Do board characteristics affect corporate performance? Firm-level evidence for India

Abstract

The study examines the association between financial performance and boards of non-financial firms. Using data on 127 listed manufacturing firms in India for 2003 the findings indicate that, after controlling for various firm-specific factors, larger boards tend to have a dampening influence on firm performance, judged in terms of either accounting or market-based measures of performance. In terms of policy implications, the analysis suggests that compensation of the CEO has a significant effect on the performance of the firm.

Acknowledgements

The views expressed and the approach pursued in the paper expresses the personal opinion of the author.

Notes

¹ Hall and Leibman (1997), for instance, report a compensation differential of nearly US$4 million for firm exhibiting below-average performance vis-à-vis those exhibiting above-average performance.

² Accordingly, these institutions have implemented new guidelines for appointment of nominee directors. According to this criterion, nominees would be appointed in companies where the combined (debt plus equity) exposure of the concerned institution exceeds Rs. 500 million (US$1≈ Rs. 45), or their shareholding is above 26%.

³ In the event that the post of Chairman and Managing Director are held by two separate persons, typically the Chairman is the topmost executive of the company. The terms Chairman and Chief Executive Officer (CEO) are used interchangeably.

⁴ There were also several recommendatory provisions, the salient among them being the number of companies in which a person can hold directorship was reduced from 20 to 10.

⁵ The Stock Exchange, Mumbai classifies firms according to several strata of which the blue-chip companies are the ‘A’ scrips. These scrips are ones in which carry forward or badla is permitted and account for over 60% of the market capitalization.

⁶ These include, in alphabetical order: Automobiles, Cement, Chemicals, Computer, Drugs and Pharmaceuticals, Electronics, Heavy industries, Textile and Others.

⁷ The financial year for corporates extends from the first day of April of a particular year to the last day of March of the subsequent year. Thus, 2003 corresponds to the financial period 2002 to 2003 (April to March).

⁸ The numerator in all the three performance-based measures, Return on Asset (RoA), Return on Sales (RoS) and Return on Equity (RoE) is pre-tax profits. The denominator in these three variables are, respectively, total assets, sales and total equity. The advantage of employing an aggregative measure like PERF is that it captures major aspects of financial performance like sales and return to shareholders, in addition to profits. The RoS measure has been employed by Khanna and Palepu (2000).

⁹ The correlation coefficient between PERF and AdjQ works out to 0.36. This relatively small correlation coefficient makes it necessary to analyze performance and board size nexus for both kinds of indicators.

¹⁰ A public company, on the other hand, is one (a) which is not a private company; (b) has minimum paid-up capital of Rs. 0.5 million. A private company is one (a) is not a public company, (b) has minimum paid up capital of Rs. 0.1 million; and (c) prohibits any invitation to the public to subscribe for any shares in, or debentures of, the company. In the present study, ‘public’ and ‘private’ firms refer to Government undertakings and Indian private entities, respectively. The Prowess database explicitly provides information as to whether a company is a public or a private entity.

¹¹ Given the nine industry groups, the dummy variable for ‘Others’ is omitted, so that the response of the other industry groups is relative to this omitted category (see Footnote 6).

¹² This variable is defined as follows. If the coefficient of variation of sales (a proxy for demand uncertainty) or the return on assets (a proxy for profit uncertainty) exceeds the mean value of these variables, the environment is defined to be uncertain (DYN = 1), otherwise, zero.

¹³ Bhagat and Black (2001) report a negative relation between board size and Tobins Q in some (but not all) specifications.