Good News Firms Research Articles

On the conditional conservatism measure: A robust estimation approach

AbstractRecent research, due to Patatoukas and Thomas (2011) and Ball, Kothari, and Nikolaev (2013), focuses on Basu's (1997) conditional conservatism measure and the existence of a denominator effect – whether the difference between the earnings‐return coefficients of bad and good news firms (‘the Basu coefficient’) is only due to the beginning‐of‐year price deflator. We address this issue head‐on by applying the Theil‐Sen (TS) estimation method, which obtains the same coefficient estimate regardless of the chosen deflator and is robust to outliers. Results show the following: (i) the Basu coefficient remains positive using TS; (ii) the Basu coefficients using TS are similar to those using OLS without scaling but much smaller than shown by scaled OLS; (iii) the scaled OLS estimates appear to be influenced by a few outliers; and (iv) OLS estimates are more volatile due to estimation error. In sum, the denominator effect does not overturn Basu's hypothesis but the magnitude and variation of the Basu coefficient is much smaller than traditional results show.

On the Conditional Conservatism Measure: A Robust Estimation Approach

Recent research, due to Patatoukas and Thomas (2011) and Ball et al. (2013), focuses on Basu’s (1997) conditional conservatism measure and the existence of a denominator effect — whether the difference between the earnings-return coefficients of bad and good news firms (“the Basu coefficient”) is only due to the beginning-of-year price deflator. We address this issue head-on by applying the Theil-Sen (TS) estimation method which does not depend on the scaling of variables; the same coefficient estimates are obtained whether or not the regression equation is scaled by beginning-of-year price. Moreover, the TS estimates are intrinsically insensitive to outliers. Results show the following: (i) the Basu coefficient remains positive using TS; (ii) the Basu coefficient using TS are similar to those using OLS without scaling but much smaller than what scaled OLS show; (iii) the scaled OLS estimates appear to be influenced by a few outlier firms with negative returns and material losses; and (iv) OLS estimates are more volatile due to estimation error. Also, placebo tests suggested in Patatoukas and Thomas (2011) do not falsify our findings. In sum, the denominator effect does not overturn Basu’s hypothesis but the magnitude and variation of the Basu coefficient is much smaller than what traditional results show.

Earnings Conservatism in an Emerging Capital Market: Evidence from Iran

In this paper we investigate whether earnings are conservative in financial reporting of firms listed in Tehran Stock Exchange (TSE)? Using the asymmetric timeliness of earnings, we measure earnings conservatism for 495 firm-years during 2001-2009, we hypothesize that earnings in all periods for sample firms are conservative. Consistent with Basu (1997), we classify the sample firms in two groups: firms with good news and firms with bad news. Our measure of conservatism is the difference between the slope coefficients for bad and good news firms. In some periods, our findings support the existence of conservatism in financial reporting of these firms; however in some periods our measure is negative which means that earnings are not conservative in these periods.

Are good-news firms riskier than bad-news firms?

This paper examines the relative risk of good-news firms, i.e., those with high standardized unexpected earnings (SUE), and bad-news (low SUE) firms using a stochastic discount factor approach. We find that a stochastic discount factor constructed from a set of basis assets helps explain post-earnings-announcement drift (PEAD). The risk exposures on the pricing kernel increase monotonically from the lowest to highest SUE sorted portfolios. Specifically, good-news firms always have higher risk exposures than bad-news firms in both 10 SUE sorted portfolios and 25 size and SUE sorted portfolios. However, the estimated expected risk premium is too small to explain the observed magnitude of returns on the PEAD strategy. Our risk adjustment can explain only about one-fourth of the total magnitude of the average realized return to the PEAD strategy. As a result, the average risk-adjusted returns of earnings momentum strategies are mostly positive and significant. Overall, our results support the view that at least some portion of the returns to the earnings momentum strategies examined represent compensation for bearing increased risk.

The impact of New Zealand's disclosure reform on differential managerial disclosure behaviour for good news versus bad news firms

PurposeThe purpose of this paper is to investigate the impact of the introduction of New Zealand's statutory‐backed continuous disclosure regime enacted in December 2002 on the differential disclosure behaviour of New Zealand firms with good and bad earnings news.Design/methodology/approachThis paper examines the level of information disclosure, analyst forecast error and forecast dispersion, abnormal returns and abnormal volumes for firms with good and bad news earnings announcements in a sample period surrounding reforms to New Zealand's continuous disclosure regime.FindingsThe authors find evidence that the pre‐announcement information flow was poorer prior to the reform for bad news firms compared to good news firms, in terms of greater analysts' forecast dispersion and a larger abnormal price reaction to the actual earnings announcement. Second, the reform reduced the asymmetry of information flow between good and bad news firms, with the differences in analysts' forecast dispersion and abnormal price reaction dissipating after the reform.Research limitations/implicationsThe findings suggest that the reforms to New Zealand's continuous disclosure regime have reduced managers' propensity to withhold bad news and improved the quality of information provided to investors by firms with bad earnings news.Originality/valueThis study improves our understanding of the impact of disclosure reform on the behaviour of managers in a market with relatively low liquidity and less litigation risk in comparison to larger and more developed markets.

Information Asymmetry and Momentum Anomalies

In this paper, we construct an information asymmetry factor (VECINF) based on the price discovery of large trades. VECINF is significantly negatively correlated with market excess return, indicating that market-wide information asymmetry is lower in bull markets, which is consistent with the view that more uninformed investors are attracted into stock markets when the markets offer high returns. In addition, VECINF has the most significant pricing effect among the considered risk factors (MKT, SMB, HML, and UMD), which suggests that ignoring the risk of information asymmetry may give rise to false discoveries of anomalies. As a case in point, we show that momentum anomalies disappear once we control for the risk of information asymmetry. This is because there are fewer informed traders, and hence lower risk of information asymmetry, in bad news firms (past losers or low earning surprise firms) than in good news firms (past winners or high earning surprise firms). The larger cost and risk of arbitrage in taking short positions makes bad news firms less attractive to informed traders. Consistent with this explanation, we find that the loading on VECINF is lower in bad news firms than in good news firms; and is lower only in the holding periods when momentum exists. This difference in loadings increases significantly with idiosyncratic volatility, and this explains why momentum is stronger in firms with large idiosyncratic volatility. Regardless of the level of idiosyncratic volatility, the significantly positive Fama-French factors risk-adjusted returns of zero investment momentum portfolios are no longer significant once we include VECINF as an additional factor for risk adjustment.

Information Risk and Momentum Anomalies

In this paper, we construct an information asymmetry factor (ECINF) based on the price discovery of large trades. ECINF is significantly negatively correlated with market excess return, indicating that market-wide information asymmetry is lower in bull markets, which is consistent with the view that more uninformed investors are attracted into stock markets when the markets offer high returns. ECINF is not only a priced risk factor, but the most significant factor in our asset pricing tests, which suggests that ignoring the risk of information asymmetry may give rise to false discoveries of anomalies. As a case in point, we show that momentum anomalies disappear once we control for the risk of information asymmetry. This is because there are fewer informed traders, and hence lower risk of information asymmetry, in bad news firms (past losers or low earning surprise firms) than in good news firms (past winners or high earning surprise firms). The larger cost and risk of arbitrage in taking short positions makes bad news firms less attractive to informed traders. Consistent with this explanation, we find that the loading on ECINF is lower in bad news firms than in good news firms; and is lower only in the holding periods when momentum exists. This difference in loadings increases significantly with idiosyncratic volatility, and this explains why momentum is stronger in firms with large idiosyncratic volatility. Regardless of the level of idiosyncratic volatility, the significantly positive Fama-French factors risk-adjusted returns of zero investment momentum portfolios are no longer significant once we include ECINF as an additional factor for risk adjustment.

Investor Sentiment, Post-Earnings Announcement Drift, and Accruals

There is growing evidence in the finance literature that investor sentiment affects stock prices. We examine whether stock price reactions to earnings surprises and accruals vary systematically with the level of investor sentiment. Using quarterly drift tests and monthly trading strategy (calendar time) tests, we find evidence that holding extreme good news firms following pessimistic sentiment periods earns significantly higher abnormal returns than holding extreme good news firms following optimistic sentiment periods. Similarly, our results suggest that holding low accrual firms following pessimistic sentiment periods earns significantly higher abnormal returns than holding low accrual firms following optimistic sentiment periods. We also document that abnormal returns in the short-window around preliminary earnings announcements for extreme good news firms are significantly higher during periods of low sentiment than during periods of high sentiment. Overall, our results indicate that investor sentiment influences the source of excess returns from earnings-based trading strategies.

How Do Firms Manage Their Earnings Forecast Strategy? A New Zealand Study

In contrast to the trend of research investigating why firms decide to release earnings forecasts to pre-empt any expected change in earnings, our study investigates how firms manage their earnings forecast strategy once they have decided to release earnings forecasts. Using a sample of 350 NZX-listed firm years with balance date ending from 31 January 1999 to 31 December 2005 for 94 companies across the statutory-backed continuous disclosure regime, we document that firms are more likely to adopt an earnings forecast portfolio approach in the post statutory sanctions period, particularly for the group of firms expecting favourable earnings change. We also document that these good news firms have a higher propensity to gradually update the market with good news earnings forecasts while those with bad news are more likely to immediately correct current market expectations of earnings or prior earnings forecast errors. These findings indicate that firms expecting better earnings performance are more conservative in their earnings forecasting compared to those expecting worse earnings performance. Even though this asymmetrical treatment between good news and bad news might not meet the objective of corporate regulators in promoting a continuously updated market, the fact that good news firms indeed increase their earnings forecast practices in the post sanction period does indicate an improvement in the information flow to the capital market.

Market Reaction to Earnings Surprise Warnings: The Incremental Effect of Shareholder Litigation Risk on the Warning Effect

We examine the incremental effect of shareholder litigation risk on market reaction to earnings surprise warnings, that is, the warning effect. Prior research examines earnings warnings by firms reporting large earnings news (at least 1% of share price), and finds a negative warning effect for bad news firms but no warning effect for good news firms. We find similar results for a larger sample of firms. In addition, we find negative and positive warning effects, respectively, for firms reporting small bad and good earnings news, suggesting that the insignificant warning effect for good news firms is restricted to large earnings news. More importantly, we find that litigation risk magnifies the warning effect—for bad news firms, the warning effect is more negative for high-litigation-risk firms than for low-litigation-risk firms, but for good news firms, the warning effect is more positive for high-litigation-risk firms than for low-litigation-risk firms.

Voluntary Disclosure and the Cost of Capital: Evidence from Management Earnings Forecasts

This paper examines the effect of management earnings forecasts on the cost of capital. We find that forecasters of bad news experience a significant increase in the cost of capital in the month after their disclosure. Conversely, the cost of capital for good news forecasters declines mildly in the same period. We also indicate that bad news forecasts are more informative by comparing the absolute value changes in the cost of capital between bad and good news firms. In addition, we find no significant relation between forecast precision and the cost of capital. Finally, we examine the combined effects of the management earnings forecasts and the earnings announcement on the changes in the cost of capital. Consistent with prior studies, we show that the effect of the subsequent earnings announcement is preempted by the management forecasts for bad news firms, but not for good news firms.

Do Analyst Earnings Forecasts Allow for Accounting Conservatism?

Recent studies show that accounting earnings are conservative, i.e., earnings tend to reflect bad news (negative stock returns) on a timelier basis than good news (positive stock returns) (Basu, 1997); moreover, the degree of conservatism is negatively associated with the price-to-book (P/B) ratio (Pae et al., 2003). If analysts correctly allowed for conservatism and its documented association with the P/B ratio, differences in earnings conservatism would be unassociated with analysts' forecast error. In contrast, we find that average yearend forecast error differs between good news and bad news firms, and between high and low P/B firms. We conclude that analysts' earnings forecasts do not fully incorporate the implications of earnings conservatism. We also find that forecast dispersion is greater for bad news than good news firms, and greater for low than high P/B firms, consistent with the hypothesis that accruals used to accelerate the recognition of bad news spawn disagreement about forthcoming earnings.

The Post-Earnings-Announcement Drift and Liquidity Risk

This paper investigates the relation between the post-earnings-announcement drift anomaly and liquidity. First, we find that, on average, bad-news firms (low standardized unexpected earnings (SUE)) are less liquid than good-news firms (high SUE), reflecting more information asymmetry and/or uncertainty among bad-news firms. Yet, we argue that this liquidity spread is less likely to explain the drift. Second, the returns of SUE-sorted portfolios are sensitive to fluctuations in market-wide liquidity. We find that systematic liquidity risk is an important determinant in explaining the cross-sectional variation of expected returns among SUE-sorted portfolios. This implies that a substantial part of the post-earnings-announcement drift anomaly can be viewed as compensation for risk associated with shocks to the information environment in the economy. Therefore, the evidence suggests that the previously reported anomalous returns are associated with model misspecification and/or hidden transaction costs.

The Voluntary Inclusion of Forecasts in the MD&A Section of Annual Reports*

Abstract. In this study, we appeal to theories advanced by Darrough and Stoughton (1990) to enhance our understanding of why some firms may voluntarily include directional forecasts in their annual reports while others do not. The data are consistent with their predictions that a firm's disclosure policy reflects its concern for both financial market valuation and product market competition. We find that for “good news firms, the probability of forecasting is increasing in the financing requirements but decreasing in the threat of competitor entry. The converse holds for “bad news” firms. These results lend further empirical support to the observation that the familiar good news hypothesis tested in the management earnings forecast literature offers only a partial explanation for the decision to forecast. Interestingly, however, even after controlling for financial and product market considerations, an overall voluntary disclosure bias still exists in the data. The data also provide support for the OSC's concern about a voluntary disclosure bias. Only 17.5 percent of our sample forecasts represent revisions downward relative to the previous year's results. However, in contrast to the OSC's concern about a general lack of forward‐looking disclosures in annual reports, 35.9 percent of our sample firms include directional forecasts in their MD&A or elsewhere in the annual report.

Callable Bonds: A Risk-Reducing Signalling Mechanism-A Comment

THE EXISTENCE OF CALLABLE bonds has been somewhat of a puzzle. The call feature provides firms with a valuable interest rate option, but the value of this option should be exactly offset by the creditor's demand for a higher interest rate as compensation. One explanation for the call provision from Myers [4], Bodie and Taggart [3], and Barnea, Haugen, and Senbet [2] is that callable bonds may be used to reduce some agency costs. Another explanation from Robbins and Schatzberg [5] is that callable bonds can signal a firm's better prospects in the presence of asymmetric information. One possible problem with the agency cost and signalling explanations of callable bonds is that short-term debt may serve the same functions in all four studies. Indeed, short-term debt is equivalent to long-term debt in the theoretical sections of the agency cost studies. However, Robbins and Schatzberg (hereafter RS) suggest that short-term debt fails to provide the risk-sharing opportunities inherent in callable bonds. Thus, they suggest that a signalling explanation provides a unique role for callable bonds. One limitation of RS's results is that they are based on an example that makes a series of strong assumptions about a two-stage lottery faced by and firms. In order for their model to have general applicability, it would be desirable to demonstrate that their results hold under more general assumptions. The results of this study suggest that their findings cannot be generalized without further assumptions. Callable bonds do not yield a separating equilibrium result when only a minor change is made in two of the lottery probabilities, suggesting that their results are far from robust to minor changes in assumptions. Moreover, callable bonds may fail to generate a separating equilibrium even for some good-news firms that conform to RS's original model because bad-news firms that do not meet their assumptions may attempt to mimic good-news firms that do. The risk-sharing result is also very sensitive to the specific assumptions made by Robbins and Schatzberg; specifically, callable bonds increase the risk borne by management if one change is made in the numeric values of one of the lottery payouts. The paper begins with a review of the relevant portions of Robbins and Schatzberg's model, Section II provides a counterexample in which callable bonds fail to provide a separating equilibrium, Section III provides a counterexample in which callable bonds fail to provide risk-sharing debt, and the paper concludes with a summary.