Positive externality of analyst coverage upon audit services: evidence from China

Abstract

Using China’s data, we find that analyst coverage helps reduce audit fees, and such effect is more pronounced for small firms, less-educated auditors, and small auditors. Our findings reveal positive externality of analyst coverage upon audit services in China, especially for those auditors weak in professional expertise, or hard to resist the pressure from their clients. Such evidence implies that the reputation protection hypothesis seems less applicable in an emerging market where the market share of big auditors is low. Our findings also provide alternative evidence justifying exchange-sponsored analyst coverage in practice intended to improve information environment of small firms.

Keywords:

• analyst coverage
• audit fees
• audit risk
• audit effort

JEL codes:

• G1
• G3
• M4

Notes

1. Lim, Monroe, and Suwardi (2013) identify a negative association between audit fees and analyst coverage in Philippines (regarded as an emerging market). But their sample size for Philippines is only 17, too small to be statistically valid.

2. http://baike.baidu.com/view/1320425.htm

3. Estimated from Gu, Li, and Yang (2013).

4. Except in the industry of mining (the average number of analyst coverage per firm in 2007 is lower than that in 2006).

5. At least one financial analyst issues earnings forecast for the firm in the year.

6. The number of financial analysts who issue earnings forecasts for the firm in the year.

7. Such relationship is from the perspective of supply side. But, if from the perspective of demand side, the existence of internal governance mechanisms may create more demand for external auditing (the assurance provided by auditors), and hence increases audit fees which firms are willing to pay (Eilifsen, Knechel, and Wallage 2001).

8. Our findings remain qualitatively the same, no matter using Big10 or Big4 to proxy for big auditors in China.

9. Salterio and Denham (1997) find that although small auditors may also have central research units or accounting consultation units, such units are typically operated as “one-person” function, and usually lack written consultation policies.

10. Untalented one’s mental cost to pass exams is presumably higher. In addition, his monetary cost to pursue a high academic degree could be also higher, due to relative difficulty for him to compete for merit-based financial assistances.

11. We admit that overconfidence may lead one to be overe-ducated.

12. http://cmis.cicpa.org.cn/cicpa2_web/public/queryofficezhpj.shtml

13. It should be noted that a listed firm’s largest controlling shareholder does not mean her ultimate controller. The Shanghai Securities Exchange has complained that some listed firms intentionally made mistakes in distinguishing such difference. http://finance.sina.com.cn/stock/y/20120810/084912813425.shtml

14. Continuous variables such as market-to-book, non-audit fee ratio, firm size, account receivables divided by total assets, inventory divided by total assets, quick ratio, return on assets, and leverage etc. are winsorized at 1% level.

15. Junior college is the minimum education level required for CPAs in China.

16. We cannot claim that, on average, every 1% increase in the number of analyst coverage reduces a firm’s audit fee by 2.2%, since we use ln (1 + the number of analyst coverage) instead of ln (the number of analyst coverage).

17. Inverse Mills Ratios are obtained by first regressing analyst coverage (dummy) on the variables as follows: firm size (natural logarithm of total assets), leverage, return on assets, market-to-book, volatility (standard deviation of the stock's monthly returns in the year), and turnover (the stock's transaction amount/number of outstanding shares). These variables are expected to affect analyst coverage (Bhushan 1989).

18. Although not shown, the coefficient of ANALYST (dummy) is still significantly negative, when IMRs are added.

19. Financial stress is defined as negative net income, negative net working capital, or negative stockholders’ equity (Carcello, Vanstraelen, and Willenborg 2009).