Seasonal Studies and Key Dates

Most agricultural commodities exhibit traditional, reliable price moves at one or more periods throughout the year. The grains grown in the Northem Hemisphere have a high likelihood of a rally during the late spring and early summer, when the chance of drought will have the greatest effect on yield. When prices show the normal harvest lows, followed by a modest rise and sidewajs pattem throughout the winter, the potential is good for a rally during the early growing season. Because prices start at relatively low levels, the risk of holding a long position is small. Once prices have moved higher, there is rarely a season in which a short sale of corn, soybeans, cotton, or sugar will not net a good profit within the two months before harvest begins.

Seasonal studies are intended to provide information on when the largest move will occur. The following studies-(l) Grushcow and Smith (1980), (2) ContiCommodily (1983), and (3) Bemstein (1986)-will be used to compare the seasonality in some of the more active commodities. A summary of the results is shown in Table -ll Each study offers a different perspective on seasonality Grushcow and Smith analyze both cadi and individual futures martlets over a fairly long period and present complete statiatics; ContiCommodily used mostly the past 10 years, bul included a unique volatility analjsis; Bemstein, the most recently published study (Seasonal Conc ts in Futures Trading), gave the most complete background on calculations, including separate studies of bullish and bearish years and an exceptionally long time period for cash maiket analjsis.

The number of years in the seasonal analjsis counts heavily in determining the normal patterns. As Table 7-11 shows, the ContiCommodily results, based on a maximum of 10 years, is often quite different from the other two studies. For trading safety, it would be best to select those patterns that have proved reliable over many years; however, because inconsistency in the past 10 years cannot be ignored, a trader must be able to identifj nonseasonal, bullish, or bearish patterns.

The periods that seem consistent throughout all studies are:

Corn and soybeans. September and October show major harvest pressure. Cattle. An end-of-the-year liquidation and midwinter rally. Coffee and juice. No common moves in the three studies.

Some commodities are clearly more seasonally consistent than others. Both the coffee and orange juice markets were expected to show patterns that reflect a rise as the possibilities of a freeze increase, it must be that the normal seasonalitj of these products is distorted by the inconsistent and highly volatile periods that follow a freeze. These martlets would be candidates for an analogous years study in which you compare only those years with common factors.

The three studies shown here, as well as most others, include recommended frades based on key dates, which reflect the patterns in Table 7-lOa. By selecting those trades common to all of them, you have found those that are most reliable. In a summarj by Bemstein, which catalogues commodities by those monttis with the highest reliability, the agricultural products are clearly the most seasonal. The only nonagricultural martlet that shows any consistent seasonality is copper. Although there may be interesting aiguments for the forces of demand on silver, currencies, and financial maikets, their inconsistency is apparent and they are not candidates for seasonal trading.

Seasonal Calendar

Because Bemsteins work covers the cash martlets over an extremely long period, it must be considered the most reliable source of basic seasonal patterns. Table 7-12 is part of the

FMlUllliiil

jTU-lic-w ano (cliaiige in pnce) prontiCjiiniodity i".. cli.-aieein seasoni

".. refej.- to the reliability ..f montijy sei

ionality

weekly seasonal calendar, which appears in Seasonal Conc ts in Futures Trading. The numbers that appear show eks with consistent historic moves. Weel of 64" and higher represent upward moves; weeks of 36° and ;r display downward trends. This calendar can be exfremely useful vhien combined with some simple hiding logic i "Is the market acting in a seasonal manner?" before the position is entered.

TABLE 7-12 Seasonal Calendar

TABLE 7-12 (Continued)