cotton prices, however, were not associated with corresponding proportional decreases in average changes in prices. As a result, the average changes in cotton prices over specified periods represented a larger proportion of the seasonal average price in 1931-32 when prices were relatively low than in any other year included in the study. The proportions of the yearly average prices represented by the average differences between the closing prices of Middling -inch spot cotton in New Orleans on Fridays separated by 4-week periods varied from 4 percent in 1934-35 to 12 percent in 1931-32; and those separated by 8-week periods varied from 5 percent in 1934-35 to 15 percent in 1931-32 and 1932-33.

RELATION OF CHANGES IN PRICES OF FUTURES CONTRACTS TO CHANGES IN PRICES OF

SPOT COTTON

The relationship between prices of New York futures contracts for the various active months and prices of Middling %-inch spot cotton in New Orleans is indicated by the data shown in figure 3. An examination of these data shows considerable irregularities in the movement of the spot-future price relationships. It will be noted that the futures prices in August 1926 were considerably below prices of spot cotton in New Orleans, but that by the end of September of the same year they were considerably above spot prices. During the season 1927-28, prices of future contracts declined from considerably above prices of spot cotton in August to approximately equal to and below spot cotton prices from January throughout the remainder of the season. Considerable irregularities are also noted during the season 1928-29 and the "squeeze” of May and July contracts in 1929-30 resulted in the irregularities shown from March through June during that season. From 1930-31 to 1933-34, a fairly uniform relationship was maintained between prices of spot cotton in New Orleans and of New York and New Orleans futures contracts. During the seasons 1933-34 and 1934-35, however, prices of spot cotton advanced in relation to prices of futures contracts so that by the end of September 1934 prices of Middling %-inch spot cotton in New Orleans were considerably above prices of New York futures contracts for all active months and remained relatively high throughout the remainder of the season.

EXTENT OF PROTECTION FROM FLUCTUATIONS IN PRICES OF SPOT COTTON AFFORDED BY HEDGES

Hedging, or the attempt to shift risks from changes in prices of spot cotton by counterbalancing purchases or sales of futures contracts, is based upon the assumption that prices of spot cotton and of futures contracts move parallel, after due allowance has been made for the cost of carrying the spot cotton. As previously shown (fig. 2, p. 8), this assumption is substantially correct for the broader price movements. Consequently, the cotton merchant or manufacturer who acquires a supply of cotton and as a result finds himself exposed to the hazards of a decline in cotton prices can hedge his purchase by selling an equal quantity of cotton in the futures market. Then, if prices of spot cotton and of futures contracts advance or decline by the same amount, any gains or losses on his spot cotton due to changes in prices will be offset by equal losses or gains on his futures contracts. To the extent that prices of spot cotton and of futures contracts do not move in the same directions and by the same amounts, the hedge either fails to give complete protection from changes in prices or yields some additional profits.

The failure of prices of spot cotton and of futures contracts to move parallel results in changes in the spot-futures-price relationship, generally referred to as

11 The term "squeeze" is used to describe a situation in the market in which more cotton is called for, or expected to be called for, in settlement of maturing futures contracts than is readily available at the point or points of delivery, with the result that prices of contracts in the month or months maturing or about to mature are elevated over prices of contracts for more distant months and generally also considerably over average prices of spot cotton throughout the country. For example, as a result of the squeeze of New York contracts maturing in May and July 1930, prices of these contracts were elevated from considerably below prices of October and December contracts in February to more than 170 points above prices of October and December contracts in May. In the same period, the elevation of the prices of May and July contracts extended from about 42 and 46 points, respectively, above the average of prices of Middling %-inch spot cotton in the 10 designated markets on Mar. 21 to about 114 and 122 points, respectively, above the sam average of spot prices on May 16, after which the "squeeze" was relaxed and the price of July contracts declined to about 63 points above the spot average on July 18.

changes in basis. These changes in basis are accounted for by (1) those items of expense in connection with carrying spot cotton, such as storage, insurance, and interest, which tend to result in basis gains, and (2) those unforeseen developments which result in unexpected basis gains or losses. In this study the gains from hedging the price of spot cotton by selling futures contracts are the same as the losses from hedging spot cotton requirements by the buying of futures contracts. In making analyses of data to show the security of hedges, the changes in basis which normally result from costs of carrying spot cotton have been estimated and largely eliminated from the data shown. The unforeseen changes in basis for the period 1926-27 to 1934-35, inclusive, are presented to show the risks not shifted or eliminated by means of hedging.

An examination of the data presented in table 2 (p. 11) shows considerable changes in the basis. The extreme changes in basis, or the gains or losses from hedged cotton, during the period 1926-27 to 1934-35, inclusive, amounted to as much as 1.45 cents a pound over 4-week periods, and 2.19 cents a pound over 8-weeks periods. It will be noted that the extreme changes in basis during the 9-year period studied varied considerably from year to year and that prior to 1931-32 these changes were much greater than since that date." The proportion of the seasonal average price represented by these changes in basis amounted to as much as 15 percent over 4-week periods, and 18 percent over 8-week periods. These proportions varied irregularly from year to year, but were greatest in 1930-31 as a result of the squeeze of the May and July contracts in 1930, and they were relatively small in 1927-28 when prices were relatively high and again since 1930-31 when delivery at southern warehouses on New York futures contracts with no differentials was permitted.

Average gains or losses from changes in basis during the period 1926-27 to 1934-35, inclusive, are also shown in table 2 (p. 11). For the 9 years combined, the average change in basis amounted to 0.17 cent a pound over 4-week periods and to 0.25 cent a pound over 8-week periods. Considerable variations in average change in basis from year to year were noted. These changes in basis varied somewhat irregularly with the level of cotton prices. The relatively small average changes in basis since 1930-31 were associated with relatively low levels of cotton prices and with southern delivery in its present form on New York futures contracts.

The average changes in basis, when expressed as proportions of the seasonal average farm price, varied from 0.7 percent in 1933-34 to 1 percent in 1926-27 for 4-week periods; and from 1.2 percent in 1934-35 to 3 percent in 1926-27 for 8-week periods. The significance of these gains or losses in basis can be appreciated in the light of testimony that the margin of profits generally expected in connection with the marketing of the cotton crop amounts to 1 percent or less of the price of the medium- and shorter-staple cotton."

A comparison of the changes in basis with the changes in spot prices shows that generally basis risks were materially less than the risks from price fluctuation (table 2, p. 11). During the period 1926-27 to 1934-35, inclusive, the average change in basis amounted to about 22 percent of the average change in spot prices. That this proportion was fairly constant for periods of dif ferent lengths is indicated by the fact that the range in these average proportions for the period 1926-27 to 1932-33 varied from about 23 percent for 2-week and 12-week periods to about 25 percent for 24-week periods (fig. 4). While these comparisons of changes in prices of spot cotton with changes in basis for periods of various lengths relate to prices of Middling %-inch spot cotton in New Orleans and prices of New Orleans futures contracts, figure 5 shows that the relationship between average changes in prices of Middling -inch spot cotton in New Orleans and prices of New Orleans futures contracts was practically the same as that between average changes in prices of Middling -inch spot cotton in New Orleans and prices of New York futures contracts.

12 The bylaws of the New York Cotton Exchange were amended in November 1928 to provide for the delivery of cotton on the New York contract at specified southern points. The price for cotton delivered at southern points was to be invoiced at 35 points less than the contract price. Trading began on the new contract in January 1929, and the first delivery month under this contract was in October 1929. The bylaws were further amended by eliminating the 35-point differential in February 1930, and the first delivery month under this contract was October 1930.

18 Garside, Alston H., Specimens of Cotton Hedging, mimeographed report by the New York Cotton Exchange, 1933, p. 1, and Meadows, William H., Commodity Short Selling. hearing before the Committee on Agriculture, House of Representatives, 72d Cong., 1st sess., January 1932, serial B, pt. II, p. 316.

The proportion of the changes in adjusted price of Middling %-inch spot cotton in New Orleans that would have been offset by counterbalancing sales or purchases of New York futures contracts varied considerably (table 2, p. 11). For the 9-year period, the hedge would have eliminated on the average about 78 percent of the gains or losses to individuals from changes in prices of spot cotton corrected for carrying charges. The protection afforded by hedges varied considerably from year to year. For example, the proportion of the changes in spot prices, over 8-week periods, that would have been offset by purchases or sales of futures contracts maturing during the season of 1928-29 amounted, on the average, to about 58 percent; whereas the hedge protection by purchases or sales of futures contracts maturing in 1932-33 amounted, on the average, to about 92 percent.

The average changes in spot prices and in basis, as previously shown, include both gains and losses to holders of spot cotton. During the period 1926-27 to 1934–35, inclusive, changes in adjusted prices of Middling %-inch spot cotton TABLE 3.-Percentage of times differences between closing prices of Middling 7-inch spot cotton in New Orleans, adjusted for carrying charges, and between adjusted "basis", on Fridays separated by 8-week periods, resulted in gains and in losses to holders of spot cotton; and average amounts of these gains and losses, by years, 1926-27 to 1934-35

1 Adjustments were made by subtracting from the changes in the prices of spot cotton the cost of storage' insurance and interest for carrying spot cotton for 8-week periods.

2 Adjusted "basis" is the spread between the adjusted prices of Middling -inch spot cotton in New Orleans and the prices of New York futures contracts for the nearest active month not in process of maturing.

in New Orleans and in adjusted basis over 8-week periods resulted in losses to the holders of spot cotton more often than in gains, and the average losses were considerably greater than the average gains (table 3 and fig. 6). The proportion of the time losses were sustained from changes in adjusted prices of spot cotton was considerably less than the proportion of the time losses were sustained from changes in basis, but the losses from changes in spot prices were, on the average, almost 5 times as great as the losses from changes in basis. Average gains and losses from changes in prices of spot cotton and also from changes in basis varied considerably from year to year. Losses from changes in prices of spot cotton and in basis were considerably less for the years 1931-32 to 1934-35 than for the preceding 5 years. This marked reduction in losses from changes in basis was associated with the change in New York futures contracts permitting delivery at southern points.

The proportion of the time changes in adjusted prices of Middling %-inch spot cotton in New Orleans and in adjusted basis over 8-week periods were of specified amounts from 1926–27 to 1934–35, inclusive, is shown in table 4. An examination of the data presented in this table shows that for the 9 years combined the gains or losses from changes in prices of Middling %-inch spot cotton adjusted for carrying charges over 8-week periods were less than 20 points, or $1 a bale of 500 pounds, less than 13 percent of the time; whereas

gains or losses from changes in adjusted basis were less than $1 a bale about 65 percent of the time. Gains or losses from these changes in prices of spot cotton adjusted for carrying charges over 8-week periods were as much as $5 a bale more than half of the time and $10 a bale more than 16 percent of the time; whereas gains or losses from changes in basis were as much as $5 a bale for only about 2 percent of the time. The data reveal considerable similarity in distribution of changes for the different contract months, but, on the whole, the changes in basis were generally somewhat smaller for March, May, and July contracts than for contracts maturing in other months.

TABLE 4.-Proportion of the differences between closing prices1 of Middling -inch spot cotton in New Orleans, adjusted for carrying charges, and between adjusted basis, on Fridays separated by 8-week periods, were of specified amounts, 1926–27 to 1934-35, inclusive

Percent Percent Percent Percent Percent Percent Percent Percent Percent Percent

1 Adjustments were made by subtracting from the changes in the prices of spot cotton the cost of storage, insurance, and interest for carrying spot cotton for 8-week periods.

2 Adjusted "basis" is the spread between the adjusted prices of Middling -inch spot cotton in New Orleans and the prices of New York futures contracts, for the nearest active month not in process of maturing.

Average changes in prices of spot cotton in New Orleans and in basis were somewhat irregular from time to time, but, on the whole, were greater in the months from June to October than during any other time of the year (table 5). This period is one when prospects for the coming crop change materially and these changes are reflected in prices. The changes were, in general, relatively small during the middle part of the season after the size of the crop becomes fairly definitely known and before speculations relative to the size of the next year's crop began to affect prices materially. The proportions of the changes in spot-cotton prices that could have been offset by contrapurchases or sales of futures contracts varied irregularly from time to time, but were greatest from June to October when price fluctuations were relatively large and were least from February to April when price fluctuations were relatively narrow (table 5).

TABLE 5.-Average differences in the closing prices1 of Middling %-inch spot cotton in New Orleans and the basis2 on Fridays, separated by 8-week periods, for specified months, 1926-27 to 1932-33

1 Adjustments were made by subtracting from the changes in the prices of spot cotton, the cost of storage, insurance, and interest for carrying spot cotton 8 weeks.

Adjusted "basis" is the spread between the adjusted prices of spot cotton in New Orleans and the prices of New York and New Orleans futures contracts.

Prices of spot cotton, after being adjusted for carrying charges, moved more nearly parallel to futures prices for the near-active months than to those for the more-distant months. An examination of the data in table 6 shows a fairly consistent increase in average change in basis along with the increase in number of months prior to the maturity of the futures contracts. In other words, the hedge protection afforded by near-month contracts was generally considerably greater than that afforded by contracts for more-distant months. For example, the average changes in basis decreased in proportion to changes in spot-cotton prices from almost 30 percent 8 months prior to maturity of the contract to about 16 or 18 percent 1 month prior to maturity.

TABLE 6.—Average differences in closing prices1 of Middling %-inch spot cotton in New Orleans and in "basis" 2 on Fridays separated by 8-week periods ending during specified number of months prior to maturity of futures contracts, 1926-27 to 1932-33

1 Adjustments were made by subtracting from the changes in the prices of spot cotton the cost of storage, insurance, and interest for carrying spot cotton 8 weeks.

Adjusted "basis" is the spread between the adjusted prices of spot cotton in New Orleans and the prices of New York and New Orleans futures contracts.

55627-pt. 2—36—15