Preferred Stock Contracts - Where’s the Accounting?

By

Alastair Murdoch

Department of Accounting and Finance

University of Manitoba

Winnipeg, Canada

R3T 5V4

and

Kathryn Hudyma

Price Waterhouse

Winnipeg, Canada

R3B 0X7

Phone 204 474 8439 Fax 204 275 0181
E-mail

First draft: 95.06.20

This draft: 97.12.11

We received helpful comments on earlier drafts of this paper from Phil Beaulieu, Tim Cairney, John Elliott, Jim Gaa, Larry Gould, Karim Jamal, Charles Mossman, Mike Stein, Don Stokes, Dan Thornton, anonymous referees and participants at the University of Manitoba, universities of Alberta and the Canadian Academic Accountants Association workshops. Financial support of the Faculty of Management at the University of Manitoba is gratefully acknowledged.

Data availability: A list of sample firms is available from the author.

An Examination of Preferred Stock Contracts

Abstract

This paper examines the covenants in the prospectuses of 134 preferred shares issued by 82 Canadian companies during the period 1986-1994. We find that the covenants

(1)do not rely on accounting numbers

(2)are more numerous and more restrictive for preferred shares that are more debt-like.

The second finding provides some support for accounting standard setters who have recently moved to require that debt-like preferred shares be treated as debt rather than as equity. The first finding confirms a trend in the United States away from the use of moving GAAP in covenants, a trend which may be due to accounting standard setters changing GAAP too frequently.

Introduction

To lower the cost of debt financing, the owners of firms, through management, voluntarily enter into covenants which limit future actions of the firm. For instance, firms issuing bonds will commonly agree not to issue additional debt of higher rank, not to pay dividends, and to permit the early redemption of the debt if certain conditions hold. Customarily, these conditions are expressed in terms of quantities reported in the firm’s annual financial statements. Covenants employing accounting variables provide one justification for the existence of accountants, standard setters (CICA Handbook section 1000.09), and auditors. The presence of such covenants is documented for American bonds (Begley 1992; Leftwich 1983; Mohrman 1996; Smith and Warner 1979), for UK bonds (Citron 1995), and for certain Canadian securities (Thornton and Bryant 1986).

The same conflict of interest between bondholders and stockholders which these covenants are designed to control may exist between holders of preferred shares and holders of common shares. However, covenants in preferred shares contracts may not be as extensive as those in bond contracts for several reasons. Firstly, in most cases preferred shares are issued in addition to, rather than instead of, debt. The preferred shareholders may then be effectively protected by the covenants in the bond contracts. Secondly, preferred shares are permanent financing, while bonds are temporary. Preferred shareholders are less concerned with management actions that change the period in which cash flows are realized as long as the present value remains unchanged. Thirdly, preferred shares that are participating or convertible are protected from most actions of the common shareholders. On the other hand, the holders of preferred shares that are retractable (i.e. redeemable at the option of their holders) would have similar concerns to bondholders.

Thus whether preferred share contracts contain the same covenants as bonds do is an empirical question.[1] And one might expect that the nature and extent of the covenants would be related to other features (e.g. participation, conversion, and retraction rights) of the preferred shares. To date only one published study has examined this issue. Stokes and Whincop (1993) find that the covenants in Australian preferred stock contracts are similar to those in debt contracts. The similarity extends to the prevalence of the contracts’ use of accounting numbers. They also document that the more debt-like the preferred share, the more probable the existence of these covenants.

This study continues in this line of research by examining the covenants in Canadian preferred shares issued in 1986 to 1994. Many of our results are similar to those Stokes and Whincop (1993) find for Australian preferred shares. We differ in finding little reliance on accounting numbers.

In finance, contracting theory (Smith and Warner 1979) addresses how firms contract with debt-holders, but not how firms decide how much debt (if any) to issue. That is, it takes as given that projects are financed from more than just internal sources (retained earnings/cash flow) and the issue of common shares. Once one admits other methods of financing, one is led to ask what are the optimal types, characteristics and combinations of these methods.

These issues are important to accountants who are responsible for determining an appropriate presentation of the sources of financing in the financial statements. The traditional presentation is to classify the sources as either debt or equity. But is this traditional dichotomy still appropriate in the light of the growth in the types and characteristics of financial instruments? If the answer is yes, then what criteria are to be used in deciding whether (or to what extent) a given instrument is debt or equity? If the answer is no, then what is the appropriate way of presenting the right hand side of the balance sheet?

This paper does not pretend to answer these questions.[2] It does attempt to provide information relevant to the questions based on recent issues of one kind of financing, namely preferred shares. Preferred share financing is a particularly useful source of facts relevant to these questions for the following reasons.

1.After debt and common equity, preferred shares are the most recognized and best understood of financial instruments. They have been used for a century and have been analyzed in the finance and accounting literature.

2.Developments in the nature and use of preferred shares parallel and are typical of developments in financial instruments generally. Although the “plain vanilla” preferred share has been around for decades, for most of this time the only features sometimes added to preferred shares were conversion and participation rights and cumulation of dividends. Only in the last two decades were many other features such as put and call options added. The use of these new features parallels the generally rapid expansion in the types of financial instruments in common usage.

3.Many of the challenges to accounting standards posed by financial instruments in general can be illustrated by the challenges posed by preferred shares in particular. Some characteristics of preferred shares are similar to those of equity, others are similar to those of debt. For a given preferred share, the proportion of the debt-like to equity-like characteristics may vary over the life of the instrument.

The growth in the types and features of financial instruments has forced accounting standard setters to address the questions raised above. The addition of these “add-ons” to the plain vanilla shares is one manifestation of this growth. The ability to answer these questions requires understanding the financial instruments and the reasons for their use. We attempt to describe in this paper the nature and extent of the use of these various “add-ons” in Canada. We hope that this will assist the reader in understanding the issues surrounding the proper accounting treatment of preferred shares in particular and financial instruments in general.

For instance, the argument for dichotomous classification of the sources of financing presumably rests on the belief that debt instruments exhibit one set of characteristics and that equity instruments exhibit another. Having described those characteristics, there should be virtually no instruments that do not fit clearly into one classification or the other. Finding increasingly many instruments that exhibit characteristics from both classifications should lead one to question whether dichotomy is appropriate. One could speculate on how useful classifying the population as male or female would be if one began encountering more and more people who had characteristics of both genders. Of course, such encounters could also lead to a re-examination of the circumstances in which the classification was useful (e.g. the classification might still be useful in the design of public washrooms but not in the design of employment criteria) or a re-examination of which characteristics are fundamental and which characteristics are coincidental. Knowledge of the evolving characteristics of different kinds of preferred shares may lead to the view that the debt/equity dichotomy is no longer appropriate. Alternatively, it may lead to a review of the characteristics of debt and equity to determine which characteristics are fundamental to classification and which are coincidental.

The paper is organized as follows. The following section reviews related literature. This segment is followed by a description of recent Canadian preferred stock issues. The penultimate section examines the lack of use of accounting variables in the contract covenants. The paper concludes with a summary of the study’s findings.

Literature review

Contracting theory proposes that bond contracts should contain covenants restricting their issuer’s investing and financing activities. We review the theory and the findings from studies pertaining to contractual covenants executed in the United States, Canada, and the United Kingdom. These studies generally support the theory. They show that some of these contractual restrictions only apply in certain circumstances and that many of these circumstances are defined in terms of accounting numbers. Table 1 provides a list of the papers discussed below and the various accounting numbers found in the covenants studied.

In the United States, Smith and Warner (1979) examined ways in which debt contracts are written to control the conflict between bondholders and stockholders and drew inferences regarding the role of contractual forms in the capital structures of firms. Their analysis indicates that covenants restricting dividend, financing, and production/investment policies are frequently specified in terms of income statement or balance sheet variables[3]. Covenants may prohibit certain actions, for instance, the payment of dividends if retained earnings or shareholders’ equity falls below a certain level. Covenants may also require that working capital, the debt/equity ratio or other accounting variables not fall below a certain level. “If a breach of the covenant occurs, the lender is in a position to use this early warning to take whatever remedial action is necessary.” [Page 453 of Commentaries, quoted by Smith and Warner.] Smith and Warner concluded that production/investment policy is expensive to monitor; whereas dividend and financing policy involve lower monitoring costs. Consequently, they found that extensive direct restrictions on production/investment policy are not often observed, but that dividend and financing policy covenants are written to give stockholders incentives to follow a firm-value-maximizing production/investment policy.

In Canada, Thornton and Bryant (1986) examined covenants of debt and preferred stock, including convertible issues but excluding participating preferred stock[4]. The securities studied were issued between 1974 and 1981; 71 of the 86 issues used accounting based ratios in their covenants.

Begley (1992) examined 130 public issues of non-convertible debt (senior and subordinated) by 126 United States industrial companies listed in Moody’s Industrial or OTC Manual issued in 1975 to 1979 (Page 10-11). 45% of them (44% of the senior and 50% of the subordinated) restricted the payment of dividends based on the level of retained earnings. 38% of them (50% of the senior and 7% of the subordinated) restricted issuing additional debt “typically ... if the ratio of funded debt to consolidated net tangible assets is expected to exceed a specified maximum” (Page 15). About 11% of them (15% of the senior and 0% of the subordinated) restricted investment in securities “based upon a variety of accounting numbers” (Page 16). She stated, “Although subordinated debt issues seldom include covenants restricting investments or additional borrowing, it can be argued that the debt holders of these issues can expect to be protected by the covenants governing the firm’s other public and private issues. If this is true then the cost of including these covenants in the current issue should be extremely low and we would therefore expect to see the covenant explicitly included. If the covenant is not explicitly included then protection is not guaranteed. The issuing firm could renegotiate the terms of the issue containing the covenant or even repurchase the issue, leaving the holders of the new debt without protection.” (Page 16).

Citron (1995) examined the covenants in 108 UK public debt contracts issued in 1987-1990. 32 of these contracts contain accounting-based covenants. All 5 of the debt contracts secured by a floating charge had accounting-based covenants. 22 of the 46 bonds and 5 of the 22 convertible unsubordinated debt contracts had accounting-based covenants. None of the debt contracts secured by a fixed charge and none of the convertible subordinated debt contracts had accounting based covenants. Various versions of the debt/equity ratio were most common in these covenants, appearing 47 times. Interest cover appeared 6 times. Asset disposals, acquisitions, and balances as a proportion of debt and/or equity appeared 22 times.

Mohrman (1996) examined the covenants in 228 debt contracts issued in the US between 1963 and 1990. She found that 76% of them relied on financial statement numbers. Surprisingly, half of the financial statement numbers were based on “fixed GAAP provisions”, i.e. the accounting methods used to calculated the ratios used in the covenants are fixed in advance. Such a provision limits the issuers ability to use voluntary accounting changes to loosen covenant restrictions. It also eliminates the impact of mandated accounting changes on the covenant restrictions.

A conclusion which can be drawn from the aforementioned studies is that the variation in contractual covenants across classes of debt financing can be explained by contracting theory. In particular, they show that the higher the rank of the financial instrument, the greater the covenants constrain management’s choices. These studies also suggest that accounting information plays an integral role in the composition of contractual covenants. Finally, they intimate that the use of covenants (particularly accounting-based ones) may be declining.

In this paper, we undertake a similar study using recent issues of Canadian preferred stock. Specifically, we test and find support for the predictions of costly contracting theory, but illustrate that accounting variables play virtually no role in the covenants of these recent issues.

In the costly contracting framework, covenants are negotiated into contracts to resolve anticipated conflicts of interest between claimholders. A covenant is a formal agreement which binds the corporation to perform or not to perform certain actions while the preferred stock or debt is outstanding. The costly contracting theory suggests that financial contracting is costly due to contract negotiation and enforcement and to activity monitoring, but that the use of financial contracts can increase the value of the firm. The costly contracting theory would predict (Stokes and Whincop 1993, 465) that holders of debt-like preferred stock rely more heavily on the covenants in their contracts than holders of common equity-like preferred stock to resolve anticipated conflicts of interest with management.[5] This anticipated conflict with management will be discussed subsequent to an explanation of the debt-like and common equity-like characteristics of preferred stock.

Debt-like attributes include entitlements to cumulative dividends (in arrears), redemption privileges, retraction privileges, parental or third party guarantees, and purchase obligations. Common equity-like attributes, on the other hand, include conversion privileges and participation privileges.

An entitlement to cumulative dividends implies that dividends not paid on any specified dividend payment date must be made up on a subsequent dividend payment date. When cumulative preferred shares are outstanding, a firm is prohibited from paying dividends on common shares until all dividends in arrears are paid in full. A redemption privilege entitles the company to call, on notice, preferred shares for repurchase at par or at a premium over paid-up value. The premium required on redemption usually reduces to nothing over time. A retraction privilege allows the holders of the preferred shares to require redemption of the shares on a specified date or after a specified date at a determined price. With a parental or third party guarantee, the firm’s parent or other affiliate guarantees the dividend or retraction payments. Finally, purchase obligations require the company to purchase for cancellation in the open market a specified number of shares each quarter or year at or below a specified price. These features make preferred stock more like an issue of debt than common stock as issues of debt are generally entitled to a fixed return over a fixed term and to the repayment of the initial investment on the date of maturity.

Alternatively, conversion and participation privileges make the preferred stock more like an issue of common stock than debt. A conversion privilege gives preferred shareholders the option of exchanging their preferred shares for common shares of the firm at a specified conversion price during a specified period of time. A participation feature entitles preferred shareholders to share proportionately with common shareholders in any profit distributions beyond the prescribed amount.

Conflicts with management derive from management’s role as an agent for the owners of the firm - the common stockholders. Preferred shareholders, like debt holders, are concerned that management, on behalf of the common shareholders, will take action to reduce the likelihood that they will receive their promised cash flows. For instance, management could issue preferred shares, distribute the proceeds as dividends to the common shareholders, and then invest in very risky investments. Management could also issue additional debt or preferred stock of equal or superior rank to dilute the claim of the incumbent security holders.

Holders of debt-like preferred stock are more prone to conflicts of interest with firm owners than holders of common equity-like preferred stock because the common stockholders have relatively more to gain from taking actions which dilute the claims of preferred stockholders with debt-like interests compared to those with common equity-like interests. Holders of preferred stock with common equity-like characteristics are better able to protect themselves from abusive actions of the common shareholders. For example, a holder of a convertible preferred share could always convert if management decides to act solely in the interest of the common stockholders.[6]