The Sobering Up Effect: Why Analysts Become Less Optimistic as the Release of Actual Earnings Gets Closer

Abstract

Using companies listed on the Taiwan Stock Exchange and the Taipei Exchange between 2000 and 2017, this study explores whether the sobering-up effect exists in analysts’ earnings forecasts. Further, this study also examined whether the aggregate bias (accuracy) of analysts’ annual earnings forecasts vary with analysts’ familiarity with the forecasted companies and whether analysts’ forecasting behavior is influenced by market sentiment. Our findings provide evidence for the existence of sobering-up effect on analysts’ earnings forecasts. When analysts forecast earnings for companies that they are less familiar with or the market is bearish, the analysts’ strength of sobering is more significant.

Keywords:

• The sobering up effect
• Analysts’ earnings forecast quality
• Familiarity
• Market sentiment

JEL CODES:

• G24
• G41

Notes

1. Under Article 36, paragraph 1 of the Securities and Exchange Act, public companies are required to publicly disclose their annual financial reports within three months after the close of each fiscal year. Thus, the data are presented on an annual basis beginning twelve months (a year) prior to the last day of each calendar year and continuing through to the end of March.

2. Psychological distance is a subjective experience people have when they are close to or far from things or events. Compared with physical distance, psychological distance is not an experience that individuals undergo directly, but is formed through explanation, speculation and imagination (Trope, Liberman, and Wakslak 2007; Liviatan, Trope, and Liberman 2008). Roughly, psychological distance can be classified into temporal distance, spatial distance, social distance and hypothetical distance. Psychological distance as used in this study refers to temporal distance.

3. Similar results were obtained when Treasury bills with maturity of 31-, 182-, or 364-days are used for analysis. For the entire 18-year period, five bear years (2000, 2002, 2008, 2011, and 2015) are identified; the rests are bull years.

4. The reason of measuring the potential conflicts of interest by aggregate volume of new equity issues for each year is because market-wide underwriting volume has a positive relationship with the aggregate underwriting-related compensation among sell-side analysts and may be appropriate to reflect the severity of conflicts of interest faced by all analysts.

5. The results obtained by just averaging all forecast quality values for each week are not reported but are similar to the results obtained by using the three-step procedure described in the text. 

6. Using a sample of listed companies in Taiwan, Lai, Lin, and Liu (2011) offered evidence that the analyst firm’s proprietary trading actions tend to be in the opposite direction as suggested by the affiliated analysts’ earnings forecasts, regardless of whether their forecasts are optimistic or pessimistic. They found that the conflicts of interest exist in analysts’ earnings forecasts. The conflicts of interest between brokerage firms’ proprietary trading and their affiliated analysts’ earnings forecasts jeopardize the quality of analysts’ forecasts. Whether the potential conflicts of interest between proprietary trading and the affiliated analysts’ forecasts are likely to be more severe for more reputable brokerage firms than for less reputable brokerage firms is a new issue worthy of further investigation.