Printer Friendly

Fiduciary decision making and the nature of private pension fund investment behavior.

Fiduciary Decision Making and the Nature of Private Pension Fund Investment Behavior

ABSTRACT

The Employee Retirement Income Security Act of 1974 (ERISA) was enacted in

order to protect and enhance the welfare of employees and others covered by private

pension plans. In order to achieve this objective, ERISA required pension plans to vest

the accrued benefits of employees with significant periods of service, meet minimum

standards of funding, guarantee the adequacy of the plan's assets against the risk of

premature plan termination, and change the legal status of plan administrators from

that of nonfiduciary agents of the sponsoring firm to that of fiduciaries whose primary

responsibility is to plan beneficiaries. This study considers the extent to which changing

the legal status of plan administrators may have had incidental, and perhaps

unintended, side effects. Empirical evidence suggests that, since the enactment of

ERISA, private pension fund portfolios have underperformed portfolios held by

comparable investment funds not subject to ERISA. This study analyzes the extent to

which this change in legal status, which holds administrators to a higher standard of

conduct and subjects them to increased liability exposure, may have contributed to this

underperformance phenomenon.

Introduction

This study analyzes the implications of the change in the legal status of private pension fund administrators from that of nonfiduciary to fiduciary economic agents that resulted from the enactment of ERISA. The analysis focuses on the manner in which differences in the legal/institutional environments within which fiduciaries and nonfiduciaries must operate lead to differences in their incentives, and in turn, their behavioral tendencies. As a result, to the extent that ERISA altered the status of private pension fund administrators from that of nonfiduciary agents of the sponsoring firm to that of fiduciary agents whose primary responsibility is to plan beneficiaries, differences in pre- and post-ERISA private pension fund investment behavior may be attributable, in part at least, to this change.

Several previous studies that have empirically studied the relationship between the enactment of ERISA and the change in the nature of private pension fund investment behavior have reported finding a statistically significant effect. For example Cummins and Westerfield [4] found that post-ERISA private pension fund investment portfolios contain a greater percentage of lower risk, lower rate of return securities than did pre-ERISA securities. More recently, Berkowitz and Logue [2] evaluated the investment performance of corporate pension plans in an attempt to determine, inter alia, if ERISA has had a significant impact on plan performance. The data revealed that private pension

plans have consistently underperformed endowment plans (not governed by ERISA) during the sample period (1968-1983); however, when the time series is divided into pre- and post-ERISA periods, a significant increase in the relative magnitude of the differential is discernible.(1)

Given the similarity of investment objectives of private pension plans and endowments, how might this marked discrepancy in relative performance be explained? One possible explanation is that the change in the legal status of private pension plan administrators, brought about by ERISA, may have exacerbated the discrepancy by altering the incentive structures under which they operate. In enacting ERISA, Congress attempted to protect interstate commerce and the interests of private pension plan participants and their beneficiaries by, inter alia, establishing standards of conduct, responsibility, and obligations for fiduciaries of employee benefit plans and by providing for appropriate remedies, sanctions, and ready access to the federal courts [ERISA Section 2]. The legislation sought to accomplish those objectives by changing the legal status of fund administrators from that of nonfiduciaries to fiduciaries.

It is important to note that this statutory change did not impose any direct constraints on the opportunity set of investment possibilities for plan administrators. Hence, in the absence of modifications in the plan administrators' incentive structure, there is no theory to explain why the performance of post-ERISA private pension plan portfolios should be inferior to that of pre-ERISA portfolios.(2) This study provides such a theory by demonstrating how this post-ERISA underperformance phenomenon may follow as a result of the change in incentive structures of private pension fund administrators brought about by altering their status from that of nonfiduciaries to fiduciaries.

Differences Between Fiduciary And Nonfiduciary Agents

The delegation of authority to act on behalf of another is central to the creation of any agency relationship, whether it is of a fiduciary or nonfiduciary type. Further, in both types of agency relationships the manner in which this delegated authority is exercised affects the economic position of the party or parties for which the agent is empowered to act. In nonfiduciary relationships, however, the costs and other economic consequences of the agent's performance tend to be borne by the contracting parties themselves, while in fiduciary relationships such consequences tend to fall on parties that were either not involved in the creation of the agency itself or are not in a position to insure adequate performance by contractual means alone.(3)

In order to gain greater insight into the differences in the tendencies of fiduciaries and nonfiduciaries, it is useful to model and analyze the decision problem faced by each type of agent. First consider the nonfiduciary. After negotiating an acceptable payoff schedule with the principal (I(*)), the nonfiduciary agent's problem is to determine which level of effort (e) to select.(4) Formally, the nonfiduciary's decision problem can be stated as(5) (1) Select e [Epsilon] Argmax [[integral of] U(I(y), e [prime]) [Phi] (y~e [prime], [Eta]) dy]

e [prime] [Epsilon] E where Argmax f(*) [approximately equal to] the set of variables which, for a function f(*), maximizes f(*); E [approximately equal to] the set of all possible levels of effort available to the agent; y [approximately equal to] the agent's level of performance as evaluated in terms of the contractually specified standards for evaluation; [Eta] [approximately equal to] the monitoring system utilized by the principal and the prime denotes the value maximizing the argument in the function. Under this formulation, the nonfiduciary's decision problem is to obtain the maximum possible payoff (I(y)) without expending an unacceptable level of effort (e). It follows that e is determined by the nature of I(*), the effectiveness of [Eta], and the nonfiduciary's incremental utility for additional compensation vis-a-vis his/her disutility for additional effort [Delta] U(I(y))/[Delta] U(e).

By contrast, the fiduciary's decision problem is complicated by the higher standard of conduct imposed, differences in the nature and source of monitoring, and the need to be able to justify decisions to two or more potentially competing parties. Due to this higher standard of conduct, and the rule against profiteering by fiduciaries, the fiduciary's payoff or compensation schedule (I(*)) is assumed to be fixed for a particular assignment. Therefore, the fiduciary's decision problem reduces to one of selecting the set of actions or arbitration schedule (a) that will provide the minimum likelihood of facing a successful legal or other challenge (c) by either an interested third party monitoring the situation or one of the parties to whom a fiduciary duty is owed, without expending an unacceptable level of effort. Formally, this decision problem can be stated as follows (2) Select a [Epsilon] Argmin [[integral of] [U.sub.d](c(y), a [prime])[Phi](Y~a [prime], [Eta])

dy] a [prime] [Epsilon] A where Argmin f(*) [approximately equal to] the set of variables which, for a function f(*), minimizes f(*); A [approximately equal to] the set of all possible actions or arbitration schedules available to the fiduciary; [U.sub.d] [approximately equal to] the fiduciary's disutility from assuming the risk of legal or other challenge; y [approximately equal to] the fiduciary's level of performance as evaluated in terms of prevailing professional or other standards for evaluating such actions or arbitration schedules; [Eta] [approximately equal to] the monitoring system(s) utilized by interested parties. Under this formation, the fiduciary's performance is evaluated in terms of its conformity to prevailing professional or other standards. Further, it is assumed that the fiduciary may expend additional effort to identify the set of actions or arbitration schedule (a) which is not consistent with prevailing professional and other standards of performance. Therefore, the selection of a is determined by the nature of the possible legal or other challenges available (c), the nature and effectiveness of existing monitoring systems ([Eta]), and the fiduciary's incremental utility for identifying an arbitration schedule less open to successful challenge vis-a-vis his/her disutility for additional effort [Delta] [U.sub.d](c(y))/[Delta] [U.sub.d](a).

Investment Implications Of ERISA

As just demonstrated, the performance incentives of pension plan administrators were substantially ameliorated by ERISA. Specifically, the incentives to maximize risk-adjusted rate of return was severely curtailed by altering the incentive structure of administrators. In this section, the effect of this radical change in incentive structures on portfolio selection decisions is analyzed. Since the performance criteria and incentives faced by fiduciaries tend to differ from those faced by nonfiduciaries, ceteris paribus, one would expect to observe corresponding differences in the pattern of investment decisions emanating from each type of agent. The pattern of investment decisions made by trusts, for example, tends to differ from those of nonfiduciary agents.(6) This difference suggests that since the institutional and legal status of pension fund administrators was changed by ERISA from nonfiduciary to fiduciary, one would expect to observe a corresponding change in the pattern of investment decisions emanating from this group. In this section the nature and direction of this change is considered.

To compare the incentives of a pre-ERISA (nonfiduciary) and post-ERISA (fiduciary) private pension fund administrator assume that the alternative investment opportunities available to the fund are a riskless asset yielding [R|.sub.F], a composite risky "market" asset yielding [R|.sub.M], and an asset yielding the allegedly superior rate of return [R|.sub.j]. Following Fama(5), it is assumed that [R|.sub.j] is determined according to the "market model"

[R|.sub.j] = [Alpha.sub.j] + [Beta.sub.j] [R|.sub.M] + [Epsilon|.sub.j] (3) [Mathematical Expression Omitted], where [Gamma.sub.j] = [Alpha.sub.j] - (1 - [Beta.sub.j])[R.sub.F],

[R|.sub.M] = E([R|.sub.M]) + [Epsilon|.sub.M],

E([Epsilon|.sub.M]) = E([Epsilon|.sub.j]) = cov([Epsilon|.sub.j], [R|.sub.M]) = 0.(7) This decomposition of [R.sub.j] is useful in determining whether asset j yields a superior return and in determining a measure of its superiority.

An asset yields a superior rate of return if the option to hold it, perhaps in conjunction with the riskless asset and the market asset, raises the fund's expected performance (in the risk-adjusted rate of return sense) over that possible from investments limited to the riskless and market assets. Thus [R|.sub.j] must contain elements of return which induce the fund administrator, given the incentive structure, to hold asset j in addition to investments yielding [R.sub.F] and [R|.sub.M]. An administrator of such a fund holding asset j must be compensated for the opportunity cost of identifying and evaluating such investments on behalf of the fund and for bearing the risk to his/her professional reputation associated with asset j. This risk associated with asset j is assumed to have two sources, [Beta.sub.j] [Epsilon|.sub.M] and [Epsilon|.sub.j]. The distribution of ([Epsilon|.sub.j], [Epsilon|.sub.M]) is multivariate normal. The source of uncertainty, [Beta|.sub.j] [Epsilon|.sub.M], depends upon the movement of asset j's return with the random market component, [Epsilon|.sub.M]. The amount of market risk in asset j is determined by [Beta.sub.j], the market sensitivity factor. Since the fund administrator can offset this risk by reducing the fund's holding of the market asset yielding E([R|.sub.M]) and instead purchase the riskless asset [R.sub.F], he/she is willing to bear the reputational and other risk associated with asset j if he/she receives additional compensation at the rate of [Psi] [[Beta.sub.j](E([R|.sub.M]) - [R.sub.F])] per dollar invested in asset j. ([Psi] is additional compensation that can be earned by a nonfiduciary administrator and is monotonically related to the rate of return earned on asset j as compared with the market return.) Hence, the term [Beta.sub.j](E([R|.sub.M]) - [R.sub.F]) in (3) is necessary to determine the possible additional compensation that might be earned by the fund administrator for bearing the risk associated with holding asset j, as opposed to holding only [R.sub.F] and [R|.sub.M].

The residual in the decomposition of [R|.sub.j] is [Gamma|.sub.j] + [Epsilon|.sub.j], where [Gamma.sub.j] is non-random and [Epsilon|.sub.j] is the source of risk specific to asset j. If [Gamma.sub.j] is more than sufficient to compensate the fund administrator for bearing specific risk [v.sub.j] (which denotes the variance of [Epsilon|.sub.j]), it is the source of the superiority of the portfolio containing asset j. The value of the opportunity to invest in asset j will vary positively with [Gamma.sub.j] and negatively with [v.sub.j]. Hence it is appropriate to refer to [Gamma.sub.j] as an index of excess return. It is assumed that [Gamma.sub.j] and [v.sub.j] are independent of the amount invested in asset j.

In order to analyze the investment implications of ERISA, the differences in the portfolios selected by fund administrators facing the decision problem specified in (1) with that specified in (2) are considered next. The administrator is assumed to be capable of assessing the expected return and risk of [R|.sub.m] and [R|.sub.j], but there are costs in the form of effort. It is further assumed that I(y) in (1) will increase if the portfolio selected outperforms the market and that c(y) in (2) will increase to the extent that the portfolio underperforms the average return earned by other private pension funds operating simultaneously. Finally, for simplicity, it is assumed that the administrator's decision reduces to a choice between a "safe" portfolio ([P|.sub.S]) containing only [R.sub.F] and [R|.sub.M] and a "risky" portfolio ([P|.sub.R]) which contains [R|.sub.j] as well. The administrator's choice of [P|.sub.S] or [P|.sub.R] can be viewed as a cost minimization problem in which he/she seeks to minimize portfolio selection costs and the opportunity costs. Portfolio selection costs include the cost of ascertaining the expected rate of return and risk associated with [R|.sub.M] and [R|.sub.j] and the cost of ascertaining the portfolio selection strategies being adopted by other similar agents. The opportunity costs include the costs due to loss of additional compensation that could be earned under the compensation agreement in effect and the cost of damage to reputation associated with a successful legal or other challenge. Since [P|.sub.S] and [P|.sub.R] differ only in terms of whether to include [R|.sub.j], the administrator will select [P|.sub.R] only if portfolio selection and opportunity costs are minimized by doing so.

Analysis of (1) and (2) reveals that the cost structures faced by pre-ERISA (nonfiduciary) and post-ERISA (fiduciary) administrators tend to vary significantly. First consider the (pre-ERISA) cost structure faced under (1). Since the performance evaluation criterion (criteria) is (are) contractually specified, and hence known to the nonfiduciary, the only significant portfolio selection cost that might be incurred is that of ascertaining the rate of return and risk associated with [R|.sub.M] and [R|.sub.j]. Correspondingly, the most significant opportunity cost faced by the nonfiduciary is that of not earning any additional compensation by outperforming the market.(8). Given this cost structure, the first decision faced by the nonfiduciary administrator is whether to incur the search and other costs necessary to ascertain the rate of return and risk associated with [R|.sub.j]. Since the most significant opportunity cost faced under (1) is not earning performance-based compensation however, the nonfiduciary's optimal strategy is to incur only enough selection cost to "document" that there is some basis for inclusion of [R|.sub.j] in the fund portfolio.(9) This investment in selection cost is only perfunctory since the nonfiduciary's optimal cost minimizing choice is to always select [P|.sub.R].(10)

By contrast, consider the (post-ERISA) cost structure faced under (2). Since there is no possibility of earning additional compensation by outperforming the market, the only relevant portfolio selection cost might involve ascertaining the portfolio selection strategies being adopted by other fiduciaries. Ascertaining the selection strategies being adopted by other private pension funds provides an indication of the criterion or criteria by which the fiduciary's performance will be evaluated by potential challengers and, thereby, helps to minimize the possibility that the portfolio selected will be successfully challenged because it underperforms the average return earned by other fiduciaries. Correspondingly, the only significant opportunity cost faced by the fiduciary is that of reputational damage associated with a successful legal or other challenge.

Under this cost structure, the first decision faced by the fiduciary is whether to incur the search and other costs necessary to ascertain the portfolio selection strategies being adopted by other private pension funds. Since the only significant opportunity cost faced under (2) is reputational damage, however, the fiduciary's optimal strategy is to incur only enough costs to "document" the general tendency of other private pension funds to avoid inclusion of [R|.sub.j] in their portfolios.(11) This investment in ascertaining alternative portfolio strategies is only perfunctory since the fiduciary's optimal cost minimizing choice is to always select [P|.sub.s].(12)

Conclusion

While ERISA was enacted in order to protect and enhance the welfare of employees and others covered by private pension plans, among the many changes brought about by this legislation was an alteration in the legal status of plan administrators from that of nonfiduciary agents of the sponsoring firm to that of fiduciaries whose primary responsibility is to plan beneficiaries. As the preceding analysis suggests, however, this change may have inadvertently induced plan administrators to adopt investment strategies that are unduly conservative and not necessarily in the best interests of either sponsoring firms or beneficiaries. This disturbing tendency on the part of plan administrators may be attributable to the fact that the incentives they now face as fiduciaries under ERISA are radically different from those they previously faced as nonfiduciaries.

It would be premature to conclude, however, that changing the status of plan administrators from that of nonfiduciaries to fiduciaries may not have had potentially compensating benefits for plan beneficiaries and others. This study has not, for example, considered the extent to which ERISA may have helped to restore public confidence in private pension plan administration by reducing the potential for sponsoring firms to use plan assets to further their self-interests at the expense of plan beneficiaries. There are, of course, criteria for assessing the consequences of ERISA other than its impact on the economic performance of plan investment portfolios.

These and other limitations of the study notwithstanding, the results suggest that even the most well-intended types of regulatory reforms may carry negative behavioral and other side-effects that need to be considered by legislators and other policy makers. By altering the incentive structure within which private pension fund administrators must operate, ERISA may have increased public confidence in such plans, but not without cost to plan beneficiaries and others.

(1)The arithmetic average of the pre-ERISA difference was only 0.28 percent, whereas post-ERISA differences averaged 1.85 percent. (2)Blair[3] has already demonstrated this analytically, albeit in an indirect manner, by modelling the impact of restrictions on asset choice for qualified pension plans. (3)Although there has been little formal analysis of fiduciary relationships in the economics literature, legal scholars have long been concerned with agency relationships of this sort. See e.g., Scott[8], Sealy[9], Weintraub[11], and Shepard[10]. (4)In order to achieve a more preferred probability distribution of investment returns, the principal may have to increase the agent's incentive to expend effort by increasing the agent's financial risk. (5)This formulation of the nonfiduciary's decision problem (equation (1)) is based on the standard agency model. It specifies that the nonfiduciary will choose that level of effort that maximizes his or her expected utility U(*), taking the principal's choice of the payment schedule and monitoring system as given. The nonfiduciary bases the decision on his or her beliefs represented by the probability distribution [Phi](*). For a discussion of the mathematical and other features of this model see e.g., Baiman[1] or Holmstrom[6]. The formulation of the fiduciary's decision problem (equation (2)) is, mathematically at least, essentially a modified version of the standard agency model. (6)One contributing factor that helps to make the investment decision of trustees different from those of nonfiduciaries is the fact that many trustees, such as those employed by full service commercial banks, are regulated by the U.S. Comptroller of the Currency. The regulations issued by the Comptroller of the Currency restrict the types of securities and funds that trust assets subject to their jurisdiction can be invested in. (7)Here [Alpha.sub.j] and [Beta.sub.j] are constants and [Epsilon|.sub.j] and [Epsilon|.sub.M] are random disturbances. (8)Another, less potentially significant, opportunity cost also faced by a nonfiduciary is the possibility of termination of the agency relationship by the principal. We assume, however, that such termination will occur only if the portfolio selected performs substantially below the market and the nonfiduciary cannot provide adequate support for his/her investment decisions. (9)This investment is intended as a form of insurance in case [R|.sub.j] performs so poorly as to cause the entire portfolio to perform substantially below the market. As discussed in note 11, documenting support of the inclusion of [R|.sub.j] helps prevent termination of the agency relationship even if such an eventuality occurs. (10)A proof of this result is available from the authors upon request. (11)The optimality of this strategy assumes that it is less costly to "document" the general investment tendency of other private pension funds than to ascertain the criterion or criteria by which potential challengers will assess fiduciary performance and then, in turn, ascertain the rate of return and risk characteristics of [R|.sub.j]. Evidence of performance in conformity with the prevailing standards of the profession in generally a credible defense to any legal challenge alleging malpractice or mismanagement. (12)A proof of this result is available from the authors upon request.

REFERENCES [1]Baiman, S., "Agency Research in Managerial Accounting: A Survey," Journal of Accounting Literature, (Spring 1982), pp. 154-213. [2]Berkowitz, Logue & Associates, Investment Performance of Corporate Pension Plans, report submitted to U.S. Department of Labor. [3]Blair, R. "ERISA and the Prudent Man Rule: Avoiding Perverse Results," Sloan Management Review, (Winter 1979), pp. 15-25. [4]Cummins, D. and R. Westerfield, "Patterns of Concentration in Private Pension Plan Common Stock Portfolios Since ERISA," Journal of Risk and Insurance, (September 1981), pp. 447-476. [5]Fama, E. "A Note on the Market Model and the Two-Parameter Model," Journal of Finance, (December 1973), pp. 1181-1185. [6]Holmstrom, B. R., "Moral Hazard and Observability," The Bell Journal of Economics, (Spring 1979), pp. 74-91. [7]Prosser, W. Handbook of the Law of Torts, 4th edition, (St. Paul: West Publishing Co.). [8]Scott, "The Fiduciary Principle," California Law Journal, Vol. 37, pp. 539-559. [9]Sealy, "The Fiduciary Principle," Cambridge Law Journal, (1962), pp. 69-88. [10]Shepard, J. "Towards a Unified Concept of Fiduciary Relationships," Law Quarterly Review, Vol. 97 (January 1981), pp. 51-79. [11]Weintraub, E. "The Fiduciary Obligation," University of Toronto Law Journal, Vol. 25, pp. 1-22.

Steven B. Johnson is Dean of Business and Law, Bond University, Gold Coast, Australia. Jack L. VanDerhei is Associate Professor of Risk Management and Insurance at Temple University.
COPYRIGHT 1989 American Risk and Insurance Association, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Johnson, Steven B.; VanDerhei, Jack L.
Publication:Journal of Risk and Insurance
Date:Mar 1, 1989
Words:4001
Previous Article:Risk aversion and the value of information.
Next Article:Risk as a discount rate determinant in wrongful death and injury cases.
Topics:

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters