Models of security markets invariably assume that market participants act intelligently since any other assumption generally leads to absurd results. Schaefer's Comment makes the incorrect assumption that one market participant, namely the government, acts unintelligently by allowing investors to reap tax arbitrage profits. Consequently, the conclusions in the Comment are erroneous. As I will show below, a government can (and the U.S. Government does) prevent tax arbitrage and any resulting disequilibrium in the bond market through two simple tax rules. My earlier paper [1] primarily showed that bond price must be a linear function of coupon for any maturity under the assumption of constant tax brackets for all investors, and then suggested that a linear relationship will also hold with investors in different tax brackets. I will prove this last result below.

Markets for exchange typically evolve over time, and the organized security exchanges in the United States have evolved in such a manner as to give specialists a major role as market makers. Specialists are responsible for maintaining a fair and orderly market in the securities listed on an exchange. They specify prices to reflect supply and demand and adjust these prices in response to shifts in supply and demand.

In a recent paper [3], Livingston analyzes the relationship between bond prices and market discount rates in a market with taxes, but which is otherwise frictionless. Two main cases are addressed. In the first, all investors are assumed to be in the same tax bracket and a relationship [17] is derived between the price of a bond, the rates of tax on income and capital gains, and the aftertax term structure. In a subsequent section, it is claimed that, when the assumption of a common tax bracket is removed, but maintaining the frictionless market assumption, the analysis is more or less unchanged:

”..there must still be a unique price of an annuity of maturity j and a unique price of a discount note of maturity j. There will be implicit tax rates in these annuities and discount notes for each maturity.… The bond pricing equation is the same except that TP and TG are replaced by TPj. and TGj., implicit marginal tax rates for each maturity” [3, p. 20].

Professors Sorensen and Hawkins (hereafter SH) have utilized regression analysis to examine the pricing of preferred stocks both before and after a particular event. This event, the NAIC event, occurred in 1979 when the National Association of Insurance Commissioners (NAIC) adopted a rule permitting insurance companies to carry sinking fund preferred issues at book value rather than at the market value required before. SH results indicate nine to 12 variables have a significant effect on the pricing of preferred stock.

In a recent paper [3] we developed a theory of financial intermediariesas information producers. We argued that financial intermediaries are one class of market participants who specialize in the production of information and sell that informationtofirms with investments to finance which can profit from its distribution. In order to focus on financial intermediation, we sacrificed considerable generality in our model ofcompetitive information production in capital markets. In particular we assumed that there were only two types of assets available in the market known as type A and type B firms. Second, we assumed that information would be produced about all assets by only one information producer or that information was a declining cost industry. The approach we utilized is actually a special case of what has been called a screening or certification process (see Stiglitz [6] and Viscusi [7]), and what we prefer to callsorting. The sorting market is a market where high value and low value assets are distinguished by market participants who specialize in sorting. It is similar, but not identical, to the process of signaling developed by Spence [5] and applied to financial markets by Ross [4] and Bhattacharya [1]. Yet, the extant literature on sorting has, by and large, ignored the opportunity for thosewho stand to lose from sorting to offer side payments to thwart the sorting process. While the problem of side payments may be of minor significance for some applications of sorting and signaling models, the prospect of side payments appears to be an important if not crucial issue infinancial markets.

Historically, preferred stock ranks well below common stock and bond issues as a source of financing in the private capital market. Relative to issuance of bonds, the issuance of preferred stock burdens the issuer with a fixed financing cost without benefitingthe issuer with the tax deductability associated with interest payments.

As is shown in the financial literature, the assumption of perfect and costless information often leads to the conclusion that corporate financial decisions are inconsequential to the value of the firm (e.g., Stiglitz [20] and Fama [6]). This conclusion seems to be in direct contradiction with the observed behavior of corporations that allocate real resources to implement financial policies. The gap between theory and observed behavior is bridged by introducing various frictions and market imperfections. A growing number of studies examine the optiraality of financial decisions when the assumption of perfect and costless information is replaced by allowing for informational asymmetry. The asymmetry is assumed to exist between corporate insiders who possess superior information about the firm's future earnings prospects and outside investors. The emphasis in this literature is on the ability of financial instruments to serve as signaling devices through which the true value of the firm can be revealed to the market without moral hazard or disclosure of confidential information. Although the signaling process is typically considered to be costly, it is advocated that firms may be better off if they employ this mechanism rather than reveal reliable, but confidential information, or not disclose at all.

The use of third-party quality certification, or sorting, to overcome asymmetrically informed markets was examined recently by Viscusi (Bell Journal, Spring 1978, pp.277–279) in a labor market context. Campbell and Kracaw (CK) have extended the Viscusi equilibrium and applied it to the potentially more interesting financial markets context. CK examine the incentives for low-quality firms to offer side-payments (a financial economists' term for bribes) to information producers, or sorters, to prevent resolution of the information asymmetry. Introducing bribes formally into the analysis maylead to a better understanding of some financial institutions; an earlier CK paper is such an examination (Journal of Finance, September 1980, pp. 863–882). Perhaps, in other applications of the sorting model, other previously unjustifiable financial activities and their associated payments might be seen as side-payments for some sorting activity.

A great deal has been written about the existence of a multi-tiered stock market while little is known about the effects [8], Elia [16,17,18,19,20,21,22], Freund [25], Farrar [23], Klemkosky [29,30], Loomis [32], Robbins [47], Rosenberg [48], Seligman [5], Smidt [55], Soldofsky [56], West and Tinic [67], and Schultz [50]). More recent discussions have considered the current nature of the tiered market (Welles [65,66], Carson-Parker [11], Ang [1], Marcial [36,37,38], Lurie [34], Janeway [27], Buhl [10], and Loomis [33]). While changes may have occurred, we believe a tiered market exists and will continue to influence trading and relative pricing (Elia [17,18,19], Marcial [37,38], and Reilly [45]). Because a multi-tiered stock market will probably continue, it becomes important to determine the effects of the tiered market on the securities and firms involved. Specifically, this paper examines common stocks in one of three market tiers (based on various measures of size), in terms of trading activity, price volatility, and financing characteristics during the 15-year period 1964–1978. The total period is divided into three subperiods representing periods of increasing trading activity by institutional investors. Specifically, the first period is generally prior to the institutional impact, the second is a transitional period, and the recent period is when institutions havebecome the dominant trading group.

The paper by Frankfurter and Hill is an interesting contribution to our understanding of the interrelationship between funding and portfolio investment decisions in pension fund management. This research summary is written with grea clarity, and the first three sections will be must reading for anyone doing serious work on the theory of corporate pension fund management. The reader would be well advised to go ahead and read the fourth section (“a numerical demonstration”) as well, for there theauthors clearly demonstrate both the power and serious weaknesses of a linear programming approach to this multiperiod, fixed horizon problem.

Controversy over the implications of debt and the rationale belying capital structure has seemingly come to rest upon a plateau defined by Miller's equilibrium analysis of aggregate corporate debt [10]. In “Debt and Taxes,” Miller reasserts his contention that whether capital is obtained through debt or equity has no bearing on the market value of the firm and is, therefore, irrelevant--a notion which has long been accepted with some reluctance by the finance academe. When the Modigliani-Miller model was first offered some 20 years ago [11], it was accompanied by a set of assumptions which portrayed the world of corporate finance in such malleable terms as to make the irrelevancy propositions palatable. Adaptation of this theoretical model (by its originators) to its secular counterpart through the imposition of corporate taxes [12] brought about a reassuring reversal of the irrelevancy doctrine, but left in its stead the disconcerting prescription to maximize firm value by financing exclusively via debt. Consideration of tax effects at the personal level by Farrar and Selwyn [7] marked the next concession to reality by capital structure theorists. Instead of alienating the original model still further from observed corporate behavior, this step provided a means of reconciling the overwhelming advantage of debt financing at the corporate level with the ultimate after-tax “consumption possibilities” afforded to individual investors. Miller's analysis explains that corporations are forced to “gross up” nominal interest rates to attract bondholders who must be compensated for their personal tax liability [10]. Potential increases in market value due to the taxdeductibility of interest payments are exhausted in the competitive drive toward equilibrium—at which point there are no gains from leverage. The sanctity of the irrelevance theorem thus appears to have been restored at the aggregate level.

The paper by Lakonishok and Sadan represents an interesting attempt to link major economic reform to fundamental issues in financial valuation. Changes instituted by the Israeli government in the fall, 1977, are related to subsequent changes in security returns on the Tel Aviv Securities Exchange. The authors employ an event study methodology to investigate the market reaction to the announced reform, the market reaction to raw data and partially analyzed data, and the market reaction to easily acquired data and less easily acquired data.

In this paper Ho and Saunders apply a model that has been used to analyze dealer spreads to banking.

A potential contribution to a field can arise whenever a well-developed framework of analysis in one problem area is applied to another field. The risk of a mechanical application, however, is that the institutional structure of the two problem areas is so different that no real insights are gained. What are the facts here?

It is important to investigate the characteristics of firms in differing size categories (tiers) in order to document any size-related differences that may exist among these firms. Such investigations may lead to improvements in asset-pricing models as empiricists test the impacts of these differences on the actual pricing of securities. Reilly and Drzycimski (R & D) have provided evidence that a number of such differences exist: volatility, debt ratios, and trading volume appear not to be homogeneously distributed across firm size categories nor across time, but there are some difficulties with the paper that need airing.