What drives the profitability of Japanese multi-business corporations? A variance components analysis
Introduction
Research has found that firm heterogeneity drives macro- as well as microeconomic phenomena importantly. Labor and industrial organization (IO) economists document that industry dynamics such as output, employment, and productivity growths are critically dependent on the heterogeneity of firms populating an industry (e.g. Davis et al., 1996, Sutton, 1997). Likewise, international economists have found that export and foreign direct investment flows are driven by firms that are more productive than industry peers (Helpman et al., 2004, Bernard et al., 2007). Though the literature abounds with evidence that firms are substantially different in economic performance, the origin of this heterogeneity is only imperfectly understood. Two strands of research have contributed to filling this gap. The first group of studies rooted in the IO tradition asks why profit rates fail to converge across firms even in the long-run and looks for product and factor market impediments to the competitive convergence. Recent contributions include Villalonga (2004a) on US firms and Maruyama and Odagiri (2002) on Japanese firms.
The second strand of research looks at the phenomenon very differently and stresses the fact that today’s firms, especially large corporations, often operate in multiple industries. The profitability of such multi-business (diversified) corporations can vary because industries in which they operate are different in profit potentials, corporate-level factors are varying across firms (but constant across industries within a firm), and/or business-level (i.e. corporate-industry-specific) factors are varying from business to business even within a firm.1 Research pioneered by Schmalensee (1985) and advanced by Rumelt (1991) has estimated the relative contribution of these factors to the dispersion of business-level profit rates. A robust finding from studies based on US data is that business-level factors strongly dominate industry- and corporate-level ones in the determination of profit rates. This finding suggests that profitable multi-business corporations are typically average performers in most of their businesses. They perform excellently because they have a few “crown jewels” where they have exceptionally strong advantage unmatched by competitors. Firms that are superior performers in many businesses are rare. The finding also shows that intra-firm heterogeneity of productive efficiency can be as large as inter-firm heterogeneity despite the common use of firm-level data in many areas of empirical research to study the behavioral heterogeneity of productive organizations.
Does the profitability of multi-business corporations follow a similar dispersion pattern in Japan? Answering this question had been long infeasible because no business-level profit data was available from either firms’ financial statements or governmental statistics in Japan. However, new accounting standards introduced in the late 1990s required firms to disclose segment-level profits to investors.2 Nowadays, large financial databases such as Nikkei NEEDS contain segment-level data that is comparable to the COMPUSTAT data widely used in the US literature. In the present study, we capitalize on this development to provide the first comprehensive evidence on the profit rate dispersion of Japanese multi-business corporations.3
Our empirical method is a variance components analysis, simple but powerful technique introduced by Schmalensee (1985) to the literature. The sample is the population of publicly traded non-financial firms in 1998–2003. Estimation results reveal that the dispersion pattern of business-level profit rates in Japan is remarkably similar to that in the US That is, the variance of profit rates is mostly brought about by business-level factors, not industry- or corporate-level ones. This finding explains a few stylized facts about the restructuring of Japanese firm after the late 1990s but also poses some puzzles to researchers. We will discuss these issues after presenting empirical results.
The plan of this article is as follows. The next section briefly reviews the literature. Our empirical method is presented in Section 3. Section 4 describes data in detail. Estimation results are reported in Section 5. Section 6 is for the discussion of empirical results. The final section concludes.
Section snippets
Literature
Schmalensee (1985) pioneered research of factors shaping the distribution of profit rates at the business-level. To estimate the relative importance (explanatory power) of corporate, industry, and business-specific factors in US manufacturing, he performed a variance components analysis, simple but powerful technique to decompose data variation into components. His estimation based on the FTC’s Line of Business cross-section data reveals that industry effects affecting all firms in the industry
Model
To examine the dispersion of business-level profit rates, we assume the following model:In this specification, an annual business-level return rikt (year t return of industry i business of corporation k) is the sum of its overall mean μ and seven effects (deviations from the overall mean). There are three main effects (industry αi, corporation βk, and year γt), three interaction effects (industry-year δit, industry-corporation ϕik, and corporation-year ωkt) and
Data
We employ 6-year panel data of non-financial segments of the universe of publicly traded Japanese firms excluding financial institutions from 1998 to 2003, which is provided by the Nikkei NEEDS financial QUEST database.9 Nikkei NEEDS assigns up to three JSIC (Japan Standard Industry Classification) 4-digit codes to each segment reported in
Empirical results
Table 2 shows the empirical results. Column (1) reports the estimation result of the full model with three main and three interaction components based on the main sample excluding single-business firms. We find that business effects overwhelmingly contribute to 53.1% of the total variance, while industry effects account for 5.0%, corporation 8.6%, year 0.3%, industry-year 1.6%, corporation-year 1.0%, and residual error 30.5%. As demonstrated in Columns (2)–(4), omitting one or two effects
Discussion
In interpreting the above results, we need to keep in mind two characteristics of our data. First, our observation period coincide with 6 years when the Japanese economy was extremely stagnant at low growth rates. We speculate that this unique economic background is partly responsible for the small contributions of time-varying effects in our decomposition results.
Conclusion
This article performs a variance component analysis to provide evidence on the dispersion of business-level profit rates of Japanese multi-business corporations. We find that business-specific effects strongly dominate corporate- and industry-level effects as earlier studies found out for US firms. Though this finding explains a few stylized facts about the restructuring behavior of Japanese firms, it also poses some puzzles about multi-business corporations.
We conclude this article by pointing
Acknowledgments
We are grateful to comments from an anonymous referee, Takeo Hoshi (the editor), and workshop/conference participants at the University of Tokyo, Aoyama Gakuin University, and Academy of Management Annual Meeting (Philadelphia). Yoshitaka Fukui acknowledges financial support from KAKENHI (20653026). Remaining errors are ours.
References (30)
- et al.
Corporate diversification, performance, and restructuring in the largest Japanese manufacturers
Journal of the Japanese and International Economies
(2007) - et al.
Firms in international trade
Journal of Economic Perspective
(2007) - et al.
Does corporate strategy matter?
Strategic Management Journal
(2001) Strategy and Structure
(1962)- et al.
Job Creation and Destruction
(1996) - et al.
Export versus FDI with heterogeneous firms
American Economic Review
(2004) Corporate restructuring in Japan part I: Can M-form organization manage diverse businesses?
Japanese Economic Review
(2003)- et al.
Strategic vs. Evolutionary Management: A US–Japan Comparison of Strategy and Organization
(1985) Internal versus external market conditions and firm profitability: an exploratory model
Economic Journal
(1990)- et al.
International evidence on the value of corporate diversification
Journal of Finance
(1999)
Does country matter?
Strategic Management Journal
Does the ‘persistence of profits’ persist? A study of company profits in Japan, 1964–97
International Journal of Industrial Organization
The performance of US corporations: 1981–1994
Journal of Industrial Economics
How much does industry matter, really?
Strategic Management Journal
The persistence of shocks to profitability
Review of Economics and Statistics
Cited by (5)
Effects of industry forces, market orientation, and marketing capabilities on business performance: An empirical analysis of Japanese manufacturers from 2009 to 2011
2016, Journal of Business ResearchCitation Excerpt :Previous empirical studies examine the question of principal sources of performance using secondary or primary survey data. Studies relying on secondary data (e.g., Fukui & Ushijima, 2011; Makino et al., 2004; Mauri & Michaels, 1998; McGahan & Porter, 1997, 2002; Roquebert, Phillips, & Westfall, 1996; Rumelt, 1991) decompose the variance in performance into components associated with external and internal factors. The findings indicate that internal factors (e.g., corporate and business effects) are more important than external factors (e.g., industry effects) in explaining the variance in performance.
Variance decomposition of firm performance: past, present and future
2020, Management Research ReviewHow much do industry, corporation, and business matter, really? A meta-analysis
2017, Strategy ScienceEmpirical Analysis of Host-Country Effects in the International Construction Market: An Industry-Level Approach
2016, Journal of Construction Engineering and ManagementFinancial management practices: An empirical study of Indian corporates
2013, Financial Management Practices: An Empirical Study of Indian Corporates