Customer concentration and over-investment

ABSTRACT

This paper investigates the impact of customer concentration on a supplier firm’s over-investment behaviour. Based on a large sample analysis on the US market, we find that a supplier firm with higher degree of customer concentration is likely to over-invest beyond the optimal investment level determined by investment opportunities. We also reveal that the positive association between customer concentration and over-investment weakens when the largest customer is a government entity, when the duration of supplier–customer relationship extends, and when the supplier firm captures larger market shares in its industry.

KEYWORDS:

• Customer concentration
• over investment
• powerful customer

JEL CLASSIFICATION:

• G31
• L14

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ We define ‘powerful customer’ as those whose purchases account for a significant portion of a supplier firm’s total sales because this trading position grants them greater bargaining power (see Hui, Klasa, and Yeung 2012). In this sense, ‘powerful customer’ includes but not limits to the widely cited ‘major customer’ whose purchases account for more than 10% of a supplier firm’s total sales. We use ‘customer concentration’ to measure the existence and importance of powerful customers: the higher degree of customer concentration indicates the supplier firm relies on relatively small group of customers who are likely to gain greater bargaining power.

² Major customers are defined as those customers whose purchases account for more than 10% of a supplier firm’s total sales.

³ Due to data availability, this paper, in line with most of related literature, focuses on the impact of customers on suppliers.

⁴ For example, Apple and Walmart have been widely cited examples in pushing suppliers to invest excessively in the product lines and research and development.

⁵ There are seven customer types in Compustat Customer Segment: COMPANY, GOVDOM (domestic government), GOVLOC (local government), GOVSTATE (state government), GOVFRN (foreign government), GEOREG (e.g. Africa, Asia) and MARKET (industries). Since we focus on firm-level data, we exclude the GEOREG and MARKET customer types.

⁶ From year 1976, Statement of Financial Accounting Standards No.14 (SFAS 14) requires that a supplier firm should disclose the information of those major customers whose purchases account for more than 10% of its total sales. Since year 1997, FAS 131 has replaced SFAS 14 but the requirement continues under SEC Regulation S-K Item 101, although it no longer requests the supplier firm to disclose the names of major customers. In addition to those major customers, firms also voluntarily disclose customers whose purchases account for less than 10% of their total sales. Our sample covers all the data with customer information.

⁷ We follow Patatoukas (2012) to rank the firms annually based on $C{C}_{i,t}$ into deciles and construct the ranked customer concentration ($RC{C}_{i,t}$) by replacing the raw value of $C{C}_{i,t}$ with the corresponding deciles. We then re-estimate our hypothesis and find consistent results (untabulated).

⁸ Following Richardson (2006), we also define over-investment when the residual from Equation (1)(1) $IN{V}_{i,t}={\alpha }_0+{\alpha }_{1}SG{R}_{i,t-1}+{\alpha }_{2}SG{R}_{i,t-1}\ast NE{G}_{i,t-1}\text{break}+{\alpha }_{3}NE{G}_{i,t-1}+{\epsilon }_{i,t}$(1) is positive. Results (untabulated) are largely consistent.

⁹ For example, if firm A starts to list firm B as a customer in the year 1985, we assume the length of this relationship is one if the observation year is 1985, two if 1986, three if 1987 and so on. Notably, the relationship duration needs to be continuous. For example, if firm A has firm B as a customer for 1985 and 1987, but not for 1986, then the length of this relationship is one if the observation year is 1985, zero if 1986 and one if 1987.

¹⁰ For example, suppose firm A has two identified customers: firm B and firm C, with the supplier-customer relationship duration of four years and five years, respectively. If firm B and firm C accounts for 40% and 60% of firm A’s total sales respectively, the weighted duration of the relationship between firm A and its identified customer base (firm B and firm C) is 4.6 years (4*40%+5*60%).

¹¹ We also test alternative MS measures by sorting the sample market share into quartiles and deciles. Results are similar.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [71702114]; Beijing Social Science Fund Project [18GLC076]; Scientific Research Plan of Beijing Municipal Education Commission [SM202010038015]; Macau University of Science and Technology under Faculty Research Grants [0447].