APPENDIX B

Economic Aspects of Institutional Exchange Membership

Irwin Friend

University of Pennsylvania

There is a strong economic presumption in favor of competition in the market for securities as in other markets so long as adequate disclosure and any needed safeguards against conflicts of interest and possible fraud are provided. For this reason, there is a burden of proof on any position that argues against permitting financial institutions to become exchange members. This is especially true of any attempt to bar institutional entry to membership on regional exchanges where it is now permitted. Such a change would represent a step backwards towards a less competitive market. Moreover, in the absence of evidence to the contrary, permitting institutional membership on the New York Stock Exchange would be expected to further increase the competitiveness and, as a result, the efficiency of the stock market.

This paper will examine the major arguments which have been made against institutional exchange membership. As a background for assessing the validity of these arguments, the paper will analyze the basic economic and related non-economic considerations involved in the issue of institutional membership and present the relevant empirical evidence. However, before proceeding with this analysis, it may be useful to define what is meant by institutional exchange membership and to clarify a common confusion about the dangers associated with such membership.

An institutional member in this paper will refer to a stock exchange member firm which is affiliated with a bank, pension fund, investment company complex, insurance company, or the manager of a similar pool of invested capital. The member of the Exchange would normally not be the institution itself, but rather a separately organized subsidiary or affiliated corporation. The affiliated member might only execute the institution's own transactions, or it might deal directly with the public, or it might do both. Not all of the arguments in favor of institutional membership apply equally to both of these kinds of activity, but the analysis in this paper indicates that there is no substantial reason for prohibiting either type.

In discussions of the desirability of institutional membership, one common fallacy is to confuse the alleged dangers of a growing institutionalization of stock market trading with those dangers which allegedly might be associated with a growing institutional role in brokerage activity. It is only the second of these "dangers" which is relevant to the issue of institutional membership. There is no reason to expect that the lower brokerage costs associated with institutional membership are likely to accelerate substantially the sales of beneficial interests in the institutions themselves, since sales of the latter are generally considered to be relatively insensitive to small changes in management costs (including brokerage). Nor is there any reason to believe that institutional membership will be associated with a significant increase in portfolio turnover-a point developed later in this paper. The increasing share of institutional trading in the market has proceeded historically and can continue to proceed without any substantial institutional membership. However, this paper concludes that neither the institutionalization of the market nor institutional membership has had or is likely to have a seriously adverse economic effect.

The case against institutional membership must rest on the impact of such membership on (1) market efficiency; on (2) the different parties at interest in the market (i.e., the main groups of brokers and investors who may be affected); and perhaps on (3) broad social policy issues which transcend the market (assuming for the moment that there are such issues which are affected by institutional membership). From an economic viewpoint, the first consideration is the most basic but the second is also of great importance for regulatory purposes since it involves fundamental questions of equitable arrangements among different participants in the market. The only significant non-market policy issue which has been raised in connection with institutional membership is its implications for concentration of power.

Impact of institutional exchange membership on market efficiency

There are two different aspects of market efficiency: operational efficiency or the minimization of transactions cost (commissions and bidask spreads) to the public for the same services; and allocational efficiency or the extent to which markets maintain equivalent rates of return (and hence costs of financing) on investments with comparable risks. It is the latter quality of markets which would ensure that funds are channeled from savers to those users who will apply them most profitably and that portfolio shifts can be made to the mutual advantage of different investors.

The basic mechanism which economic theory indicates is necessary to attain efficiency is free competition. Institutional membership enhances the degree of competition in the provision of broker-dealer services on exchanges. It could provide substantial additional long-term capital to an industry whose permanent capital base is generally regarded as inadequate. The industry has recently required the dramatic infusion of capital from outside sources and is likely to need large additional amounts in the future. A market which is becoming increasingly institutional requires more capital for a given volume of trading activity to handle the growth in the size of orders.

In addition to providing needed new capital, institutional membership would permit taking advantage of any joint economies among such activities as financial management, distribution of beneficial interests in the institution, and securities brokerage. Such economics may lower costs in-and thereby lower prices charged for some or all of these activities. Institutional membership would also permit a diversification of business interests which would lower risk in the combined enterprise, hence again potentially leading to lower prices charged for financial services. Expanded institutional membership would discourage the inefficient use of large numbers of brokers for execution of institutional orders as compensation for services or benefits unrelated to the efficient execution of orders.

The argument that institutional membership could lead to a market dominated by dealers dealing for their own account with presumably adverse consequences for market efficiency (and perhaps intensifying conflicts of interest) seems without justification. It is not institutional membership but the growth in institutional trading and large blocks which has stimulated dealer activity. Moreover, the Institutional Investor Study (p. 1856) suggests that the activity of market makers has been generally stabilizing, contributing to rather than detracting from market efficiency. (The question of conflicts of interest will be treated later in this paper.)

Attempts to exclude institutions from exchange membership will only divert trading to the 3rd and 4th markets, reducing market liquidity and efficiency at least until that point of time when all stock quotes and transactions, including those on the over-the-counter markets, may appear on a central tape (perhaps an expanded NASDAQ).

The argument that institutional membership would drive out the individual investor, with a consequent loss of liquidity, also seems unfounded. Again it is not institutional membership but the vast growth in institutional stock ownership and trading which has led to a reduction of the relative importance of the individual investor. Even so, there has been a virtually uninterrupted growth in the absolute number of individuals owning stock directly in U.S. corporations. It has been asserted that institutions would turn over their investment portfolios much more rapidly (particularly to realize capital gains) if they did not have to pay relatively high commissions every time they bought or sold a stock and that the greatly expanded volume of large blocks would cause sharp market fluctuations, both harming the small individual investor and deterring him from trading. This line of reasoning seems dubious because the incentive to excessive trading by non-brokerage institutions in the past was not exchange membership but its absence, with the existence of surplus brokerage a potentially important factor motivating trading for give-up purposes. The assumption that institutional members returning commission income to the institutions would be more likely to churn portfolios than in a system of non-institutional membership without such a return of commission income is especially questionable. The two largest institutional exchange members which return virtually all commissions, the Investors Diversified Services groups of funds and Prudential Life Insurance, do not evidence higher turnover than the averages for their industry groups. Prudential exhibited substantially lower rates of stock activity than the average life insurance company during 1970 and the first half of 1971. During this same period the Investors Diversified funds had a stock activity rate roughly the same as the average mutual fund. Neither Prudential nor Investors Diversified significantly changed their stock turnover relative to other institutions after becoming exchange members.

As a matter of fact, all broker-affiliated institutions, including those which did not return any commissions, had only slightly higher than average turnover ratios historically when other institutional characteristics including size are kept constant. (See Institutional Investor Study and Wharton School Mutual Fund Study.) For most institutional groups, the effect of brokerage affiliation on turnover has been statistically insignificant. The only notable exception has been individual and personal trust accounts advised by advisory affiliates of firms doing a brokerage business, and even here the difference from the industry average is quite small and in any case these are hardly what is normally meant by institutional members.

Similarly, it is difficult to rationalize the suggestion that, in view of the "danger" that institutional membership poses to individual trading activity and the supposed dependence of market efficiency and the efficient execution of institutional trades on such activity, institutional investors should subsidize the small individual investors. Even if we were to assume that institutional membership poses a real danger to individual trading which seems questionable, there is little basis for the supposition that market efficiency and the efficient execution of institutional trades depend

on such trading. It is true as pointed out by the Institutional Investor Study that a significant volume of block trades are offset by individual trading, but the Study also indicated that in block trades of $1 million and over probably close to 70% of the total layoffs of block positions are to institutions (P. 1608).

Much more important, there is no reason for believing that market efficiency, and hence the quality of market prices paid or received by either institutional or individual investors, is (or will be) impaired by institutional trading. Thus, despite the greatly increased stock market activity by institutional investors over the past decade or so, a recent Twentieth Century Fund Study (Irwin Friend, Marshall Blume and Jean Crockett, Mutual Funds and Other Institutional Investors: A New Perspective, McGraw Hill, 1971), could detect no changes in the allocational efficiency of the market over this period (pp. 91-94). To assume thai this situation is likely to change in the future is to imply that institutions are less likely than small individual investors to make informed decisions when they buy or sell stock, i.e., institutions have less insight into intrinsic stock value than the rest of the market-which would appear to be an implausible assumption.

While there is no reason for believing that the basic allocational efficiency of the market is impaired by expanded institutional trading, there is some evidence both in the Institutional Investor Study and in a more recent study (Alan Kraus and Hans R. Stoll, Price Impacts of Block Trading on the NYSE, Rodney L. White Center for Financial Research, University of Pennsylvania, 1971) that sales of large blocks may have a small adverse effect on market price in the sense that prices recorded on the ticker temporarily fall. (Purchases of large blocks apparently do not have a corresponding temporarily stimulating effect.) The Kraus and Stoll study (pp. 1-28) suggests that the adverse price effect of sales of large blocks may average somewhat over .7 of 1% but since only blocks associated with down ticks are included and total sales of the stock are not held constant even this relatively small figure may overstate the block effect. Moreover, the small temporary discount necessary to bring in willing buyers quickly seems to be largely dissipated by the end of the day of the block trade.

This effect on price may be regarded as an increase in transactions costs to the institutional sellers of these blocks necessary to mitigate a short-term supply-demand imbalance. However, while the institution may have to pay a small cost for speedy execution of a large block, there is no significant harm to other investors. Investors buying the stock from institutional sellers on a day of a block trade-a period when individuals and other non-institutional investors are presumably net purchasers of the stock-benefit from the temporarily depressed prices.

Impact of institutional membership on different groups

The benefits of exchange membership to institutions and their beneficial owners of relatively small means are fairly obvious. These include the reduc

1. The Institutional Investor Study concluded that registered investment companies and insurance companies advised by advisory affiliates of brokerage firms had somewhat higher turnover rates than the industry averages, but the T values indicate that these small departures from the averages are statistically insignificant (p. 190). The Wharton Study concluded that the largest mutual funds with brokerage affiliations generated turnover rates lower than the comparable industry group in all six years covered (p. 225).