Financial analysts' bundling across firms: Target prices and stock recommendations

• Published date: 01 December 2023
• Author: Yu‐An Chen,Dan Palmon
• Date: 01 December 2023
• DOI: http://doi.org/10.1111/jfir.12348

Received: 27 June 2022

|

Accepted: 7 July 2023

DOI: 10.1111/jfir.12348

ORIGINAL ARTICLE

Financial analysts' bundling across firms:

Target prices and stock recommendations

Yu‐An Chen

1

|Dan Palmon

2

1

Department of Accounting and Taxation,

Stillman School of Business, Seton Hall

University, South Orange, New Jersey,USA

2

Department of Accounting & Information

Systems, Rutgers Business School, Rutgers

University, Newark, New Jersey, USA

Correspondence

Dan Palmon, Department of Accounting&

Information Systems, Rutgers Business

School, Rutgers University, Newark, NJ

07102, USA.

Email: dpalmon@business.rutgers.edu

Abstract

A rising trend is that analysts bundle earnings forecasts for

multiple firms on the same day, and such forecast bundling

is associated with low‐quality forecasts. We explore target

price bundling and recommendation bundling. Factors

driving bundling revisions of an output for multiple firms

vary across three outputs: forecasts, target prices, and

recommendations. Target price (recommendation) revisions

bundled for firms are less informative than stand‐alone

revisions. Although consistency in the direction of revisions

between different outputs is typically associated with

higher perceived quality, consistent revisions between

outputs are not associated with higher informativeness of

revisions bundled for multiple firms.

JEL CLASSIFICATION

G24

1|INTRODUCTION

Since the regulatory, technological, and market structure changes in the 2000s, analysts have increasingly relied on

batch production for their earnings forecasts (Drake et al., 2020). Regulatory shocks, such as the Global Settlement,

change the funding sources and business models of research departments in brokerage firms, and technological

improvements increase clients' demands for thematic research reports. These market structure changes drive

analysts to promote specialized services, such as investor conferences or marketing trips, in which analysts have

incentives to update their forecasts for all related firms. To cut costs and promote production efficiency, analysts

J Financ Res. 2023;46:1047–1102. wileyonlinelibrary.com/journal/JFIR

|

1047

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and

reproduction in any medium, provided the original work is properly cited.

© 2023 The Authors. Journal of Financial Research published by Wiley Periodicals LLC on behalf of The Southern Finance

Association and the Southwestern Finance Association.

bundle earnings forecasts for multiple firms on the same day. However, earnings forecast bundling gives analysts

less time to incorporate firms' idiosyncratic natures into analyses, resulting in lower earnings forecast quality in

terms of lower forecast accuracy, less boldness, and lower informativeness (Drake et al., 2020).

Analysts' three primary forecast outputs are earnings forecasts, target prices, and recommendations. Extending

Drake et al.'s (2020) study on earnings forecast bundling, we focus on analysts' target price bundling and

recommendation bundling for multiple firms on the same day.

1

We ask (1) whether factors that determine an

analyst's decisions to bundle revisions of a forecast output for multiple firms are similar across earnings forecasts,

target prices, and recommendations; (2) whether the market perceives bundled target price revisions or bundled

recommendation revisions to have lower quality and be less informative than stand‐alone revisions; and (3) how

consistency in revision directions between different forecast outputs is related to the perceived quality of revisions

bundled for firms. To facilitate comparisons, we consider market‐based perceived quality to be the way the market

reacts to revisions of forecasts, target prices, and recommendations.

2

Ononehand,theproductioncostandefficiencyargumentbyDrakeetal.(2020) may apply similarlyto all

three forecast outputs under the observed regulatory, technological, and market structure changes. If analysts

follow similar batch production processes to make different forecast outputs, factors determining bundling

revisions of a forecast output for firms will not vary much across these outputs. On the other hand, Yezegel

(2015) notes differences in the purpose of earnings forecasts and stock recommendations. Earnings forecasts

reflect analysts' predictions of various future financial statement line items, whereas stock recommendations

reflect the divergence or convergence between an analyst's valuation and the market's valuation. Therefore,

factors contributing to bundling revisions of one output can differ from factors contributing to bundling

revisions of another output. Specifically, incorporating shared industry or market factors for bundled firms

appears to be a natural part of predicting target prices and making recommendations, but less so for forecasting

earnings.

Commonalities and variations coexist in the factors determining earnings forecast bundling, target price

bundling, and recommendation bundling. Although analysts consider some firm characteristics, such as firm size and

prerevision trading volume, in similar ways for their bundling decisions, they also consider certain factors differently

across different forecast outputs. For example, we find that accounting accrual is a strong predictor for earnings

forecast bundling but less so for target price bundling or recommendation bundling. Given the differences between

the determinants of bundling target prices and bundling recommendations compared to the determinants of

bundling forecasts, the unfavorable effect of bundling on earnings forecast quality may not directly extend to the

quality of target prices or recommendations.

The second question is whether an analyst's revisions of target prices or stock recommendations have lower

perceived quality when bundled for several firms on the same day. Bundled target prices or recommendations have

lower quality if an analyst generates forecasts, target prices, and recommendations sequentially through

fundamental analysis. Analysts' earnings forecasts serve as crucial inputs for their valuations and recommendations

1

The bundling of a forecast output refers to “across‐firm”bundling of revisions to a certain type of forecast output (i.e., earnings forecasts, target prices, or

stock recommendations) for multiple firms on the same day. Drake et al. (2020) look at analyst earnings forecast bundling, whereas we look at analyst

target price bundling and recommendation bundling. However, another literal meaning of bundling can be an analyst's “across‐output”bundling on the

same day, either for one firm or for multiple firms. Consistent with Drake et al. (2020), we mainly use the term bundling to refer to the bundling of revisions

to a certain forecast output for multiple firms on the same day.

2

We follow Drake et al. (2020) and apply the market‐based tests of the “perceived”quality of revisions of a forecast output bundled for multiple firms to

earnings forecasts, target prices, and recommendations. Unlike earnings forecasts, target prices or stock recommendations do not have realized values on

specifically forecasted dates (i.e., earnings announcements) that can serve as benchmarks to infer accuracy. However, market‐based tests facilitate

comparing the premiums or discounts for the market reactions to bundled forecasts, bundled target prices, or bundled recommendations. Therefore, in

empirical tests, we mainly use market‐based informativeness measures in terms of signed or unsigned stock returns and abnormal trading volume. A

limitation of focusing on the perceived quality is that the results do not speak to the actual forecast output quality of target prices or recommendations

bundled across firms. Also, we do not identify the causal explanation for the documented lower informativeness of bundled target prices or bundled

recommendations. Potential explanations include information overload, low‐quality bundled forecasts as inputs to target prices or recommendations, or

low‐quality target prices or recommendations themselves as results of an analyst's batch production. Readers should be cautious in interpreting the

baseline results.

1048

|

JOURNAL OF FINANCIAL RESEARCH

(Bradshaw, 2002,2004; Brown & Huang, 2013; Brown et al., 2015).

3

However, if “garbage in, garbage out”(GIGO)

holds, bundled target prices or recommendations can be perceived as low quality when made together with and

based on poor‐quality bundled earnings forecasts (Ertimur et al., 2007; Gleason et al., 2013; Loh & Mian, 2006).

Even if target prices or recommendations are revised separately from earnings forecasts, they can be perceived

as low quality when bundled for multiple firms because analysts do not customize analyses about individual

firms (Clement, 1999; Drake et al., 2020; Plumlee, 2003). Either case predicts that bundled target prices or

recommendations will have lower perceived quality than stand‐alone target prices or recommendations.

The process analysts use to determine target prices or recommendations differs from the process they use to make

earnings forecasts (Asquith et al., 2005;Brav&Lehavy,2003;Francis&Soffer,1997). Differences in economic or other

factors that influence analysts' behavior can drive the differences in their activities of making earnings forecasts, target

prices, and recommendations (Iselin et al., 2021;Palmonetal.,2020). In fact, Drake et al. (2020) do not find an increasing

trend for recommendation bundling, and analysts might not apply the same batch production used for bundled forecasts

to target prices or recommendations. Observed variations in the determinants of bundling revisions of a forecast output

for firms across different outputs also motivate the examination of whether target price bundling or recommendation

bundling, like forecast bundling, is associated with lower informativeness.

Furthermore, analysts' bundled target price or recommendation revisions may offer more, rather than less,

information value through thematic research compared to stand‐alone revisions (Drake et al., 2020). Bundled target

prices or recommendations may better meet investors' information needs for evaluating firms on a relative basis

(Boni & Womack, 2006; Da & Schaumburg, 2011). Given that analysts' industry expertise matters to the perceived

quality of their forecast outputs (Bradley et al., 2017; Brown et al., 2015; Clement & Tse, 2003; Clement, 1999;

Kadan et al., 2012,2020), the relation between bundling revisions of a forecast output and the perceived quality of

revisions of that bundled output may vary with whether an analyst bundles a firm with its industry peers or other

firms. Besides extending Drake et al. (2020) from forecast bundling to target price bundling and recommendation

bundling, we consider the role of how an analyst bundles firms.

Additional benefits likely to result in higher perceived quality for an analyst's target prices or recommendations

arise when the analyst analyzes a group of companies in the same industry simultaneously, as opposed to analyzing

each company separately at different times. For instance,target price bundling may reflect ananalyst's ability to react

to several firms' market risks, which can signal higher quality valuation. These benefits and the related impact of

bundling on the perceived quality of the analyst's outputs arelikely higher when the analysis aims to produce target

prices or recommendations, not just earnings forecasts.Some consumers of analysts' outputs, such as portfolio

managers, may benefit by comparing two or more companies in the same industry (Boni & Womack, 2006;

Da & Schaumburg, 2011), and the benefits and related impacts of bundling on the perceived quality of analysts'

outputs may be higher for target prices or recommendations than for earnings forecasts. Whether Drake et al.'s

(2020) results about bundled forecasts can be generalized to bundled target prices and bundled recommendations,

and whether analysts' choice to bundle firms within the same industry or across industries plays a rolein the market

reaction are empirical questions.

In this article, we reproduce Drake et al.'s (2020) findings of a rising trend over time in analysts' forecast

bundling for multiple firms on the same day. There is a similar increasing trend in target price bundling but not in

recommendation bundling. When analysts bundle revisions of two or three types of outputs for multiple firms on

the same day, analysts often revise these outputs together. When an analyst bundles multiple firms' earnings

forecasts and bundles target prices on the same day, 87% of these observations are issued simultaneously, and 77%

are revised in a consistent direction; that is, both earnings forecasts and target prices are revised upward, or both

are revised downward. The presence of simultaneous revisions bundled for multiple firms between different

3

Several studies support this view on the linkages between analysts' various forecast outputs (Bradshaw, 2002,2004). For example, Bradshaw (2004,

p. 26) examines a predictable linkage that “analysts use their earnings forecasts along with other information to estimate a stock's value, which is then compared

to the actual trading price of the stock and forms the basis for the recommendation.”

FINANCIAL ANALYSTS BUNDLING

|

1049