1 Commissions based on pre-December 5, 1968 schedule. 2 Cost estimates as calculated by National Economic Research Associates, "Stock Brokerage Commission: The Development and Application of Standards of Reasonableness for Public Rates" (July 1970), vol. II, table XI-2. Costs based on survey for 1969. During the fixed commission era, profits on large trades were substantial compared to those on small trades because costs did not rise with transaction size as fast as commission revenue did. For example, prior to December 5, 1968, commissions on 10,000 shares of a $10 stock were $3,900, which is 100 times the commission on 100 shares. Costs were not 100 times as great. Table 2 reports estimates of profits on transactions of various sizes on a $40 stock. The estimates were prepared by National Economic Research Associates, a consulting firm employed by the NYSE to determine reasonable commission rates." Cost estimates are based on a survey of firms for 1969. For the purposes of this discussion, concern is less with the absolute level of costs (which appears to be too high in view of the fact that firms were accepting small orders) and more with the relationship among costs of orders of different sizes. Clearly, large orders were extremely profitable relative to small orders. 13 NERA, "Stock Brokerage Commissions: The Development and Application of Standards of Reasonableness for Public Rates" (July 1970). TABLE 3.-COMMISSIONS ON LARGE ORDERS OF STOCK ON THE NEW YORK STOCK EXCHANGE 60 Mar. 30, 1959 to Dec. 5, 1968. Dec. 5, 1968 to Apr. 5, 1971.. Apr. 5, 1971 to Mar. 24, 1972. 44.9 28.9 24.8 20.9 18.6 17.9 .75 .75 .75 .75 .75 45.0 45.0 45.0 45.0 45.0 45.0 .75 .59 .57 56 .55 .55 45.0 35.4 34.2 33.6 33.2 33. 1 Mar. 24, 1972 to Sept. 25, 1973. 1 Commissions on portion of order over $500,000 negotiated as of Apr. 5, 1971; on portion over $300,000, negotiated as of Apr. 24, 1972, Commissions on orders involving negotiations based on NYSE membership survey of discounts on negotiated portion. For the purposes of this table, the change on Apr. 24, 1972, assumed to occur on Mar. 24, 1972, the date of a general commission rate structure change. Source: NYSE. .31 .29 45.0 35.4 31.3 23.5 19.0 17.5 .37 31 .29 44.2 31.2 27.6 22.0 18.7 17.5 .41 .54 54 47.0 43.8 43.4 43.2 43.1 43.0 35 30 .29 47.0 43.8 35.0 27.9 23.7 23.1 33 .29 .28 52.2 39.2 31.9 26.4 23.4 22.2 38 35 60.0 41.3 34.1 30.7 28.7 28.0 43 40 39 64.8 44.9 37.7 34.2 32.2 31.5 49 49.0 49.0 49.0 49.0 49.0 49.0 49 49.0 49.0 49.0 49.0 49.0 49.0 30 49.0 49.0 36.4 30.1 26.3 26.4 31 28 27 60.2 47.2 36.3 31.0 28.0 27.0 36 35 34 69.2 47.0 40.0 36.6 34.7 34.0 41 39 38 74.7 51.4 44. 4 41.0 39.0 38.4 With the growth of institutional trading, large orders became even more frequent, and the institutional side of the business became extremely profitable. The volume discount instituted on December 5, 1968 reduced these profits, but not substantially. Further reductions were accomplished by the SEC mandated requirement that commissions be negotiated on that portion of the order over $500,000 (as of April 5, 1971) and later (April 24, 1972) on that portion of the order over $300,000. Commissions as a percent of value and in cents per share are shown in Table 3 for orders of various sizes and stocks of various prices. The trend relative to commissions on round lots is reflected in Figure 1. Commissions on transactions exceeding $500,000 after April 5, 1971 and exceeding $300,000 after April 24, 1972 are based on quarterly surveys of members conducted by the NYSE which indicated that the discount on the portion of the order that was negotiated was on the order of 30 percent off the old (December 5, 1968) commission. The difference between commissions on large orders and sma!! orders as reflected in Figure 1 is quite startling. What is perhaps more startling is that further reductions in commissions on large orders would prove possible. 14 Limited membership Membership in the NYSE is limited to 1,366 individuals each of whom owns a "seat." Membership entitles one to do business on the floor of the NYSE, to pay a share of the costs of running the exchange, and to own a prorata share of the assets of the exchange-primarily the real estate and building. The 1,366 members are grouped into about 500 firms, of which about 410 do business with the public. Figures on the number of firms and on the allocation of the 1,366 members by primary business of the member are given in Table 4 for recent years. There has been a steady decline in the number of firms. Categories of members which have increased in number are floor brokers and registered floor traders. The number of specialists has fluctuated between 377 and 392. An important characteristic of NYSE membership is their salability. This permits more efficient firms to expand by buying seats from less efficient firms and tends to ensure that the level of services rendered is rendered efficiently (even though prices may be too high and the level too low). End of month seat prices reached a high point of $500,000 in 1929,15 fell to a low of $18,000 in 1942, reached an all time high of $515,000 in 1968 and 1969 and have fallen since then, to $83.000 as of July 1976. Recent monthly seat prices are in Appendix Table A. 14 I am grateful to the NYSE for supplying these data. TABLE 4.-NUMBER OF NYSE MEMBER FIRMS, MEMBERS, AND TOTAL PERSONNEL OF MEMBER FIRMS The market price of a seat depends on the monopoly profits (over and above all costs and a normal profit) that a potential buyer could earn. Annual profits are (P-C) V where P commission in cents per share C=average cost of supplying brokerage services in cents per share. Includes a normal profit. V=number of shares traded by seat holder. The price of a seat is the discounted value of expected profits (plus the value of NYSE property). Since profits are related to the volume of trading and commission rates, seat prices have tended to be related to volume of trading and commission rates because individuals base their expectations of the future on current events. Only the recent advent of competitive commission rates has broken this association. Figure 2 plots a simple form of this relationship; namely, seat prices and contemporaneous dollar volume of trading. Each variable is an annual average of monthly values. The association over time is quite close until 1970 when seat prices and volume diverge. At that time there was good reason to expect the introduction of negotiated rates which would tend to eliminate monopoly profits and therefore reduce the value of seats.16 16 More elaborate models of the determinants of seat prices are in Doede (1967) and Schwert (1975). Since new members must pay to join the cartel an amount equal to the capitalized value of monopoly profits, the only clear beneficiaries of a cartel are the original founders. New members benefit if they are more efficient than those selling a seat. The resistance to competitive rates is easily understood when one recognizes that many members bought seats at price ranges as high as $515,000 (in 1968). Elimination of fixed commissions imposes a windfall capital loss. (A large portion of that loss is realized when the intention to have competition is announced.) Rules and regulations A cartel is an agreement among many firms to raise prices and limit output. Since a monopoly profit is earned on each transaction, each member of the cartel has an incentive to expand output in order to increase his level of monopoly profits. Rules prohibiting price cutting and limiting output are therefore generally found in cartels. On the NYSE, rebating any part of the commission to a nonmember was prohibited. Rule 394 prohibited members from trading away from the |