Size of Financing Need and the Choice between Asset Sales and Security Issuances

• Author: Manu Gupta,Chintal A. Desai
• DOI: http://doi.org/10.1111/fima.12242
• Date: 01 June 2019
• Published date: 01 June 2019

Size of Financing Need and the Choice

between Asset Sales and Security

Issuances

Chintal A. Desai and Manu Gupta∗

Westudy the effect of the size of financing need on a firm’schoice between selling assets and issuing

securities to finance its investments. The balance sheet effect predicts that a firm prefers to sell

assets when the financing need is relatively small as thereis less information asymmetry regarding

the value of a (small) subset of its assets. When the financing need is large, a firm prefers issuing

securities to selling assets. We find evidence supporting the prediction. Our findings remain

unchanged when we employ measures of financing need that are relatively independent of the

actual amount of financing raised.

Firms raise external capital by issuing financial claims in the form of debt or equity. The

pecking order theory recommends first using internal funds and then issuing a security with

the least information asymmetry. The market timing theory suggests issuing a security that is

overvalued in the market (Baker and Wurgler, 2002). An often-overlooked source of capital is

asset sales. In this paper, we study the choice between two financing sources: issuing securities

versus selling assets. We focus on one potential determinant of this choice: the size of a firm’s

financing need.

The theoretical work of Edmans and Mann (2018) identifies the balance sheet effect. It helps

establish a relation between the size of financing need and the choice between an asset sale and

a security issuance. In their model, the assets-in-place of a firm consists of core and noncore

assets.1The values of both assets are unknown.The f irm finances the assets-in-place by securities

such as debt and equity, which gives the existing securityholder a claim on the firm’s total assets.

The information asymmetry regarding the value of the firm’s relatively small asset is less than

that regarding the value of its financial securities. Therefore, when the financing need is small,

the firm prefers an asset sale. In contrast, when the f inancing need is large, the firm prefers a

security issuance. The underlying economic arguments are as follows. The new securityholders

have claims on the entire firm, that is, on the entire balance sheet of the firm, which includes the

funds raised. The value of the funds (cash) raised is known with certainty, which helps reduce the

valuation uncertainty of the firm’s assets-in-place. The asset purchasers, however, have claims

only on the purchased assets. They do not have claims on the entire balance sheet, thus on the

We are grateful to an anonymous reviewer formany insightful comments, which have significantly improved the paper.

We also thank Utpal Bhattacharya(Executive Editor) for his comments. We acknowledge comments from Etti Baranoff,

Alisa Brink, Puneet Prakash, BrentSmith, Mira Straska, Lan Xu (discussant), and the seminar participants at the Virginia

Commonwealth University (VCU), and Eastern Finance Association (2018) meetings. Gupta acknowledges the summer

research support fromthe VCU School of Business. We are solely responsibleof errors and omissions.

∗Chintal A. Desai is an associate professorof f inance in the School of Business at the VirginiaCommonwealth University

in Richmond, VA. Manu Gupta is an associate professorof finance in the School of Business at the Virginia Commonwealth

University in Richmond, VA.

1Weconsider noncore assets as a subset of the f irm’s total assets.

Financial Management •Summer 2019 •pages 677 – 718

678 Financial Management rSummer 2019

funds raised. When the financing need is sufficiently large, the information asymmetry regarding

valuation of the firm’s financial securities is less than that for a relatively large asset. Thus, the

balance sheet effect predicts that a firm will sell assets when its financing need is relatively small

but will issue securities when its financing need is large.

For empirical analyses, we use firm-year observations from flow-of-funds statements for

1971–2016. Using the cash-flow identity of sources for funds and uses of funds, we identify

2,403 firm-years as asset sales and 26,275 fir m-years as security issuances. In both cases, firms

use proceeds for investments purposes. We begin our analysis by using the actual amount of fi-

nancing raised to measure a firm’s financing need. The average value of financing raised through

an asset sale is 4.5% of total assets. In contrast, the average value of financing raised through a

security issuance is 12.3% of total assets. Furthermore, the number and percentage of asset sales

decline with an increase in financing need. For example, when the amount raised is between 1%

and 2% of total assets, the number of asset sales is 628, which is 15% of 4,130—the combined

number of asset sales and security issuances for that financing size. When a fir m’s financing

need exceeds 12% of its assets, we do not observe any asset sales. The number of security is-

suances, however, is 8,835. This data pattern provides evidence for the balance sheet effect.2The

multivariate regressions confirm the data pattern by showing that the odds of an asset sale over a

security issuance decrease with the size of financing need.

In our initial analysis, we do not directly observe a firm’s demand for financing. Instead, we

observe the amount the firm is able to raise. This equilibrium outcome is the result of mutual

decisions by financing providers (asset or security buyers) and the firm requiring financing.

Therefore, our measure of the size of financing need is not exogenous. To address this concern,

we use a firm’s level of external finance dependence as a measure of its size of financing

need (Demirguc-Kunt and Maksimovic, 1998; Ross et al., 2009).3We define external finance

dependence as the difference between a firm’sactual growth rate and the internal growth rate. The

internal growth rate is the theoretical growth rate from the percentage-of-sales-based approach

of the financial planning model. A f irm can achieve the internal growth rate by using only

internally generated funds. A firm’s large external finance dependence indicates a higher level of

financing need. Our main results remain unchanged with this alternative measure of a fir m’ssize

of financing need.

In theory, the balance sheet effect applies equally to debt and equity. The purchasers of newly

issued debt or equity have claims on the entire firm, similar to those of existing securityholders,

and the cash raised from both sources appear on a firm’s balance sheet. The data, however,

reveal differing patterns for a firm’sf inancing choice between asset sales and debt issuances, and

between asset sales and equity issuances.4We observe a significant decline in the proportion of

asset sales relative to debt issuances with a small change in the size of financing need (2% of total

assets to 3% of total assets). In contrast, the proportion of asset sales relative to equity issuances

remains almost constant until the financing size increases to 11% of total assets. Therefore, these

patterns suggest that a firm switches from asset sales to equity issuances at a much higher level

of financing need than when switching from asset sales to debt issuances. This is intuitive. Equity

financing is expensive and the firm will issue equity only when the financing need is suff iciently

large so that issuing debt is no longer adequate. In a reduced sample where the size of financing

2Figure 1 in Section II.E shows this pattern.

3Wealso perform a two-stage regression. In the f irst-stage regression, we regress actual amount raised on supply-related

variables. In the second-stage regression, we use the residual from the first-stage regression as a proxy for a firm’s size

of financing need.

4Figure 3 in Section III.C shows these patterns.

Desai & Gupta rSize of Financing Need and the Choice between Asset Sales and Security Issuances 679

need is less than 12% of total assets, we find that the balance sheet effect is four times larger for

debt issuances versus asset sales than for equity issuances versus asset sales.

We also analyze how the balance sheet effect plays out in the Myers and Majluf (1984)

framework. Myers and Majluf (1984) consider the possibility of splitting assets, stating that “if

[part of assets-in-place] can be sold at intrinsic value, the firm treats the proceeds as additional

slack and looks again at its issue-invest decision” (p.202). They pose a question for future

research: What will happen if the firm is able to sell an asset only at a discount? Intuitively,

the firm sells the asset at a discount, if the valuation uncertainty of the asset precludes the asset

purchaser in determining its “true” value. This situation is more likely to hold when the asset

under consideration is large. In this case, the firm prefers issuing a f inancial security to selling an

asset at a discount. Thus, the balance sheet effect under the Myers and Majluf (1984) framework

predicts that for small investment expenditures, the firm prefers internal financing, but when

investments increase, the likelihood of external financing also increases. The data patterns and

regression results support this prediction. The odds of external financing over internal fi nancing

increase with the size of investment need.

Our research is relevant to academics as well as practitioners. We believe our paper is the

first to carry out an empirical investigation of a fir m’s financing choice between asset sales and

security issuances. We analyzethe effect of investment need on a firm’s choice of internal versus

external financing, which is also novel to the corporate finance literature. Our comprehensive

data set of all US industrial firms, emphasis on data patterns, and rigorous empirical analyses are

particularly useful to practitioners in their corporate decision making.

The paper proceeds as follows. Section I provides the related literature and underpinnings of

our hypotheses. It also shows how cash-flow identity guides our empirical analyses. Section II

describes the sample and variables. Section III reports the results. Section IV presents the balance

sheet effect in the Myers and Majluf (1984) framework of internal versus external financing.

Section V demonstrates the sensitivity of the balance sheet effect to firm characteristics. Section

VI provides concluding remarks.

I. Related Literature and Motivation

A. Hypotheses Development

Our research is at the intersection of the corporate restructuring and financing literatures. In an

asset sale (divestiture) transaction, a conglomerate sells an existing division, segment,subsidiar y,

or product line to a third party. In return, the selling firm often receives cash, and in some cases

receives shares or a combination of cash and shares from the buying firm. The shareholders’

wealth, on average, increases when a conglomerate announces an asset sale (see, among others,

Rosenfeld, 1984; Jain, 1985; Mulherin and Boone, 2000; Dittmar and Shivdasani, 2003).5Firms

engage in corporate restructuring via an asset sale to modify the firm’s asset portfolio and scope.6

Corporate restructuring is also associated with efficient resource allocation, increased focus on

5Eckbo and Thorburn (2013) survey empirical research on asset sales.

6The other commonly used forms of restructuring are spinoffs and equity carve-outs. In a spinoff, a conglomerate

distributes subsidiary shares to its existing shareholders. In an equity carve-out, a conglomerate consummates an initial

public offering of some portion of the subsidiary. We exclude spinoffs and equity carve-outs because cash is not raised

in a spinoff, and the firm sells partial ownership of a subsidiary during the first stage of a carve-out, thereby retaining

the option of reacquisition, spin off, or sell-off of the remaining portion in the second stage (Perotti and Rossetto, 2007;

Desai, Klock, and Mansi, 2011).