It is further evident from the chart that the ratio of short sales to total sales generally varies inversely to the trend of stock prices. As the market advanced from a level of 290 (DJI) in January 1954, to above 500 in early 1956, short selling as a percentage of total sales generally remained well below the 5-percent level and declined to below 4 percent near the peak of the market. Only during the price decline in the latter part of 1956, did the ratio rise above 5 percent. As prices rose in July 1957, to match the earlier peak, the ratio again dropped to the 4-percent level. Then on the more severe decline of prices during the following 6 months the percentage of short selling increased to more than 8.5 percent of total sales, reaching this high point near the bottom of the price decline. The same general pattern is seen during the following rise and subsequent fall in stock prices. Prices advanced from the 440 level in January 1958, to nearly 700 at the end of 1959, while the short sales ratio dropped from 8.5 percent to about 3.5 percent. During the decline in prices in 1960, the ratio once again rose to above 6 percent, this high coinciding very nearly with the low point of the market in October of that year. The pattern was repeated in the third general bulge in prices during 1961 and on the subsequent decline through June 1962. While prices rose from 580 to nearly 740, the ratio of short selling dropped from 6 percent to about 3 percent, the latter once more virtually coinciding with the peak in prices. Thereafter, the ratio rose as the 1962 decline set in, increasing again to one of the highest points in the 81⁄2-year period-nearly 8 percent of total sales. b. Members' and nonmembers' short sales Though reports by the exchange do not differentiate among the various types of short selling, the weekly reports do provide a breakdown of short selling in the round-lot market by principal classes of members; i.e., specialists, floor traders and members off the floor. Since the total sales figures for each class of members are also known, it is possible to state the proportion of each class's activity which is short selling, and by subtracting members' short sales from total short sales, the volume of nonmembers' short selling in round lots is indicated. It is thus possible to note the differences, if any, in the patterns of short selling by the several classes. It is also possible to note the extent to which the different clases account for portions of total short sales. The following analysis is directed first to a comparison of short sales by all exchange members with those by nonmembers, and then to a comparison of short sales by the several classes of members. Chart VI-e shows total volume of short sales by members and by nonmembers, together with the trend of the market. During the 8year period, 1954-61, when the market averages showed an overall advance from 290 to 740 (DJI), members more than doubled their volume of short selling, increasing it from an average of 1,390,000 shares a month during 1954, to 3,060,000 a month during 1961. Nonmembers, on the other hand, averaged well under a million shares per month during most of the 8-year rise, exceeding their average almost entirely during periods of falling prices, especially during the decline in the last 6 months of 1957 and early 1958, and also during the first 7 months of 1962. The latter period, one of the sharpest drops in prices for many years, witnessed a greatly increased amount of short selling by members and nonmembers alike. This short selling is examined in more detail below; but turning again to its general features it will be noted from chart VI-e that, on an aggregated basis, the volume of short selling by members far exceeds that of nonmembers. This is also indicated in chart VI-f which pictures, in percentage terms, how the total of short selling is divided between the two groups. It is immediately apparent that the monthly volume of members accounts for more than half of all short selling, generally varying from 60 to 80 percent of the total and sometimes reaching as high as 90 percent. Members appear to increase their proportion of all short selling as an advance in stock prices progresses, and it follows that nonmembers are then accounting for less. The reverse is true during declines. Even though in absolute terms short selling by both members and nonmembers increases during sharp price declines, the greater increase is by nonmembers. Actually, though the monthly totals do not reveal it, nonmembers' sales exceed those of the member class on some days during declines. This was true near the end of the decline in 1957 and again at the bottom of the more drastic decline of 1962. Charts VI-g and VI-h have been plotted on a daily basis from April through July 1962, to show in more detail the volume of short sales and the proportions of total short selling by the two groups as compared with the trend of the market during this brief period of rapid market change. During the decline, as indicated, nonmembers increased their proportion of short selling and accounted for more than 50 percent during the near-panic decline in the final days of May. Later in June as the market dropped to a lower point, the nonmembers accounted for as much as 60 percent of total short sales. In the slight sell-off in mid-July, they once more contributed the greater part of short selling. Thus, while the bulk of short selling is generally effected by members, on some days of unusually sharp price declines short selling by nonmembers has predominated. c. Members' short sales That the greater part of short selling on the New York Stock Exchange is done by members has been known from earlier studies by the Commission; and it was partly for this reason that the Commission in 1936 asked that the exchange, in its weekly report of trading, report short sales separately for the three principal classes of members. The series of weekly reports since then have shown that the great bulk of such selling has been by specialists, followed in order of magnitude by members off the floor (firms and office partners), and by members trading for their own account while on the floor (floor traders). Chart VI-i shows the short selling by the three classes of members as percentages of total short sales, on a monthly basis from January 1954 through July 1962. The ratios in the case of specialists vary from 40 to nearly 70 percent of all short sales, for members off the floor from 10 to 25 percent, and for floor traders from 2 to 10 percent. (Obviously, the three classes time their short selling at least somewhat differently in relation to the market.) |