This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
The performance premium found in this studies is not undisputed. Contributions such as Moskowitz and Vissing-Joergensen (2002), Gottschalg, Phalippou, and Zollo (2004), Nielsen (2006), or Conroy and Harris (2007) provide contradicting results. Based on the research approach, this study uses the studies by Ljungqvist and Richardson (2003a) and Ick (2005) to form the study’s hypotheses.
- 2.
One however has to note that recent trends in the private equity industry seem to put more weight on relatedness of the assets in a PE portfolio. In particular the increasing importance of strategic growth through “buy-and-build” strategies requires the PE firm to acquire firms that have a strong relatedness in resources as well as strong strategic fit.
- 3.
Some of the mentioned studies contain larger numbers of private equity funds. However, if consolidated to a private equity firm level, the sample size is reduced substantially.
- 4.
A more detailed overview of the characteristics of the selected samples of publicly listed corporations and private equity firms is provided in the appendix of this publication.
- 5.
SIC (Standard Industry Classification) Codes are a widely used approach to classified firms’ business activity into generally accepted categories. The SIC Code methodology will be introduced in detail in Sect. 5.2.2 “Diversification Measures” and will be basis for all further empirical analysis. An overview of all SIC codes and industry groups is furthermore provided in the appendix of this publication.
- 6.
All data items have been drawn on December 31st 2006; accounting items are therefore figures for the year 2005 or 2005/2006 for companies reporting on a different business year. All market data items are for the full year 2006.
- 7.
An overview of all SIC codes and industry groups is provided in the appendix of this publication.
- 8.
In accordance with the above argumentation, the study focuses on the status view of diversification. Changes in the degree of diversification during the 10-year analysis period are not considered relevant to the study’s research objective.
- 9.
Diversity within industry groups is measured by the four-digit diversification of industry group j DR j =å P j i ln (1/P j i ) as presented in Palepu (1985; 252). P j i is defined as the share of the segment i of industry group j in the total sales of the respective industry group j.
- 10.
Pitts and Hopkins (1982) urge researchers to be more creative in their approach to diversification measurement in order to open up new areas of research. In the author’s view, the proxy for a PE portfolio used in this study is the best outside-in assessment of the diversity of a PE portfolio available to academic research. Insider data would be more accurate but commonly does cover only individual funds rather than entire PE firms. It furthermore does not allow the coverage of a sample of 100 leading firms.
- 11.
Public corporations are required to publish ten SIC codes in their annual report, typically representing at least 10% of the firm’s consolidated revenues or assets. While only few firms apply the latter reporting guideline, all firms limit reporting to ten SIC codes. The study therefore concentrates on the rule of a maximum of ten SIC codes. Residual or non-classifiable businesses are in accordance with public corporations grouped under SIC code “9999 non-classifiable establishments”, which have been removed in comparable studies. The international work of Lins and Servaes (1999) across varying international reporting standards has particularly provided guidance for the approach chosen in this study.
- 12.
The focus of the study is to derive conclusions about how firms differ in managing diversified portfolios. The performance component is therefore more important than the aspect of valuation, which can be influenced by numerous factors as highlighted in Chap. 2.
- 13.
The findings of Cunningham’s (1973) study are basis for the calculation of risk figures in all relevant databases such as Datastream, Worldscope, or Research Insight/Compustat.
- 14.
Metrick and Yasuda’s (2007) study about the economics of private equity funds for instance also uses private equity Intelligence as primary source of private equity performance data.
- 15.
The studies by Rosenberg and Guy (1976, 1995) use technical and fundamental indicators to predict the beta of individual firms. Both studies show that the strongest predictive power of a firm’s business risk is associated with the industry mix and degree of diversification of the business portfolio as opposed to other fundamental characteristics such as capital structure, market capitalization, or earnings growth.
- 16.
The study by Kaplan and Schoar (2004) was based on a sample of venture capital and buyout firms. Their analysis therefore distinguishes different stages of investment. This distinction is not relevant for this research, which is focused on the leveraged buyout segment. Other studies such as Ick (2005) and Conroy and Harris (2007) use historical performance of private equity investments to measure risk. These studies however face substantial problems given the lack of continuous market information about private equity investments.
- 17.
Industry segments are measured on the four-digit SIC code level; industry groups are assessed on the two-digit SIC code level. The characteristics of the SIC code method are highlighted in Sect. 5.2.2 “Diversification Measures”.
- 18.
According to modern financial theory, risk adjustments should convert superior performance in selected equities back to the mean after adjusting for different risk profiles. However, data in this analysis shows that investments in high growth industries seem to outperform investments with a stronger focus on mature markets – even after adjusting for different risk profiles.
- 19.
Industry groups are measured on the two-digit SIC code level.
- 20.
Statistical comparison of the two sub-samples does not allow a rejection of the null hypothesis of equal means.
- 21.
Excluding liquidated funds guarantees that the variable “number of funds” is not a mediator variable for measuring the experience of private equity professionals. Experience will be assessed using the lifetime of a firm since the year of vintage of a firm’s first fund.
- 22.
Given the setup of former private equity studies, information about the influence of number of funds has not been subject for research. Prior studies typically evaluated private equity questions either on a transaction or a fund level instead of addressing the private equity firm as a larger entity.
- 23.
The Herfindahl index reflects low diversification in a number close to “1” and high diversification in a number close to “0”. A negative correlation of Herfindahl and performance therefore shows that increasing diversification leads to increasing performance figures.
- 24.
Portfolio size and number of funds are positively correlated with all performance indices, first year of vintage is negatively correlated with all performance measure; experience can be assessed by calculating 1-(first year of vintage), which would lead to a positive correlation of experience and performance.
- 25.
The seminal study of Berger and Ofek (1995) shows R 2 values between 2 and 8%, R 2 values in the international benchmark study by Lins and Servaes (1999) range from 3 to 14%, and the study of German corporations by Glaser and Mueller (2006) shows R 2 values between 1 and 9%.
- 26.
The sales of public corporations in the sample appear to be more concentrated than the distribution of assets across businesses. This would most likely result in substantial differences in return measures across the different businesses within a corporation. This analysis however is not part of this study.
- 27.
The ranges presented in Table 5.24 further support this result. Although the mean is significantly below the value in private equity, the range of relatedness in public corporations exceeds the maximum values in private equity settings. This indicates that firms are either highly related diversified or show a low overall level of diversification.
- 28.
The study provides the results of the 5-year performance analysis at this point as the sample of firms that contain 5-year performance figures (overall private equity n = 96) is larger than that with 10-year performance figures (overall private equity n = 80).
- 29.
Section 5.4.3 “comparison private equity vs. public equity” provides possible explanations for this empirical finding, which appears contradictory to various prior publications. However, other academic contributions such as Bettis and Mahajan (1985) or De (1992) provide consistent empirical results.
- 30.
Lossen (2006) makes similar remarks in his study about diversification in private equity funds. His data shows that specialization in particular financing stages can improve the performance of private equity investments whereas no impact of country specialization and a negative impact of industry focus is found. The study however fails to distinguish between related and unrelated diversification and consequently uses a very narrow definition of industry specialization.
- 31.
Gompers (1995) focuses his research on venture capital investments. His findings however appear applicable more generally to private equity as way of financing.
- 32.
In addition to these studies, Loos (2005) finds a concave curve in his study about performance and fund manager experience. He however provides no results about the relationship between performance and private equity firm experience but focuses on the characteristics of individuals.
- 33.
These results are based on the analysis using independent diversification ranges. The only mode of diversification with below-average performance of 10% is both high unrelated and high related diversification. This particular segment of multi-business companies however is eliminated if joint ranges are used for the analysis.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2009 Physica-Verlag Heidelberg
About this chapter
Cite this chapter
Klier, D.O. (2009). Empirical Part I: Diversification and Performance. In: Managing Diversified Portfolios. Contributions to Management Science. Physica, Heidelberg. https://doi.org/10.1007/978-3-7908-2173-4_5
Download citation
DOI: https://doi.org/10.1007/978-3-7908-2173-4_5
Published:
Publisher Name: Physica, Heidelberg
Print ISBN: 978-3-7908-2172-7
Online ISBN: 978-3-7908-2173-4
eBook Packages: Business and EconomicsBusiness and Management (R0)