Michael Firth

Institutional monitoring and opportunistic earnings management

Abstract Investment institutions with substantial shareholdings in a firm have the resources and incentives to monitor and influence management decisions. Whether the institutions actually monitor and exert pressure on managers is an empirical question. Previous studies have reported mixed results on this question. We examine whether large institutional shareholdings in a firm deter earnings management by its managers when those executives otherwise have incentives to increase or decrease reported profits. Using discretionary accounting accruals as the measure of earnings management, we find that the presence of large institutional shareholdings inhibit managers from increasing or decreasing reported profits towards the managers desired level or range of profits. The evidence is consistent with institutional investors monitoring and constraining the self-serving behavior of corporate managers.

Stock market linkages: Evidence from Latin America

Abstract This study investigates the dynamic interdependence of the major stock markets in Latin America. Using data from 1995 to 2000, we examine the stock market indexes of Argentina, Brazil, Chile, Colombia, Mexico and Venezuela. The index level series are non-stationary and so we employ cointegration analysis and error correction vector autoregressions (VAR) techniques to model the interdependencies. We find that there is one cointegrating vector which appears to explain the dependencies in prices. The results are robust to sensitivity tests based on translating indexes to US dollars (i.e., a common currency for all the markets) and to partitioning the sample into periods before and after the Asian and Russian financial crises of 1997 and 1998, respectively. Our results suggest that the potential for diversifying risk by investing in different Latin American markets is limited.

Auditor-Provided Consultancy Services and their Associations with Audit Fees and Audit Opinions

This study examines the relationships between non-audit services fees (consultancy fees) paid to auditors and (1) audit fees, and (2) the occurrence of qualified audit opinions. The positive association between consultancy fees and audit fees is shown to be due to certain company specific events that generate a demand for consultancy services as well as requiring additional audit effort. Identified company specific events are mergers and acquisitions, new share issues, new accounting and information systems, new CEOs, and corporate restructurings. When these events are absent, there is no statistically significant relationship between audit fees and consultancy fees after controlling for company size. Companies that have relatively high consultancy fees are more likely to receive a clean audit opinion. This may be due to the non-audit work clearing up problem areas at the client company or it may be due to high consultancy fees impairing auditor independence. With the available data it is not possible to distinguish between these two reasons. Copyright Blackwell Publishers Ltd 2002.

Auditor conservatism and reported earnings

Abstract Conservatism is an underlying concept of financial accounting. We argue that the auditor forces conservatism on client companies and that the amount of conservatism depends on the economic performance of the company and on the type of audit firm. In particular, we contend that Big Six audit clients use more conservative accounting than non-Big Six audit clients when the clients are performing poorly (as reflected in stock prices). We attribute the more conservative accounting used by Big Six audit clients to the influence of the audit firm. By regressing excess earnings to price ratios on excess stock returns and other variables, we find evidence consistent with our hypothesis that Big Six auditors influence their clients to adopt more conservative accounting than non-Big Six auditors only when the clients financial performance is worse than expected.

The Accuracy of Profits Forecasts in Initial Public Offering Prospectuses

Abstract The paper examines the accuracy of profits forecasts contained in prospectuses of companies newly listing on the New Zealand Stock Exchange. Accuracy is measured by forecast errors, absolute forecast errors and squared forecast errors. The level of forecast accuracy appears to be poor in comparison to studies conducted elsewhere. An attempt to model the errors as a function of firm specific characteristics and of Big Eight/non-Big Eight auditor choice proved to be of little value. The forecasting errors were found to be unrelated to the stock price premium upon listing.

IPO profit forecasts and their role in signalling firm value and explaining post-listing returns

The study examines the role of profit forecasts published in the prospectuses of initial public offerings. While it is very rare to see such forecasts in American IPOs, the practice is pervasive in some other countries. For example, Singaporean new issue prospectuses have an established history of providing point estimates of post-listing earnings and, since 1993, profit forecasts have been mandated by the Stock Exchange. Earnings forecasts are hypothesized to be a signal of company value and the accuracy of the forecasts are hypothesized to explain post-listing returns. Using data from new issues in Singapore during the period 1979-1992, various tests are employed to examine the relationships between profit forecasts appearing in prospectuses and market valuations. The results indicate that the earnings forecasts are a major signal of IPO value and that they are more important than other signalling mechanisms such as the retained share ownership of the entrepreneurs. The accuracies of the forecasts are positively related to post-listing stock returns. Stock returns in the three years after listing are close to zero and this contrasts with results from many other countries that find significant and substantial negative returns. The three-year post-listing stock returns are positively associated with IPO profitability.

The Information Content of Concurrently Announced Earnings, Cash Dividends, and Stock Dividends: An Investigation of the Chinese Stock Market

The firms listed on China’s stock market are less than ten years old and to date there has been relatively little research on the usefulness of their accounting disclosures for investors. This study focuses on the information content of annual earnings and dividend announcements made by listed Chinese companies. Earnings, cash dividends, and stock dividends are announced concurrently in China and so this allows for tests of their information usefulness and of the interactions between the three signals. Based on a data set of up to 1,232 announcements, we find that unexpected earnings, proxied by earnings changes, are positively related to abnormal returns. Thus, earnings are used by investors in setting market prices. Stock dividends corroborate or attenuate the earnings signal. If the sign of the unexpected stock dividend (increase, decrease) is the same as the sign of the unexpected earnings, then the earnings signal is stronger. If the signs are opposite, the earnings signal is weaker. Unexpected cash dividends have little impact on the earnings signal. Stock dividends per se have a small association with stock returns. In contrast, cash dividends have no discernible association with stock returns and this is consistent with dividend irrelevance arguments. Our results are robust across a number of sensitivity tests.

A multivariate analysis of the determinants of Moody's bank financial strength ratings

Abstract In 1995 Moody’s Investors Services inaugurated a new rating service, bank financial strength ratings (BFSRs), that assesses the safety and soundness of banks in over 50 countries. Our study sets out to do some preliminary investigations of this new type of credit rating. We develop logistic regression models to help explain or predict BFSRs. Using bank-specific accounting and financial data we are able to correctly classify or predict BFSRs. These fundamental variables cover the dimensions of risk, loan provision ratios, and profitability. Of the three, loan provisions is the most important factor, followed by risk, and then profitability. Country risk ratings do not appear to be significant explanators of BFSRs. We also find that traditional debt ratings accurately classify BFSRs and this raises the question of whether BFSRs add incremental information. The paper also highlights future directions for our research. One such area is to examine how well BFSRs predict banking crises such as the credit problems currently affecting Asia and Latin America.

The determinants of top management pay

Agency theory argues that companies need to structure their top management pay so as to attract, retain, motivate, and reward senior executives. It is implicit in this literature that managers should be rewarded for performance and that company size should not be a significant determinant of compensation. Empirical evidence in many countries has concluded, however, that size is a major determinant of management remuneration and the pay-for-performance link is very weak. This study examines the determinants of senior executives remuneration and bonus payments in Hong Kong companies using recently available data. We examine both the level of pay and changes in pay. Corporate size is found to be a major explanator of remuneration levels and of changes in the pay of the CEO and executive directors. Accounting profitability is also a significant explanator of compensation. Performance, as measured by stock returns, has little or no statistically significant relationship with pay; in fact, some of the results show negative relationships. Some share ownership characteristics have influences on the levels of remuneration. In particular, share ownership by directors and share ownership by institutional investors moderate the compensation levels. In contrast, corporate governance variables have little association with change in pay. Overall, the results imply agency arguments that advocate pay-for-performance compensation schemes are not major factors in setting top management remuneration in Hong Kong.

Asset pricing in segmented capital markets: Preliminary evidence from China-domiciled companies

Abstract A number of Chinese companies have issued shares to investors within China (A shares) and issued shares to foreign investors (B, H, and N shares). All these shares have equal rights although A shares can only be sold to, and traded among, PRC citizens and B, H, and N shares can only be issued to, and traded among, foreign investors. The paper examines the impact of the initial listing of B-share issues on the prices of already listed A shares. Our analyses test the joint characteristics of market segmentation and seasoned equity offerings. We find that the abnormal returns on A-share companies that also offer B shares are significantly negative, a result consistent with the hypothesis that the demand curve for equity shares is downward sloping. Interestingly, these negative abnormal returns can be explained by our proxies for the investor recognition theory of Merton (1987) and the liquidity theory of Amihud and Mendelson (1986) .