This paper was partially financed by the Henry Crown Institute of Business Research in

Israel.

The Institute’s working papers are intended for preliminary circulation of tentative

research results. Comments are welcome and should be addressed directly to the authors.

The opinions and conclusions of the authors of this study do not necessarily state or

reflect those of The Faculty of Management, Tel Aviv University, or the Henry Crown

Institute of Business Research in Israel.

Page 2

* Allen Michel is a professor of finance and economics at Boston University School of Management and a

senior expert at the Michel-Shaked Group. Jacob Oded is an assistant professor at Boston University

School of Management and a lecturer at The Faculty of Management, Tel Aviv University. Allen Michel:

(617) 353-4167, amichel@bu.edu. Jacob Oded: (617) 353-2679, oded@bu.edu.

Stock Repurchases and the EPS Enhancement Fallacy

Jacob Oded and Allen Michel*

This Draft: March 17, 2008

Abstract

A common belief among practitioners and academics is that the increased EPS associated

with a stock repurchase creates value for a firm’s shareholders. In this paper we show that

this belief is flawed. We demonstrate using a numerical example the magnitude of the

distortion that arises from using EPS to make such repurchase decisions. We similarly

analyze alternative payout policies, namely the payment of dividends and cash

Introduction

Record numbers of firms have announced share repurchases over the last several years.

Firms as diverse as MMM, Capital One Financial, Caterpillar, CBS and Accenture have

each recently announced repurchases of more than $1 billion. One commonly cited

reason behind the increased use of share repurchase programs in company press releases

and executive surveys is the increase in earnings per share (EPS).2 Over the last two

decades, the use of repurchases as a payout tool has increased dramatically. Firms now

distribute as much cash back to shareholders with repurchases as they do with dividends.3

This article investigates the effect of stock buybacks on both a company’s earnings per

share and the value of an investor’s holdings. It demonstrates an important, but generally

ignored effect of increasing a company’s EPS growth through share repurchase. Namely,

we show that the process of buying back shares, while increasing EPS, leaves the value of

an investor’s holdings unchanged. Furthermore, we demonstrate that the EPS increase

associated with a buyback is merely a risk-return tradeoff. This is because, in a

repurchase, the firm retires safe cash and as a result its assets become more risky. With

the increased risk the expected return increases and this is reflected in the higher expected

EPS.

Firms generally choose among several alternative options for their excess cash. These

typically include dividend payments, share repurchases, and cash accumulation (no

payout). We consider each of these alternatives. We show that the value of an investor’s

holdings is invariant with respect to the choice of payout policy, yet each alternative

provides a unique risk-return tradeoff which is reflected in the EPS pattern. These results

conflict with the commonly accepted intuition that increasing EPS through repurchase

creates economic value for the investor.

Miller and Modigliani (MM) (1961) demonstrate that in a perfect world payout policy

does not matter. The literature that followed their seminal work focused on explaining

how market imperfections make payout policy relevant. A broad range of reasons are

the creation or destruction of value for the shareholders. While MM is well known,

managers continue to use EPS as a significant input into their repurchase decision. Our

analysis thus demonstrates the magnitude of the distortion from using EPS to make share

repurchase decisions. At the same time it guides financial analysts how to interpret EPS

changes, namely, how to distinguish between EPS changes that are associated with

changes in expected shareholder wealth from EPS changes that are not.

Arnott and Asness (2003), and more recently Zhou and Ruland (2006), demonstrate

empirically that despite the negative effect of dividends on earnings growth, high

dividend payouts are not associated with weak future earnings growth. In fact, high

dividends actually predict earnings growth. Given that repurchases are increasingly used

to substitute for dividends as a payout tool, their effect on EPS growth is important to

understand. Our analysis is closely related to these investigations and suggests that the

positive effect that repurchases have on earnings growth should not be confused with real

economic growth at the firm.

As an application, we consider the effect that alternative payout policies would have had

on Exxon-Mobil’s EPS, given its recent history of sizable share repurchases and large

dividend payments. We then compare these results to Exxon-Mobils actual payout

policy. It is shown that more than 16% of the firms EPS growth over the last four years is

an artificial result of its repurchase program and cannot be associated with improvement

in operating performance.

Our results have important implications beyond the effect on EPS. For example, many

managers suggest they repurchase shares to reverse the dilution from employee stock and

option compensation. Our results suggest that repurchasing shares, while preventing

dilution of EPS, does not prevent dilution of value to shareholders. This is because the

granting of stock and options does dilute value whereas repurchases do not enhance it.

Similarly, a new practice which is gaining increasing popularity among firms is the

accelerated repurchase. In an accelerated repurchase the firm borrows a large quantity of

shares from its shareholders through an investment banker and retires those shares upon

receipt. Then, generally within six to nine months, the investment banker buys shares on

behalf of the firm in the open market and returns them to the lending shareholders.

Accelerated repurchases thus enable the firm to quickly increase EPS while repurchasing

shares slowly over time.5 The results in this paper imply that boosting EPS with an

accelerated repurchase does not create value for shareholders.

Our analysis adds to other recent investigations such as those by Cornell (2005) and

Fairchild (2006) who also question whether share repurchases enhance value. Cornell

demonstrates that because repurchases change the number of shares outstanding, per-

share data is misleading. For example, if the repurchased shares are reissued to

employees, repurchases can actually reduce shareholder value. Fairchild demonstrates the

misconception that repurchases create shareholder value by spreading total market value

5 The firm incurs an obligation to absorb any subsequent share price changes. The alternative of buying the

shares quickly may adversely affect the market price and exposes the firm to law suits claiming price

manipulation.

Page 5

over fewer shares. We not only demonstrate that repurchases may not enhance value, but

also explore the manner in which repurchases affect EPS. It is shown that the increase in

EPS associated with repurchases is misleading and merely reflects a risk-return tradeoff.

EPS under Different Payout Policies

To understand the effects of alternative payout policies available to a firm considering

repurchasing shares, we consider the following financial options: 1) accumulating all

funds earned each year without paying out any cash, 2) paying dividends with all of the

firm’s earnings, and 3) repurchasing shares using all of the firm’s earnings. Without loss

of generality, we demonstrate our insights using a numerical example.

Consider an all equity firm that initially has one million shares outstanding. The firm’s

balance sheet consists of $5 million in risky assets and $5 million in cash. Thus with $10

million of assets, the firm’s one million shares are traded at $10 per share. The risky

assets are expected to generate a 15% return (in cash earnings) each year. The firm’s

cash is expected to generate a 5% return each year. The expected return on the firm’s

total assets is thus 10%. During Year 1, the firm earns $750,000 from its risky assets and

$250,000 from its cash assets. Thus, at the end of Year 1, the firm has earnings of $1

million and its EPS is $1 per share.

To simplify the analysis, and in order to focus on the effects of payout policy, we assume

that the firm does not have new risky investment opportunities for its cash. Thus, the firm

can either pay out free cash that it earns to its shareholders or invest the free cash at the

risk free rate. We also assume that cash payouts can only be distributed from the firm’s

earnings. We further assume no taxes, no signaling motivation, and no agency problems.

Under this framework we consider the consequences of each of the three possible payout

policies: cash accumulation (no payout), dividends, and share repurchase.

The Policy of Cash Accumulation:

With a policy of cash accumulation, the firm reinvests all earnings in safe assets. Thus, at