qualities in those stocks and during those periods when they are most needed, rather than fortuitously or when and where least needed. Improving specialist capital requirements, for instance, or assigning floor traders responsibilities as "auxiliary specialists," would constitute more direct approaches to these problems, and the latter approach would enjoy the further merit of tempering special market advantages with definite market obligations.

Attempts to retain or expand the benefits of floor trading and at the same time curtail its undesirable characteristics have been nominally successful at best. In 1945 the Commission proposed the abolition of floor trading, but withheld action in light of repeated assurances that the exchanges would develop effective self-regulation of this activity. Despite the great variety and complexity of exchange rules experimented with to date, however, floor traders still retain their significant private trading advantages in a public market, continue to concentrate their activities in the more active stocks, and continue to accentuate price movements.

Self-regulation in this particular area has not only been generally ineffective, but in a most important respect it has been misdirected. The exchanges' regulation of other categories of members on the floor is generally to assure adherence to obligations designed to benefit the market. In the case of the floor trader, on the other hand, the exchanges have established elaborate rules and complicated enforcement mechanisms, the sole purpose of which is to restrict activities that are primarily of private benefit. The public interest cannot ignore this administrative burden required to police private investing practices on the floor.

The Special Study concludes and recommends:

1. Floor trading in its present form is a vestige of the former "private club" character of stock exchanges and should not be permitted to continue on the NYSE or Amex. The Special Study therefore recommends that, except as permitted under any program adopted pursuant to the following paragraph, (a) floor trading on the part of members and member firms of the NYSE and Amex whose income from floor trading in each of the years 1961 and 1962 amounted to less than 25 percent of their total gross income from all activities in the securities business (including floor trading) should be prohibited by a Commission rule under section 11(a) on and after January 2, 1964; and (b) floor trading on the part of members and member firms whose income from floor trading in either of the years 1961 or 1962 exceeded 25 percent of their total gross income from all activities in the securities business (including floor trading) should be prohibited by such rule on and after January 2, 1965. There should be excepted from these prohibitions, however, (i) transactions by specialists or odd lot dealers in stocks in which they are registered, if reasonably necessary in terms of the functions served by such members; and (ii) transactions effected to offset transactions made in error.

2. It has been noted, in part D, above, that the financial capacity of some specialists or of the specialist system generally is in need of strengthening, and it is possible that some present floor traders could perform a highly useful function as "auxiliary specialists."

The NYSE and Amex should undertake studies, in conjunction with the Commission, as to the feasibility and desirability of a program under which present floor traders or other members of such exchanges might register with the exchanges as "auxiliary specialists," with permission to trade on the floor in any security on condition that (a) the auxiliary specialist meets special capital requirements equivalent to those applicable to a specialist registered in (say) 10 average-priced stocks; (b) all transactions of such auxiliary specialist on the floor are either undertaken at the unsolicited request of a specialist and, in accordance with rules similar to those governing specialists, or are effected for the purpose of reversing in whole or in part a transaction so undertaken. If such studies indicate the feasibility and desirability of such a program, it should be put into effect promptly with appropriate procedures for surveillance by the respective exchanges.

3. Since floor trading on regional exchanges in dually listed stocks does not appear to influence price movements or involve special advantages, a different approach or approaches to floor trading on regional exchanges may be warranted and should be the subject of separate consideration by the Commission. Among other things, consideration should be given to whether floor trading in solely listed stocks on regional exchanges is or is not comparable to floor trading on the NYSE and Amex.

G. MEMBERS' OFF-FLOOR TRADING

1. INTRODUCTION AND METHOD OF STUDY

Approximately 5 percent of all shares purchased and sold in the round-lot market of the NYSE is attributable to members' trading for their own accounts in transactions which originate away from or "off" the floor. During the congressional hearings preceding enactment of the Securities Exchange Act of 1934 this type of member trading received considerable attention, due in large measure to member participation in pool operations.499 Under section 11 of the act, the Commission is empowered "to prevent such excessive trading on the exchange but off the floor by members *** as the Commission may deem detrimental to the maintenance of a fair and orderly market," but it has never exercised this authority. Rule 435 of the NYSE, adopted at the Commission's request in 1935 and since expanded by the exchange, sets forth certain general prohibitions on member trading, the first of which provides that no member shall effect transactions on the exchange which "*** are excessive in view of his *** financial resources or in view of the market for such security." This rule has not played a significant role in the exchange's disciplinary actions, however, and generally speaking member trading from off the floor has excited little Commission or NYSE interest since 1935. Although members must file weekly forms with the exchange setting forth the aggregate number of shares traded from off the floor for their own accounts each day (total pur

“***

499 See Senate Committee on Banking and Currency, "Stock Exchange Practices." S. Rept. 1455, 73d Cong., 2d sess., pp. 30-45 (1934). See also ch. VI.C for a brief description of the legislative history regarding member trading.

chases, total sales, total short sales), no identification of the stocks, number of transactions, or time of transactions is required. Less is known, as a result, about the nature of member trading from off the floor than is known about the trading of specialists or floor traders.

In order to study the trading of members from off the floor, the Special Study requested all NYSE members to set forth on questionnaire forms all such round-lot trading by stock and by day for the 3 weeks ended January 27, March 24, and June 16, 1961.500 These data were then analyzed on a "stock day" basis,501 facilitated by the computer process described in appendix VI-A.

2. VOLUME OF MEMBER TRADING FROM OFF THE FLOOR

Member trading from off the floor of the NYSE between the years 1937 and 1961 accounted for 2.9 to 6.1 percent of total round-lot purchases and sales each year. Prior to 1955 this trading exceeded 4.5 percent only once (6.1 percent in 1949), but it has since ranged from 4.8 to 5.3 percent (table VI-73). This trading declined from approximately 40 million shares in 1937 to less than 8 million in 1942, but has since shown a relatively constant increase to over 102 million shares in 1961.

Over the three 1-week periods studied, NYSE members off floor accounted for 5 percent of total round-lot purchases and sales on the NYSE, a slightly higher rate of participation than their percent participation for the entire year of 1961. They traded 6,803,800 shares over the 3 weeks in 6,728 stock days, or an average of 1,011 shares per stock day in which they participated (table VI-a). In about half of the stock days in which they participated their purchases and sales equaled 9.75 percent or more of the total stock-day volume, or approximately 4.87 percent or more of total stock day purchases and sales (table VI-2).502

Unlike specialists and associate brokers, whose trading is limited primarily to their marketmaking functions, and unlike floor traders, whose trading consists largely of short-term investment or speculation, members seem to trade from off the floor for a variety of purposes, ranging from personal investment to arbitrage, and including the offsetting of positions by dual members who are specialists on the regional exchanges. No data are available, however, indicating the extent to which member trading from off the floor is attributable to each of these various trading purposes.

Although virtually every exchange member participates in this type of trading (if only, in many cases, for personal investment) the bulk of such trading tends to be highly concentrated in relatively few member Över the 3 weeks studied, the 30 member accounts trading most actively from off the floor each week accounted for 51.9 percent of all shares traded by members from off the floor (tables VI-74 to VI-76). A total of 49 accounts were among the 30 most active, with 16 accounts appearing among the most active in all 3 weeks, and 9 accounts appearing in 2 of the 3 weeks.

500 These data were reported on form EX-2, a copy of which appears in app. VI-J. Form EX-3, also in app. VI-J, obtained data for each transaction by members on Mar. 23, 1961, but the Special Study has not been able fully to analyze these data in time for inclusion in the report.

501 The "stock-day" concept is explained in note 63, p. 51, in pt. C. above, and in app. VI-A.

502 For an explanation of these two different measures of trading, see note 1, table VI-b. of pt. C.

3. CONCENTRATION OF MEMBER TRADING FROM OFF THE FLOOR Members traded from off the floor in less than half of the total stock days over the 3 weeks. As measured by the number of stock days in which they participated, members off floor exhibited a slight preference for higher priced stocks; approximately 55 percent of the stock days in which they participated were above the median price of all stock days over the 3 weeks (chart VI-1). Within the stock days in which they did trade, however, members off floor revealed a slight tendency to account for a greater percent of total stock-day volume in low-priced stocks relative to high-priced stocks (app. VI-A, chart 13 and table 13). Overall, therefore, the conclusion appears to be warranted that the trading of members off floor does not concentrate to a significant extent in either high-priced or low-priced stocks.

Members revealed a more pronounced tendency to trade from off the floor in those stock days with wider daily price ranges; approximately 61 percent of the stock days in which members traded from off the floor were those with daily price ranges wider than the median price range over the 3 weeks (chart VI-2). Again, however, within the stock days in which they did trade, members off floor tended to account for a greater percent of total stock day volume in stock days of narrow price range relative to stock days of wide price range (app. VI-A, chart 14 and table 14).

503

Most pronounced is the tendency for members to trade from off the floor in stocks experiencing high volume on any given day. Approximately 77 percent of the stock days on which members traded from off the floor were stock days of higher share volume than the median stock day (approximately 1,800 shares) for the 3 weeks (chart VI-3).5 Again, however, within the stock days on which they did trade members off the floor tended to account for a greater percent of total volume on the stock days of lower volume (app. VI-A, chart 15 and table 15). That these members nonetheless tend to trade a disproportionate number of shares in the more active stocks is indicated by the concentration of their trading in the 25 most active stocks in each of the 3 weeks studied. Whereas the 25 most active stocks accounted for 19.6, 22.1, and 19.7 percent (tables VI-58 to VI-60) of total NYSE volume in these weeks respectively, members trading from off the floor did 27.9, 36.1, and 29.7 percent of their trading in these stocks (tables VI-77 to VI-79), and accounted for 7.3, 8.7, and 7.8 percent of the total purchases and sales in these stocks.504

In several cases members trading from off the floor accounted for substantial proportions of the weekly purchases or sales in individual stocks among the 25 most active. For the week ended January 27, 1961, for instance, they accounted for 12.5 percent of all shares purchased in Loew's Theatres, Inc., 17.7 percent of all shares purchased in American Telephone & Telegraph Co., 29.9 percent of all shares sold in Alleghany Corp., 47.6 percent of all shares sold in Royal Dutch Petroleum Co., and 59.9 percent of all purchases in Chance-Vought Corp. Member trading in American Telephone & Telegraph Co.

503 See also chart VI-4. This chart reflects stock day activity by number of transactions rather than by share volume, but produces results almost identical to chart VI-3.

604 This concentration is corroborated by the fact that members trading from off the floor had high partiicpation rates in a significant number of very active stock days over the 3 weeks (app. VI-A, chart 15 and table 15).

was heavy in each of the other weeks studied, accounting for 49.2 and 23.5 percent of all sales in the March and June wecks, respectively. It appears that many of the days or weeks of concentrated trading by members from off the floor are attributable to arbitrage operations.505 Not all cases are explainable in this fashion, however. On Monday, January 23, for instance, Chance-Vought Corp. opened late at 43%, up 34 from the prior close. Members trading from off the floor purchased 23,600 of the 29,200 shares traded over the day. In this case, it seems that members were utilizing their commission-rate advantages to benefit from an offer by Ling-Temco Electronics to purchase all offered shares of Chance-Vought at $43.50 a share.506

4. RELATIONSHIP OF MEMBER OFF-FLOOR TRADING TO PRICE MOVEMENTS

These cases of concentrated member buying or selling raise a question as to the effect of such trading on stock prices. To the extent that arbitrage is involved, this trading keeps NYSE prices in line with other domestic or foreign markets, or with rights, warrants, convertible securities, et cetera. Some trading may also reflect member responses to specialist requests for bids or offers on large blocks of stock. In any event, a question remains as to the effect of such trading on price movements.

The few studies that have been made in this area have done little more than reveal a unique feature of trading by NYSE members from off the floor—that they usually sell more shares than they buy on the exchange. Over the years 1937 through 1961, members trading from off the floor sold over 90 million shares more than they purchased, for an average yearly sale balance of more than 3.5 million shares (table VI-80). In the period November 2, 1959, to November 1, 1961, members trading from off the floor had net sale balances on 330 days, and purchase balances on only 174. On the 6,728 stock days in which members traded from off the floor during the 3 weeks studied, they posted even balances in 512, purchase balances in 2,928, and sale balances in 3,288, and had a net sale balance for the 3 weeks of 458,000 shares. Because of this tendency to post rather consistent sale balances, it is difficult to show a correlation between member off-floor purchase and sale balances and price movements, as has been done with floor traders.507 There is some evidence, however, that suggests that the sale balances of members trading from off the floor may occur relatively more frequently on days of price decline than on days of price rise.508

In a majority of the stock days in which members traded from off the floor during the 3 weeks studied their net balances were relatively small. In 7.6 percent of the stock days they posted even balances, and their balances were between minus 100 shares and plus 100 shares (including zero balances) in 40 percent of the stock days. In only 16 percent of the stock days in which they traded did they post net pur

505 The Royal Dutch Petroleum Co. instance noted in the text is probably traceable to foreign arbitrage, and the Alleghany Corp. situation was due to arbitrage between the common and preferred stock. The March trading in American Telephone & Telegraph Co. apparently involved trading in rights. The nature of arbitrage transactions is explained in pt. H.1.c(3).

de Wall Street Journal, Jan. 24, 1961, p. 23. Nonmembers, due to higher commission rates, could not purchase the stock at 43% or 434 (the range for the day) and sell it at a profit to Ling-Temco Electronics.

B07 See pt. F, above.

608 See Securities and Exchange Commission, "Report on the Feasibility and Advisability of the Complete Segregation of the Function of Dealer and Broker," p. 48 (1936).

96-746-63-pt. 2—17