owned firms to refinance from the start of the credit crunch crisis period until present time (2008 – 2015). A dummy variable is created to define the start of the crisis until present year, range (2008 – 2015). Crisis dummy: values 1 for all years after 2007, and 0 before 2008. The data is set as panel data. The time series includes the years 1999 to mid-2015. Interaction to analyse results is created between the dummy variable Private Equity Firms (PEF) and the crisis dummy variable (2008–2015). The regression model and variables are similar to the model used in the hypothesis test (shown in previous Table 4).
28 The results are presented in the new Table 4.1, Non-PE owned firms have statistically significant and the highest coefficient (0.7514) for refinancing from the start of the crisis to present period (2008—2015). PE owned firms have no statistically significant results for the period (2008—2015) and a lower coefficient (0.0344) than non-PE firms. PE owned firms show statistically significant results for the period before the crisis (1999— 2007) with a coefficient (-0.3276).
This finding indicates the Non-PE owned firms have the tendency to refinance more in the period during and after the credit crunch crisis (2008—2015) that PE owned firms.
The outcome is consistent with the first regression model and leads to the same conclusions, thus confirming the key findings discussed under the Results section 5.1.
5.8 Spread: are the spreads for non PE owned firms going up after the crises? In this research we found evidence on the spread for non PE owned firms going up dramatically after the crises, whereas they were similar to the PE owned companies prior to the crisis as shown in Table 11 figure 11. (The spread analysis is carried out in detail on section additional analysis 5.5.5 Spreads. )
In order to ascertain our findings we are performing hereby an empirical test for the spread variable in which we compare the different behavior of PE and non-PE owned companies in the crisis period (2008—2009) and the start of the crisis until present, period (2008—2015). We utilize the same probit regression model as before, in Table 4.2 we present the results for the crisis period (2008—2009), and in Table 4.3 for crisis period (2008—2015). Interaction to test results is created between the dummy variable Private Equity Firms and the crisis dummy and the Spread, ln(YieldAveragespread).
29 Table 4.2 show the results are statistically significant at 10% level for non-PE owned firms during crisis period (2008—2009) where the coefficient is (-1.5953), whereas for the period of start and after the crisis (2008—2015) as shown in Table 4.3, the results are higher and statistically significant at 5% level with a coefficient (-0.8599). This result provides with suggestive evidence of spreads going up dramatically for non-PE owned firms after the crises.
5.9 Refinancing volume: Euro Equivalent Deal Size (EDS)
Last but not least, we run a probit regression to test for the Euro Equivalent Deal Size (EDS) variable. In this test we compare the different behavior of PE and non-PE owned companies in the crisis period (2008—2009) and the start of the crisis until present, period (2008—2015). We utilize the same probit regression model as before. The new Table 4.4 presents the crisis period (2008—2009), and Table 4.5 presents the period during and after the crisis (2008—2015). Interaction to test results is created between the dummy variable Private Equity Firms and the crisis dummy and the Euro Equivalent Deal Size (EDS), ln(EuroEquivalentDealSize).
As shown in Table 4.4, for the period (2008—2009), non PE owned firms during the crisis (2008 and 2009) have a positive coefficient (0.6706) significant at 5% level. This coefficient is much higher than the other coefficients on this table which indicates that non PE owned firms do more refinancing during the crisis period (2008—2009).
As shown in Table 4.5, related to the period (2008—2015), PE owned firms during and after the crisis (from 2008 to 2015) have a negative coefficient (-0.2288) significant at 1% level. This coefficient is lower than the rest of the coefficients in this table which
30 indicates PE owned firms do less refinancing than Non PE owned firms in the period (2008 to 2015).
It can be deduced that these results are in line with the rest of the empirical analysis carried out in the previous regressions meaning non PE owned firms do have a higher propensity to refinancing during and after the crisis than non PE owned firms.
5.10 Pearson goodness-of-fit test
In addition to the robustness check, we employ the Pearson goodness-of-fit test or the Hosmer–Lemeshow goodness-of-fit test. The Pearson goodness-of-fit test is a test of the observed against expected number of responses. Table 6, Probit model for refinancing goodness-of-fit test, shows the results of this test, with 2,072 number of observations versus 1,034 number of covariate patterns. Our model fits reasonably well, thus we cannot reject our model.
Furthermore, we present classification statistics for the empirical test probit regression data in Table 7. The overall rate of correct classification is estimated to be 66.40%, with 91.48% of the normal weight group correctly classified (specificity) and 16.67% of the low weight group correctly classified (sensitivity). Positive predicted value is 49.67% and negative predicted value is 68.52%.
Finally, we calculated the marginal effects, as presented in Table 8. The coefficient (-0.2018) on private equity firms (Refinancing) is negative and statistically significant
throughout our specifications during the time series (1999–2015). This outcome is consistent and complementary to the results presented above.
31 5.11 Additional variables analysis
In this section we perform an additional analysis and results presentation on several variables mentioned as follows, structure, terms to maturity, spreads, Debt to EBITDA ratio, GDP, EURIBOR, industry allocation and country distribution.
5.11.1 Structure
The structure of debt refinancing could be arranged according to whether senior or junior lenders provide it. Senior lenders are categorized as traditional bank financing (Term Loans A and Credit Facilities) and institutional loans (Term loans B, Term loans C, or higher). These lenders are always senior to the other junior debt tranches, such as loans (Second Lien) and Mezzanine Facilities. Banks and institutional investors usually provide senior debt, while junior debt is provided by high-yield investors.
When looking at the composition of the tranche structure for private equity firms refinancing debt, the most predominant is Term Loan A (TLA) for the years between 1999 and 2005, as shown in Table 9 and Figure 9.
Prior to the crisis period, Term Loan B (TLB) became more prevalent in 2006 and 2007, with total private equity refinancing debt peaking in 2007 at 35,825 million Euro. From 2008 onwards, after the crisis period, there is a clear pattern that indicates TLB as the main tranche in the structure within the private equity firms’ refinancing debt. These results are consistent with the findings from Demiroglu and James (2015), who argued that the percentage of LBOs with Term B and Term C loans increased from 44% in 1997 to approximately 80% at the end of 2006.
32 As can be observed in Table 9 and Figure 9, the composition of the refinancing debt structure is different for non-private equity firms than for private equity firms. During the period 1999 to 2003, the predominant structure for non-private equity firms is RC (Revolving Credit Facility). During 2004, there is a maximum in TLA with 9,784 million Euro. From 2005 to 2007, there is a change in the structure trend towards TLB with the highest peak in 2007 of 10,350 million Euro. As of 2008, after the credit crunch crisis, the structure of debt refinancing for non-private equity firms is based again predominantly on RC, followed by TLA and TLB as secondarily predominant tranches.
As shown in the provided data and research carried out in this paper, private equity firms’ refinancing debt structure differs from that of non-private equity firms. This is particularly true in the period prior to the crisis, which shows the largest difference. During the years 2005, 2006 and 2007, not only is the volume of refinancing transactions higher for PE firms, but also the structure of private equity refinancing debt issuance is mainly accounted for by TLB, followed by TLC and TLA tranches. For non-private equity refinancing debt, however, the tranches TLB, TLA and RC will be the most predominant. For both private equity and non-private equity firms, Tranche TLB increases largely in the structure of debt refinancing deals in the year 2007. De Maeseneire and Brinkhuis (2012) stated that the importance of Term Loan A has diminished in favor of Term Loans B and C, implying a trend towards the institutional debt market.
As shown in Table 9 and Figure 9, the structure of the debt for private equity firms is mostly based on tranches TLB, followed by TLA, TLC and RC. Non-private equity firms on
33 the other hand, seem to be using RC more often, followed by TLA and TLB. It is also observed in the time series that private equity firms are using a pattern of consistent debt structure in accordance with economic cycles and market conditions. Demiroglu and James (2015) stated that PEGs are more active in the LBO market when credit risk spreads are low and lending standards in credit markets are lax. Non-private equity firms appear to be using a random selection of tranches in their structure, without a defined strategy pattern of organized structure in debt refinancing deals. These facts suggest private equity firms might have a better debt structure strategy than non-private equity firms.
5.11.2 Terms to maturity
As shown in the Table 10, the terms of the deal tranches for the private equity firms have longer maturity periods than non-private equity deal tranches. When looking at the total average in the time series, PE firms have an average of 5.9 and 6.1 years maturity for RC and TLA, respectively. Non-PE firms, on the other hand, have a lower average with 4.7 and 5.1 years maturity, respectively. These differences are also recognizable in the rest of the tranches as well, giving advantage to PE firms over non-PE firms in having longer terms to maturity. In these variables’ observation, once again, the private equity firms’ deals have outperformed the practices of the non-private equity firms. Not only do the deals have better conditions with longer maturity periods, but the private equity firms also show a more stable trend during the whole time series. The longer maturity term conditions, together with trend stability, ensure good performance on the loans. This performance predictability benefits PE firms in obtaining
34 a good reputation on performing loans, which will have a positive impact on future lending conditions for refinancing practices. Demiroglu and James (2015) indicate in their research that reputable PEGs pay narrower bank and institutional loan spreads, have longer loan maturities and rely more on institutional loans.
Longer terms to maturity can be observed as a trend for both PE and non-PE firms when looking at the period prior to the credit crunch crisis. As soon as the credit crunch crisis hit the economy, the terms of the loans were reduced to a shorter number of years per tranche. Table 10 shows how the average number of years to maturity declines after 2008. Plastiras and Cappelle (2011) discuss in their article that after 2007, banks and PE funds in Europe became increasingly cautious. Wright, Cressy, Wilson and Farag (2014) also found that the 2008 global financial crisis severely restricted debt availability for PE deals.
5.11.3 Spreads
PE firms have an average of lower spreads than non-PE firms, again proving their better refinancing conditions. We examine how the different choices between senior and junior refinancing debt affect the spreads of the tranches RC/TLA, TLB, TLC and Second Lien. The seniority is reflected in pricing and hence, senior debt will be cheaper than the junior debt. As shown in Table 11, PE firms’ total average spread of senior debt for tranches RC/TLA is 267 bps, but higher for non-PE firms at 280 bps. TLB tranches would be quite similar for both PE and non-PE firms with an average of 320 bps. For TLC tranches, non-PE firms have higher spreads again, with an average of 339 bps compared to the lower 320 bps from PE firms. For junior debt, Second Lien shows an average of
35 663 bps for non-PE firms versus a lower average for PE firms with 529 bps. Hence, PE firms show lower spread averages than non-PE firms. It is also plain to observe the difference in tranches when sorted by seniority between cheaper senior debt provided by banks and institutional investors (TLA, TLB and TLC) and expensive junior debt provided by high-yield lenders (Second Lien).
Furthermore, Table 11 shows that spreads for PE firms can also be higher in some instances. In 2007, PE firms report 215 bps for tranche RC/TLA versus 205 bps for non-PE firms. The cause of higher spreads for non-PE firms could be explained by the fact that spreads can increase with longer maturity terms. PE firms will have regularly longer maturity terms than non-PE firms; this could provoke the increase in the spreads for PE firms in some cases. Colla, Ippolito and Wagner (2012) expect spreads and leverage to be related, and that spreads should increase with leverage, thus reflecting the higher risk of default. As also mentioned in the literature, tranche spreads can be explained by the same factors as corporate bond yields: spreads increase with equity volatility and the market-wide cost of debt (i.e. the credit spread).
The average spreads for PE refinancing deals will look stable from 1999 to 2007, but there is an increase in 2008 mostly triggered by the start of the credit crunch crisis. From 2010 onwards, the spreads are at the highest point and they continued that same tendency for the following years until 2015. For non-PE deals the spread average looks quite different. First of all, from 1999 to 2007, the average spreads are not as stable and predictable as for the PE refinancing deals. Furthermore, in 2009, the average spread tranche RC/TLA reached the highest peak at 475 bps, a very costly refinancing approach.
36 Considering this, we analyze the spread moving average from PE refinancing deals versus non-PE refinancing deals in Figure 11. This figure shows how the moving average for the PE refinancing deals is lower during the years prior to the credit crunch crisis, 2005 to 2007, hence less expensive as a refinancing strategy.
5.11.4 Debt to EBITDA ratio
Practitioners measure firm’s leverage by implementing the ratio debt to EBITDA (earnings before interest, tax, depreciation and amortization). This ratio (Debt/EBITDA) also defined as Leverage ratio, is frequently employed when determining the debt package for companies, as the amount of debt is largely based on the cash flow the firm can generate in order to support debt repayment. In our analysis (Table 12 and Figure 12), we examine the debt to EBITDA ratio for our data sample and we observe that there is higher Public Debt/EBITDA ratio for PE deals (4.97) versus non-PE deals (3.99), as total result from the time series 1999–2015. Prior to credit crunch crisis, the Public Debt/EBITDA ratio of PE equity deals shows a rising inclination from 4.92 points in 2005 to the utmost peak of 5.51 in 2007. However, for non-private equity deals during the same period, the Public Debt/EBITDA ratio was 3.25 points in 2005, reaching 4.20 points in 2007. This difference in the Debt/EBITDA leverage ratio shows there is evidence found in this research which proves that PE firms have a higher leverage ratio than non-PE Firms during the years prior to the credit crunch crisis.
The analysis also reveals that PE firms had a higher leverage ratio during the rest of the years as well, with the exemption of years 2004 and 2010, when non-PE firms had a higher ratio. Our results are supportive with other studies revealing that high leverage
37 ratio is related to firms with good reputation, active financial management and consequently more profitable. Demiroglu and James (2015) find buyouts of reputable PEGs are more highly leveraged. Wright, Cressy, Wilson and Farag (2014) argue solvent buyouts have much higher leverage than non-buyouts. They also claim in their analysis that PE-backed buyouts already contained provisions to optimize recovery rates under insolvency and that this raised their recovery rates significantly. They carried out this research using a relative control group on PLCs, both before and after the crash. De Maeseneire and Brinkhuis (2012) elucidate that LBOs have higher leverage when debt market liquidity is stronger and a reputable PE sponsor is involved. Colla, Ippolito and Wagner (2012) find that leverage in LBOs is positively related to profitability, supporting a dynamic version of the trade-off theory of capital structure. A dynamic trade-off theory predicts that at refinancing points higher profitability reduces expected bankruptcy costs and increases leverage.
5.11.5 GDP EU28 and EURIBOR
When observing the Euribor trend along the time series (Table 13 and Figure 13) one can derive a clear inflexion point in 2009, where the Euribor rate has a huge decrease, going from 4.83% to 1,61%, probably triggered by the effects of the credit crunch crisis. This is very much in line with the GDP (Table 14 and Figure 14), which also shows a big downturn in 2009 as part of the credit crunch impact. It can be clearly observed in the analysis that both Euribor and GDP were impacted by the credit crunch.
38 The refinancing volume for PE firms was already down by the time the EURIBOR and GPD decreased, thus PE firms reacted quickly to the economic turmoil by managing their debt structure.
Non-PE refinancing deals on the other hand increased in volume in 2009, despite the downward trend in the economic cycle. By observing this pattern of behavior, we could envisage that PE refinancing practices are more in line with economic momentum. The refinancing volume comparison between PE and non-PE firms provides insights into the different strategic approach of both groups. Based on our findings, we deduce that private equity refinancing practices are more proactive and concurrent with market expectations and demands during a tumultuous economic period than non-private equity refinancing practices.
5.11.6 Industry allocation
The refinancing industry allocation for PE firms differs substantially from the industry allocation of non-PE firms. For instance, during the period prior to the credit crunch crisis from 2004 to 2007, the most representative industries of the PE firms were Services & Leasing, Retail Point, Food & Beverage, Entertainment & Leisure, Chemicals, and Building Materials. On the other hand, for non-PE firms the most representative industries were led by Metal & Mining, Telecom, Cable and lastly, Services & Leasing.
It is also interesting to observe that the main differences in industry allocation are not driven by the purpose of the deal (i.e. refinancing or not refinancing). We derived from our analysis that the main driver behind industry allocation could be related to origination of the deal, meaning whether the deal was PE related or non-PE related. It is
39 notable that industry allocation for PE deals is much more diversified than for non-PE deals during the whole time series. These factors support the theory that PE firms have different approaches to refinancing than non-PE firms, in this case PE firms are more diversified in terms of industry allocation for debt refinancing.
5.11.7 Country distribution
When looking at the refinancing deals for PE firms during the years 2004 to 2007, the most predominant countries are the UK, followed by Germany, France, The Netherlands, Spain, Denmark and Switzerland. Other European countries are also represented in the distribution, but not so predominantly as the ones mentioned.
In 2008, there is almost no volume for refinancing deals since there is a decrease in deal activity. During the years following the crisis period, the country distribution maintains the same trend.
In general, country distribution is less diversified for non-PE firms than for PE firms. The most representative countries for non-PE firms are the UK in most of the years and in the years prior to credit crisis, 2005 to 2007. Netherlands, Germany and Belgium were amongst the countries with highest volume in million Euro. After the credit crunch crisis, from 2009 onwards, Germany has the most predominant volume in million Euro. Other EU countries are also present, but less so, such as the UK, The Netherlands, Italy, France, Spain, and Belgium.
Altogether, PE firms have a more diversified representation of EU countries than non-PE firms. From the geographic allocation perspective, non-PE firms outperform the non-non-PE
40 firms on country diversification, and therefore have better protection from country risk exposure.
6. Conclusion
This paper studies the refinancing practices of private equity firms versus non-private equity firms during a time series comprehending the years 1999 to 2015. Special focus is placed on the credit crunch crisis, which hit the economy most drastically in 2008. This event will be taken as an inflexion point to determine behavioral changes in the refinancing practices of PE and non-PE firms along the time series sample data.
The results of this study suggest that private equity firms were more likely to engage in refinancing in the years prior to—but not during—the credit crunch crisis of 2008 and 2009. Refinancing during the years prior to the credit crunch crisis proved to be an optimal strategy for PE firms since lending conditions were more attractive during the boom. This finding is in line with the research from Demiroglu and James (2015) where they concluded that reputable PEGs capitalize more during favorable market conditions. Also, Colla, Ippolito and Wagner (2012) stated in their research that banks and institutional investors systematically loosen lending conditions during hot buyout markets. In this paper we provide suggestive evidence of how private equity firms changed their behavior in the years before, during and after the credit crunch crisis. PE firms were able to adapt their strategy and debt refinancing practices according to the economic demands and markets conditions of those specific periods. This study coincides with existing research in showing how PE firms excelled by taking an active
41 role in managing portfolio debt during periods of economic and financial momentum. In our empirical analysis, PE firms achieved a higher volume and a greater number of refinancing deals from 2004 to 2007, prior to the credit crunch. Our research shows that refinancing activity among PE firms during the crisis decreased drastically. This change in behavior is clear evidence of PE firms adapting their refinancing strategy to the demands of the economy and financial markets during a particular period in time. Wright, Cressy, Wilson, and Farag (2014) found that in a rapidly deteriorating financial climate, active involvement by PE firms helps portfolio companies better deal with trading difficulties in a timely way. Non-private equity firms on the other hand, do not show a clear change of behavior in their refinancing strategy under the same circumstances.
When looking at the structure of the refinancing debt, it can be concluded that private equity firms are leveraging on all the possible tranches (RC, TLA, TLB, TLC and Second Lien) more often and more evenly than non-private equity firms along the whole time series. This research provides suggestive evidence that PE firms employ a more balanced and active utilization of all possible tranches in structuring debt refinancing deals than non-PE firms. It is also concluded in this research that private equity firms are using a pattern of consistent debt structure over time and in accordance with the economic cycle and market conditions.
From the analysis of the maturity terms, we conclude that PE firms obtain longer maturity terms than non-PE firms along the whole time series.
