What determines the value of recommendation change? A preliminary analysis

ABSTRACT

This article determines the condition under which recommendation changes relative to consensus recommendation and recommendation changes relative to analyst’s previous recommendation are valuable. We show that recommendation upgrades (downgrades) relative to consensus recommendation are followed by significantly positive (negative) short-term returns whenever consensus recommendation represents convergence of analysts’ opinions. We also show that as the standard deviation associated with consensus recommendation increase, the value of recommendation changes reduce significantly. This article also shows that recommendation upgrades (downgrades) relative to analyst’s previous recommendation are followed by significantly positive (negative) short-term returns whenever interval between two consecutive recommendations is relatively short. We also show that, as the interval between two consecutive recommendations increase, the value of recommendation changes reduce significantly. Our results are robust across various subsamples based on size and region.

KEYWORDS:

• Value
• analyst recommendation
• recommendation change
• consensus recommendation
• previous recommendation

JEL CLASSIFICATION:

• G15
• G20
• G24

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Consensus recommendation is, generally, defined as the average of outstanding recommendations.

² For the example presented, one third of analysts (one out of three analysts) have already issued Buy recommendations.

³ In addition to reporting considerably long average duration between two consecutive recommendations, Wu and Lin (2014) also report high standard deviation for these averages. They report standard deviation of 139.96 and 120.20 for favourable and unfavourable reiterations, respectively, while standard deviation for upgrades and downgrades is reported to be 139.66 and 138.50, respectively.

⁴ There is ample evidence that suggests that analysts revise their forecasts much more frequently than recommendations. For instance, Myring and Wrege (2009) document that analysts revise their forecasts around four times a year.

⁵ Consistent with the I/B/E/S convention, we assign the rating of 1 to Strong Buy, 2 to Buy, 3 to Hold, 4 to Underperform, and 5 to Sell recommendations.

⁶ We would like to mention that descriptive statistics provided in Table 2 do not include those upgrades or downgrades that give ambiguous investment advice. For instance, following upgrades are not considered as upgrades in Table 2: (1) Upgrade from Sell to Underperform, (2) Upgrade from Sell to Hold, and (3) Upgrade from Underperform to Hold. Similarly, following downgrades are not considered as downgrades in Table 2: (1) Downgrade from Strong Buy to Hold, (2) Downgrade from Strong Buy to Buy, and (3) Downgrade from Buy to Hold.

⁷ Above footnote also holds for Table 3.