Going-public vs. private sales: A two-tiered agency approach

Abstract

This paper models a situation where an entrepreneur with assets in place and uncertain development opportunities decides whether to sell the business to public capital markets or to place it privately to a conglomerate. It finds that the two-tiered managerial hierarchy of a conglomerate is likely to cause more adverse effects of agency problem. Thus, going-public dominates private sales in motivating the entrepreneur to acquire more information about investment opportunities and in the profit performance of the business. The entrepreneur obtains less wealth if he sells the business privately at a price representing its profit potential when the entrepreneur and the manager of the conglomerate have the same managerial interests.

Introduction

Firms have a variety of sources to fund their new investment projects. One measure is to raise investment funds through public capital markets. The other is financing through internal capital markets (Williamson, 1975); that is, raising the funds for real investment from retained earnings and allocating them to various divisions within the firm. A special case of such a fundraising issue is that of a private and start-up firm, a venture business in particular, obtaining investment capital required for its growth. Should it be sold to dispersed investors in public capital markets, or to a private financier such as a conglomerate? What kind of ownerships will lead the firm to perform better?

From a managerial incentive perspective, private sales seem likely to dominate going-public, since a single shareholder or a small group of shareholders may provide better monitoring than dispersed atomistic shareholders (Shleifer & Vishny, 1986). But Aghion and Tirole (1997), Burkart, Gromb, and Panunzi (1997) and Bolton and Scharfstein (1998) notice that tight monitoring has an adverse effect on the incentives for the managerial and entrepreneurial activities of managers. Subrahmanyam and Titman (1999) examine the financing alternatives from an information perspective and argue that “information considerations can favor either public or private financing depending on how investors obtain information.” A key assumption in their analysis is that the insiders (managers) of a firm can only obtain information from outsiders, being either public investors or private financiers. Chemmanur and Fulghieri (1999), assuming the information asymmetry in favor of insiders, present a model featuring going-public's advantage of a low risk-premium and private financing's advantage of a low information cost. The entrepreneur's choice between going-public and private financing is then determined by the tradeoff between minimizing the information duplication in the Initial Public Offering (IPO) market and the reduction of the risk-premium.

In this paper, we turn back to the agency approach in line with Bolton and von Thadden (1998), Holmström and Tirole (1993), Stein (1997) and Williamson (1975). However, distinguished from these authors, we consider a two-tiered managerial structure within a conglomerate similar to Scharfstein and Stein (2000): a headquarters manager (HM) of a conglomerate on the top of the managerial hierarchy is intermediated by a group of division managers. More specifically, we develop a model in which an entrepreneur chooses to sell his business and its growth opportunities either to a public capital market or to a private financier. If he decides going-public, he (partially) sells his business and its growth potential and in turn the business becomes an independent corporation with him as the manager. In the case of a private sale, the entire business together with its growth opportunities is sold to a conglomerate run by an HM. Although the entrepreneur's business becomes a separate division within the conglomerate and he is the manager of the division, his management is subject to severe monitoring of HM.¹ Of course, conglomerates are only one type of private financiers. Alternatively, entrepreneurs can, for example, seek financing from private equity companies, who draw financial resources from capital markets such as pension funds to invest in private firms. Therefore, after developing a model for private sales to conglomerates, we extend the model to analyze the effects of sales to private equity companies.

In our basic model, the agency problem exists in three dimensions. One is that the efforts exerted by the entrepreneur and HM to learn the payoffs of investment projects deviate from the optimal level required by the outside investors. Another is that the project selected by the entrepreneur or HM when they are informed of the project payoffs is likely to differ from the outside investors' choice. The third is that public capital markets and the HM value the entrepreneur's assets and investment opportunities differently. The agency problem exists wherever owners delegate the managerial tasks to professional managers. However, in the case of public financing, where the entrepreneur acts as the manager of the independent corporation, the problem is mitigated by the entrepreneur's partial ownership of equity. In the case of private sales, the entrepreneur's agency problem as the manager of the division of the conglomerate is coupled by the agency problem of the HM; that is, the entrepreneur's distortion in learning effort and project selection is likely to be amplified by the HM's distortion. Moreover, tight monitoring by the HM can also weaken division management's incentive to conduct more entrepreneurial and innovative activities, as suggested by Aghion and Tirole (1997), Burkart et al. (1997), and Bolton and Scharfstein (1998). Thus, different strategies of business trading lead to different agency effects, and consequently different profit performances and exchange values of the business. The dominance of private equity placement under the conventional agency approach no longer exists in our model. In contrast, we can conclude that going-public is more likely to result in a better performance of the entrepreneur's business in terms of expected profits.

In related works, Stoughton and Zechner (1998) analyze the effects of different IPO mechanisms on the structure of share ownership, emphasizing the role of underpricing and rationing in determining investors' shareholdings. In a more dynamic setting, Mello and Parsons (1998) study the optimal IPO strategy with a tradeoff between selling shares to large controlling investors and small passive shareholders. Interestingly, because of different assumptions about information structure in their models, these two studies reach contrasting optimal strategies. Pagano and Roell (1998) also analyze ownership structure in terms of a company's controlling shareholding vs. dispersed shareholding. Their focus is on the tradeoff between the cost of providing a liquid market and over-monitoring. In contrast to these works, we investigate the effect of two-tiered management on the financing choice of entrepreneurs.

The rest of the paper is organized as follows. 2 Going-public decision, 3 The decision of private sales present a model of going-public and private sales to conglomerates, respectively. Section 4 compares the entrepreneur's strategies of selling his equity and their consequences. The empirical relevance and implications of the model's predictions are presented in Section 5. Section 6 extends the model to cover the case of sales to private equity companies. It also discusses the assumption of dispersed public ownership adopted in our analysis and the extension of the model that conglomerates provide incentive packages to entrepreneurs to align their interests with those of outside investors. The final section concludes the paper. For simplicity, we assume a zero discount rate and that all agents are risk neutral throughout this paper.