Divisional Informativeness Gap and Value Creation from Asset Sales

• Author: Chintal A. Desai,Manu Gupta
• Published date: 01 November 2016
• DOI: http://doi.org/10.1111/fire.12112
• Date: 01 November 2016

The Financial Review 51 (2016) 559–578

Divisional Informativeness Gap and Value

Creation from Asset Sales

Chintal A. Desai and Manu Gupta∗

VirginiaCommonwealth University

Abstract

Nanda and Narayanan (1999) show that the information asymmetry between the managers

and market participants regarding divisional cash flows helps explain the value creation on

asset sales. Based on their theoretical framework, the divisional informativeness gap hypoth-

esis predicts that the announcement-period return increases with the difference in cash-flow

informativeness of retained and divested divisions prior to the divestiture. Our results, using

industry-average earnings response coefficient as a proxy for cash-flow informativeness of

a division, support this prediction. The effect is stronger when a conglomerate retains the

division with relatively greater growth opportunities.

Keywords: asset sales, divestitures, cash-flow informativeness, corporate restructuring

JEL Classifications: G34, G14

1. Introduction

In an asset sale, the conglomerate divests a division, subsidiary, or product line.

Nanda and Narayanan (1999), hereafter NN, show that the information asymme-

try between the managers and the market regarding the divisional cash flows can

cause undervaluation of a two-division firm. In their model, the difference in cash-

∗Corresponding author: Department of Finance, Insurance and Real Estate, School of Business, Virginia

Commonwealth University, 301 West Main Street, Richmond, VA 23284; Phone: (804) 828-7175; Fax:

(804) 828-3972; E-mail: mgupta2@vcu.edu.

C2016 The Eastern Finance Association 559

560 C. A. Desai and M. Gupta/The Financial Review 51 (2016) 559–578

flow informativeness of divisions is the key driver of undervaluation, and hence,

the source of wealth gain on an asset sale. We form the divisional informativeness

gap hypothesis based on the theoretical framework of NN, which predicts that the

announcement-period return increases with the difference in cash-flow informative-

ness of the retained and divested divisions prior to divestiture.1The objective of this

research is to test this prediction using a sample of U.S. divestitures.

We estimate the divisional informativeness gap of a conglomerate by first iden-

tifying the three-digit Standard Industrial Classification (SIC) code of the divested

and the retained divisions based on their business operations for the year prior to

the asset sale. Then, we perform an ordinary least squares regression of earnings-

announcement abnormal-return on earnings surprises for all firms with the same

three-digit SIC code as the given division. The residual of this regression is the noise

in cash-flow informativeness of a division. The divisional informativeness gap is the

absolute difference between noise of cash-flow informativeness of the retained and

the divested divisions.

Our results, based on a sample of 242 asset sales for the period 1990–2010,

support the divisional informativeness gap hypothesis. The asset sale announcement-

period abnormal returns increase with the difference in cash-flow informativeness of

divisions prior to the asset sale. This effect is stronger for a conglomerate that retains

the division with relatively greater growth opportunities.

The NN model is built on two key assumptions. First, firms sell assets to finance

an investment opportunity of the retained division. There is empirical support to this

assumption. Hovakimian and Titman (2006) empirically show that the cash from

a voluntary asset sale is an important determinant of the firm’s future investment

expenditure. Borisova and Brown (2013) find evidence that the cash from asset sale

is useful for financing future research and development expenditure. Second, the

market participants are unable to observe a division’scash flow for a given period. In

reality,the firms must report performance information of segments that represent 10%

or more of consolidated sales. However, we can still not eliminate the possibility of

cross-segment shifting of earnings and expenses. Managers are likely to manipulate

divisional cash flows to avoidreporting losses of a particular division as losses attract

more external monitoring (Hann and Lu, 2009). Chen and Zhang (2007) show that the

undervaluation of a conglomerate is a result of industry wide cross-segment shifting

of earnings. They show that a superior quality firm divests a businessunit to separate

itself from industry peers, and thus overcomes undervaluation.

Our research makes two contributions to the corporate finance literature. First,

it identifies additional source of shareholders’ wealth gains on asset sales. Second,

it uses a measure based on the industry-level earnings response coefficient for the

divisional informativeness gap of a conglomerate.

1In the next section, we review the related literature on asset sales and conceptually explain how the NN

framework helps form the divisional informativenessgap hypothesis.