of their Exchange seats. Total income was compared to this capital figure (table VI-20).119 It should be emphasized that these are gross and not net return figures.120 The anlaysis showed a fairly wide dispersion among units; most had a return ranging between 10 and 50 percent. There was no pronounced relationship between capitalization and rate of return, although some of the more heavily capitalized units fell into the higher ranges.

In summary, NYSE specialist units show great diversity in size, number of specialty stocks, source of income, capital, and rate of return. The larger units tend to have more than their proportionate share of capital, indicating a greater ability to service the market adequately. However, as unit size increases, the major source of income shifts from brokerage income to trading profits, which suggests that trading may become an end in itself. As also suggested by these data and as discussed more fully below, specialists' market participation seems to be keyed to their capital ability rather than to any uniformly interpreted standard of responsibility.

5. THE BASIC BROKERAGE FUNCTION-THE SPECIALIST'S "BOOK"

As noted above, the continuous auction market and its physical organization into different posts require that there be some mechanism to store and execute limited price orders.122 The specialist brokerage function has evolved to meet this need,123 although specialists also accept market orders.

Specialists are compensated for executing agency orders by the firms which transmit such orders to them. The fact that the forwarding firm pays a portion of its commission to the specialist does not affect the total commission paid by the customer who originated the order. The specialist's portion of the customer's commission is known as "floor brokerage" and there is a prescribed minimum amount set by the NYSE constitution which is the same as that paid to independent brokers when they execute orders on the floor.124 On the Amex, contrary to the NYSE practice, when an order is transmitted to a specialist, floor brokerage may be shared by the specialist with the forwarding broker. 125 A substantial percentage of all round-lot orders on the Exchange are executed by specialists in an agency capacity. As indicated previously, aggregate specialist brokerage income for 1959 and 1960 was $19,590,000 and $18,919,000 respectively. Since

119 It might be argued that the proper rate of return should be restricted only to trading profits, since this is the direct employment of the capital. However, an integral part of the specialists' activities consists of executing brokerage orders and the two functions are always viewed as interdependent.

120 Just as there are no uniform concepts used to identify specialist capital, there are likewise no uniform methods by which costs may be properly allocated, and therefore it was not possible to arrive at a net return. The representatives of the NYSE have recognized that for the determination of profitability, it would be necessary to standardize accounting procedures.

121 See sec. 6.d, below.

122 Such a mechanism must also include a system to determine priority between such orders. Under Exchange practices, orders entrusted to specialists are generally handled on a first-in-first-out basis.

123 Legend has it that the first specialist was a floor broker whose activities were confined to a particular section of the floor by a broken leg. Apocryphal or not, it is fairly clear that the first specialists were differentiated from their fellow floor brokers only because they "specialized" in specific stocks, accepting brokerage orders from other

members.

124 NYSE constitution, art. XV.

125 Amex constitution, art. VI. There is one exception to the NYSE practice that specialists must receive floor brokerage for orders they execute, which relates to orders executed at openings. See sec. 6.j, below. See pt. I of this chapter for further discussion of the minimum commission schedules.

average floor brokerage is about $3.85 per round lot, specialists executed as agents about 5,088,300 round lots in 1959 and 4,914,000 in 1960 out of a maximum of 17,164,020 possible round-lot orders in 1959 and 15,866,700 in 1960.

This section contains an examination of the kinds of orders accepted by the specialist in his brokerage capacity, and of the specialist's book and its uses.

126

a. Orders accepted by specialists

Section 11 127 of the Exchange Act prohibits, without exception,128 a specialist's effecting "any transaction except upon a market or limited price order." 129

The statute does not define limited or market orders. In common usage, a limited order is one that limits the discretion of the broker who holds it he may execute it only at the price specified or better. For example, an order to buy a particular security with a limit of 50 may be executed only at that price or lower. If the market moves below 50, it is the broker's responsibility to execute it at a lower price if possible.

When the price moves to a point that the public bids held by the specialist at 50 are capable of execution, the order first received by the specialist at that price will normally be executed first. Specialists do not guarantee that all orders on their book at a particular price will be executed merely because the senior order at that price has been executed. Each order must await its turn for execution under technical auction market rules.180 If another customer who had entered an order at the same price sees an execution on the tape at 50 (when the senior order is executed) and inquires as to the status of his order, he will be given a report "stock ahead," and on further inquiry he can usually determine the number of shares ahead of his order.

Although the statute refers to limited price and market orders, specialists also accept "stop" orders, a variant of both limited price and market orders. These are orders to sell securities when the current price declines to a specified point 131 or to buy when the price increases to a specified point.182 When the price reaches the specified point, the stop order becomes a market order and is handled as such.133

126 The various problems which arise from the fact that specialists act as agent for many customers on both sides of the market, and also act as dealer, are discussed in sec. 7, below.

127 See sec. 3.b, above.

128 Under sec. 11 (c), upon a finding of limited volume on the particular exchange, the Commission may grant appropriate exceptions to all provisions of section 11 (a) and (b). 129 The section also restricts disclosure of information regarding orders on the specialist's book which is not generally available to other members. The Commission is given authority to require disclosure to all members of all orders held by specialists, but this authority has not been exercised.

130 See pt. B of this chapter.

181 These orders are commonly known as "stop-loss" or "sell-stop" orders.

183 These orders are known as "buy-stop" orders. While the term stop-loss orders implies the reason for their use by investors, the function of buy-stop orders may not be as apparent. One principle reason for the use of such orders is similar to that for the use of stop-loss orders, in that an investor with a short position may desire to protect his profits or cut his losses when prices increase to a certain level. The other principle reason is that investors who make their decisions on a so-called "technical" basis may believe that a stock is not a "buy" until the price increases to a particular point.

133 There is a variant of stop-orders known as "stop-limit" orders. Such orders contain a second specified price away from the market beyond which they may not be executed. In other words, a stop-loss order with an effective price of 50 becomes a market order at that price and may be executed at the best available price-not necessarily at 50. A limit such as 48 may be placed on the stop-loss order, preventing an execution at a lower price than 48. NYSE specialists ordinarily accept both stop-loss and stop-limit orders. Amex specialists are permitted by that Exchange to accept only stop-limit orders.

A market order is a direction by a customer to execute "at the market." Beyond the fact that a broker has a reasonable time to transmit the order to the floor of the Exchange and then to the trading post, opinions seem to differ as to the broker's actual responsibility. The following simplified example may serve as an illustration: a broker receiving a market order to sell 100 shares proceeds promptly to the post at which the stock is traded, where he finds the stock quoted at 49 bid, offered at 50. Upon inquiring as to the size of the market, he is told that there are bids for 1,000 shares at 49. Upon further inquiry, he ascertains that the last sale was at 493⁄4 and that the price has been increasing during the day. The broker may decide to "hit" the bid and sell at 49 or he may decide to underoffer the best existing offer by offering the 100 shares at 49%, in which case the market becomes 49 bid, offered at 49%. He might do this in the exercise of his brokerage judgment, based on the previous trend of the market, the last sale, and the size (1,000 shares) of the existing bid which indicates that he would probably be able to sell at 49 even after several small transactions at that price. If a buyer takes his offer the broker has achieved an execution at 497%-a better price than the 49 bid. However, if the market moves the wrong way before he can act, e.g., a seller with a large order sells 1,000 shares at 49, forcing the price down, the broker may not be able to execute his order at 49 but only at some lower price, e.g., 4812.

The question arises whether in such a situation the broker is liable to his customer for the 1/2 point between 482 and 49 (the price which he could have obtained when he arrived at the post), even though, by hypothesis, he exercised reasonable care and reasonable judgment. Most specialists take the view that the broker would be liable, since a market order "holds the broker to the tape," i.e., the first possible sale in which he could have participated under the technical rules of the auction market. A few seem to believe that the broker would not be liable under such circumstances. The rules of the Exchange are silent on the point, an Exchange publication stating merely that a market order directs a broker "to buy or sell at the best price available" when the order arrives on the trading floor.134 The Restatement of the Law of Agency says that a "*** direction to a stock broker to buy or sell at the market' is interpreted, under normal circumstances, as a direction to buy or sell immediately irrespective of price and prospects." 135 In 1934, alternate definitions of market orders were proposed by the Commission staff, but none of them were adopted.

As a practical matter, since most brokers execute routine market orders immediately and adjust the price if they use brokerage judgment but "miss the market," the definition of orders is basically of importance only in determining the obligation of the specialist. It is important there because the specialist holds orders in particular stocks for many customers on both sides of the market, and the degree of brokerage judgment which he is permitted to exercise may bring his

134 NYSE, "The Language of Investing," 18 (April 1960).

185 Restatement of the Law of Agency (Second) sec. 424, comment B (1958). See also the Segregation Report, at p. 184, which defines market orders as those to be executed *** at the most advantageous price as promptly as reasonably practicable,"

obligations to his several customers into conflict.136 The question becomes even more important because of the development over the last several years of the "not-held" order, which NYSE specialists have been accepting for execution and which allows a degree of discretion to the executing broker.137 However, even without this special problem, more precise definitions of orders seem necessary to effectuate fully the policies underlying the restriction of section 11.

b. Contents of the book

The "book" refers to both the accumulation of market and limit orders entrusted to the specialist and to the actual book in which limit orders are entered while awaiting execution. Market orders are not physically entered in the book; the order slip is held by the specialist. Almost all NYSE specialists enter limit orders in a looseleaf binder approximately 4 by 11 inches, with buy orders on the left side and sell orders on the right.138 Each page is used to record orders at each of a point from one even dollar level to the next, e.g., $35, 35% ... 35. Limit orders and stop orders are entered in the sequence in which they are received, at the appropriate price and with a notation of the number of shares and the name of the firm which forwarded the order; the NYSE practice is not to make a notation of the time the order is received at the post.139 As an order expires, is executed or canceled, it is crossed off. In an active stock the book may be thick, with several pages devoted to orders at each price, while in an inactive stock the book may be devoid of all bids and offers.

Books will tend to become "heavier" or "lighter," depending on volume in the stock. But it should be noted that the "book" changes rapidly because "day" orders, which predominate in some stocks, are canceled at the end of the day of entry, while orders "good until canceled" remain on the books until executed or withdrawn.140 This means that it is extremely difficult actually to reconstruct the book as it was at any particular point in time. To do so requires working with a mass of orders, executions, and cancellations, and even then such efforts may not be wholly successful. Thus, what should be a primary source of data in the surveillance of specialists' activities cannot be effectively utilized.111

The Special Study examined a sample of the specialist books for 103 stocks as of the close of business on February 16 and the opening on February 19, 1962.142 The sample was roughly divided among stocks that were active, inactive, and of average activity. Specialists testified, and examination of the sample books showed, that the thickness of a book was a function of the activity and price of the stock: 143 An active stock by definition attracts orders, some of which are placed with the specialist. Lower priced stocks have a tendency to attract more round-lot orders than higher priced issues.

136 One specialist testified that a very large market order implied a direction to take a reasonable amount of time in the execution of the order. This is a somewhat unusual but not untenable construction of brokerage responsibility in executing such orders. If a customer gives an order to sell 10,000 shares at the market, it does not seem reasonable that the customer is giving a direction that the order be mechanically executed whatever the price consequences. However, this does not solve the question of whether a specialist who accepts such an order compromises his fiduciary obligations to his other customers. 137 See sec. 7b.

138 A sample book is set forth in app. VI-D.

139 On the Amex, orders received by specialists are stamped by a timeclock upon receipt. 140 These orders must be renewed every 6 months.

141 See secs. 6.e and 7.f, below.

142 See app. VI-C for a list of the 103 stocks.

143 Also, a stock that has an active arbitrage will often have a thick book.

The study of the books in 103 sample stocks disclosed that, among the 79 stocks selling at $20 or over, active stocks had thicker books than inactive ones. It was also found that, for both active and inactive stocks in this category, the amount of stock on the books within 10 percent of the market price was a fairly constant proportion of the whole book (table VI-21). In total, the most active group in the sample had an average of 45,600 shares on the book within 10 percent of the previous closing price, while the most inactive group averaged about 3,600 shares. With respect to individual issues, General Electric, American Telephone & Telegraph, and Bethlehem Steel, which were in the most active group and traded approximately 20,000 shares each on the study day, had 53,800, 44,300, and 64,000 shares, respectively, within 10 percent of the closing price. On the other hand, Empire District Electric, Amsted Industries, and Gerber Products, each of which traded 400 shares on the study day, showed only 1,400, 4,400, and 4,500 shares, respectively, within 10 percent of the closing price.

There was some variation even within the active group. One such issue, Westinghouse Electric, had 52,500 shares on the buy side and 11,000 shares on the sell side within 10 percent, while Burroughs, with a similar volume, had 8,100 shares and 22,300 shares respectively. The "book market" (the spread between the best public buy and sell orders held by the specialist) tended to be the closest for the most active stocks (table VI-22). For active stocks, the most common spread was 14, with a range from 1 in General Telephone & Electronics to 34 in Texas Gulf Sulphur, each of which traded about 20,000 shares on the study day. In inactive stocks (trading 1,000 shares a day or less) the most frequent spread was 1⁄2 but spreads as large as 3 points appeared.

Stocks selling below $20 per share had much thicker books than stocks selling at $20 and over. On the study day, 24 of the 103 stocks studied were in the former category. These stocks had a total of 2,792,400 shares on their books, which was 564,000 shares greater than the total on the books of the remaining 79 stocks. Five of these stocks had books containing over 250,000 shares, whereas none of the stocks selling at $20 or over had books this thick. These five included two of the lowest priced issues traded on the Exchange: Rhodesian Selection Trust selling at 12 had 727,000 shares on its book and GrahamPaige selling at 234 had 419,200 shares. Two of the most active stocks in the low-priced group, Studebaker-Packard and American Motors, had 401,700 and 331,300 shares on their books, respectively.

A significant point with respect to the book is contained in the testimony of some specialists with long periods of service on the floor of the Exchange. Almost all of these testified that in the 1920's and even in the early 1930's, the books generally contained many more orders than in recent years, and that markets today are generally "thinner." 144 Of those who attempted to explain the change, most gave two reasons. The first is that there are fewer speculators; i.e., less in-and-out trading for small profits,145 and thus less concentration of orders close to the market. The second is that institutional investors have increased in

144 See also sec. 3.a, above, discussing the changes in volume characteristics which have led to thinner books in active stocks.

145 Among the reasons the specialists advanced for the decline in speculative trading are the capital gains tax holding period, higher margin requirements, and higher commissions.