Do Analysts Anchor Their Recommendation Revisions to Valuation Multiples?

Abstract

Purpose of the study

The key objective is to study whether analysts partly anchor their recommendation revisions to the 52-week high and low stock prices as well as to the median forward-looking consensus valuation multiples, particularly to the 2-year high and low P/E ratios. The potential anchoring effects of valuation multiples have previously been left without academic attention even though analysts often justify their recommendation revisions by stating P/E and PEG ratios, indicating that they may use these valuation multiples to reach their recommendation decisions. As recommendations are forward-looking predictions, uncertainty and intuition are involved making analysts prone to use simple heuristics not only to communicate with investors but also to revise their recommendations.

Data and methodology

My sample consists of 35,270 analyst recommendation revisions from 5,193 analysts for 1,454 unique US stock-listed companies. The dataset covers time period from November 1993 to September 2015. Analyst stock recommendations and earnings estimates data were gathered from IBES database, stock exchange data from CRSP database, accounting data from Compustat database, institutional ownership of stocks from Thomson Reuters and idiosyncratic volatility from Kenneth R. French’s database. To study the effects of reference points on recommendation downgrades, I employ a series of conditional fixed effects logistic regression models by controlling price and earnings momentum effects, valuation, fundamental and growth indicators as well as company and analyst specific factors related to analyst stock recommendation decisions.

Findings

Partly consistent with prior literature, I find that analysts tend to downgrade recommendations near the 52-week high and 52-week low stock prices. Furthermore, from the most common relative valuation multiples used by analysts, only P/E ratio and PEG ratio show some statistically significant evidence of anchoring to valuation multiples. With all control variables included, the odds of being downgraded by analysts are over 40% higher for companies approaching the 52-week high or low than that for other companies. In the case of approaching the 2-year high and 2-year low P/E ratios, the odds of being downgraded by analysts are 15% and 20% higher, respectively, although the evidence on the statistical significance is mixed. Provided evidence suggests that the 52-week high and low stock prices are more important reference points for analysts than high and low points of valuation multiples. Despite these anchors, my main results suggest that two other factors create by far the largest effects on revisions: price momentum in the recent weeks increases the odds and a high number of earnings forecast revisions decreases the odds of downgrading the most.