— CRB Yearbook CD — Seasonality of Volume and Open Interest

By Kenneth H. Shaleen

Classical bar chartists in the futures markets have a general rule for analyzing the interaction of price change, volume, and open interest change to assess the health of a price trend:

Volume and open interest should increase as prices move in the direction of the major price trend

According to this rule, the most bullish condition is price moving up on increasing volume and increasing open interest; the longs are in control and the price uptrend is expected to continue. On a daily basis, this rule implies that on price-up days (a higher close) in a healthy bull market, volume should expand and open interest should increase. When price is moving downward, the ideal healthy bear market is also categorized by expanding volume and expanding open interest.

Generations of futures technicians have used this general rule. They realized that the ideal interaction of price, volume, and open interest is subject to considerable interpretation. Many idiosyncrasies to the general rule exist. One is the possibility of a seasonal influence.

The Commodity Research Bureau (CRB) has produced seasonal charts of volume and open interest for nineteen U.S. futures contracts covering the 10-year period from January 1986 thru September 1996. This article will examine the seasonality of these futures contracts and discuss the necessity of a secondary level of analysis when applying the general volume and open interest rule for a healthy price trend.

Methodology

The monthly variation in volume and open interest each calendar year is measured by constructing seasonal indexes. These are percentages representing each month’s relationship to the average for the year, which is taken as a base equal to 100%. Thus, a Soybean seasonal open interest index of 84% for the month of August indicates that, typically, the average total open interest for August is expected to be below the average monthly open interest by 16%. The 105% figure for U.S. Treasury Bond futures for November means that open interest in this month is usually 5% higher than the monthly average. The arithmetic mean of the 12 monthly index values is 100%.

The Commodity Research Bureau has periodically updated the seasonal charts of the agricultural futures since the mid-1950’s. The absolute level of volume and open interest is different from year to year, but the percentage changes do exhibit a regular pattern. There is a distinct difference between the look of seasonal variations of volume and open interest in the agricultural commodity futures versus the financial instrument futures. This is the tendency for the contract specifications of the financial futures to dominate the monthly swings in volume and open interest. The calendar year-end effect is particularly evident.

Application of the General Rule

The change in open interest is the important condition monitored by technical analysts. In its most minute application, the open interest change from one trading day to the next is analyzed in conjunction with the price change that day. If the investigation is expanded to encompass several weeks, a seasonal influence may have to be factored into the analysis.

For example, given a price uptrend on a chart, declining open interest implies that a weak price uptrend is occurring. But what if total open interest normally declines at that time of year? A second step in the analysis is necessary. The following question must be answered: Is the current decline in open interest equal to or greater than what would normally be expected? If so, the same (not bullish) conclusion would be reached. But if the current open interest is only falling slightly in the face of an expected severe seasonal open interest decline, a bullish reading of the current price rally would result.

Volume also exhibits seasonal characteristics but variations in volume are more important in day-to-day comparisons than over extended periods. Thus no seasonal adjustment is necessary. Each day’s volume is categorized as high, average, or low. The result is then analyzed in conjunction with the price change to determine the direction of the major price trend. For example, low volume sell-offs are bullish; low volume rallies are bearish.

The historic seasonal variation charts can aid a technician in determining how often to re-calibrate the definitions of high, average, and low volume. In the unusually low volume month of July in T-Bill futures (75%), looking back as far as the month of May for volume comparisons would not be productive. The seasonal volume index in T-Bills declines from above 100% in May, to 90% in June, and bottoms in the lowest volume month of the year, July.

Overview of the Seasonal Graphs

Year-End Effect

The tendency for December to be a relative low volume and open interest month is general throughout the U.S. futures markets. Since open positions are marked-to-the-market for U.S. taxpayers the tendency is for speculative positions to be reduced by December 31. The main exception to this overview is Heating Oil. The dominant seasonal tendency for the physical product generates high volume in December.

I. Agricultural Products

A. Wheat, Corn, and Soybeans

The Chicago Wheat, Corn, and Soybean futures exhibit strong seasonal tendencies for open interest to increase as the U.S. harvests approach. The move from low to high average monthly open interest occurs in the short time span of two months. The size of the crop becomes more assured and is reflected in the need to place hedges by both the producer and user of the commodity.

The seasonal high in open interest for Soybeans (104%) is in the month of October. Corn open interest reaches its seasonal peak (107.5%) in November. Wheat open interest is bi-modal with a peak in June/July as the big Winter Wheat crop is harvested, and another high in October when the Spring Wheat crop comes in.

Relative lows in open interest in Wheat and Corn (below 95%) are posted in December. A possible reason: Wheat or Corn deliveries against the Dec. future can begin on the first business day in December. Conservative traders have exited from Dec. futures by the end of November. It is apparent from the seasonal open interest graphs that these positions are not automatically rolled into the next (March) contract month. Soybeans do not have a Dec. future listed so open interest holds fairly constant (at above 100%) thru the end of the calendar year.

Volume in the three main Chicago Board of Trade (CBOT) commodity futures tends to build in the early summer months, with a relative peak in June or July. Volume posts a relative low in the month of December for the same reasons discussed above.

B. Cattle, Hogs, and Pork Bellies

Cattle and Hog futures record their highest seasonal open interest index in March. The monthly average open interest then drops sharply at the beginning of the summer into a yearly low in August.

Pork Belly futures also exhibit a steep drop in open interest into the month of August. The seasonal decline produces the largest movement in the index readings of any of the 19 markets surveyed. Within a two month period (June — August) the open interest index declines from 117% to 70%. This very strong seasonal tendency must be taken into consideration by technical traders analyzing Pork Belly futures.

Volume in the Chicago Mercantile Exchange’s livestock futures is more stable than the well defined variation in seasonal open interest would imply. In fact, Cattle and Hog volume shows less month-to-month variation than grain or oilseed futures. Volume in Pork Bellies coincides with the unusually strong seasonal open interest influence in that market.

II. Interest Rates: Treasury Bonds, Treasury-Bills, and Eurodollars

Relative open interest highs in Treasury Bond futures occur every three months coinciding with the quarterly U.S. Treasury refundings in February, May, August, and November. Options on T-Bond futures (the most active quarterly series) also expire in these months. And physical delivery in T-Bonds can occur beginning on the first business day of the next month so speculative positions are often removed in February, May, August, and November, the months prior to the delivery month. With the U.S. Treasury now reducing the auctions of the 30 year Bonds to less than 4 times per year the seasonal influence may change, but not dramatically.

Open interest in the futures at the short end of the U.S. Dollar yield curve (Eurodollar and T-Bill futures) exhibit peaks at similar three month intervals and in the same months as the T-Bonds.

Volume movements in interest rate futures coincide with the relative monthly highs in open interest with one major exception. T-Bill volume is unusually low in the month of July, down to an index reading of 75%. The 75% figure is only reached on the T-Bonds and Eurodollars in the month of December. This year-end phenomenon, especially in T-Bond futures, has been observed by technical traders since the mid-1980’s.

III. Currencies: Deutsche Mark, Japanese Yen, and British Pound

The distinct peaks every three months on the D-Mark seasonal open interest graph can be explained by the contract specifications of the Chicago Mercantile Exchange (CME) currency products. Physical delivery takes place on a single day, typically the third Wednesday of the last month of each calendar quarter. Substantial deliveries cause the open interest to drop sharply (on a single day) in the months of March, June, September, and December.

A note on how the seasonal graphs are plotted is in order. The average volume and open interest for the month is plotted at the middle of the month. Total open interest in the CME currency futures usually rises into the delivery month. The method of plotting the monthly averages creates the impression that the relative highs in open interest occurs in the prior months of February, May, August, and November; on average it does, but the actual highest number of contracts open will typically occur in the delivery month. This must be taken into consideration when examining the seasonal open interest indexes of the currency futures.

The Japanese Yen and British Pound conform to what would be expected in open interest behavior with one curious deviation. Instead of an open interest peak in the month of August the peak shifts one month later to September. Open interest then returns to posting the expected cyclical highs in November, February, and May. The answer to the question of why this happens remains a mystery.

Volume in the currency futures records spikes in the quarterly expiration months of March, June, and September. This is different from the T-Bond and T-Bill financial futures which record the higher volume one month earlier. This is expected. The CME’s foreign exchange deliveries occur in the delivery month but, unlike the physical delivery contracts at the CBOT, the currency deliveries occur two days after the last day of trading. This means there is considerable trading of the expiring future in the delivery month without the threat of delivery.

There are two additional factors that contribute to the unusually high currency futures volume in the months of March, June, September, and December (although December is by far the lowest of the four months.) The Chicago Mercantile Exchange broadened its definition of volume in January of 1994. It now includes expired in-the-money futures options and futures deliveries in reported volume. Both of these anomalies occur on specific days in the quarterly delivery months. For proper day to day comparison of volume a classical bar chartist must remove the exogenous volume associated with those two days every quarter. The serial (monthly) futures options expirations do distort the volume at expiration as well, but to date these option series are not as heavily traded as the quarterly options.

IV. Metals: Gold and Silver

One only has to look at the contract months listed for trading in Silver to determine why the open interest hits relative lows in those months. Open interest drops when the lead contract enters the (physical) delivery month. Gold exhibits the same characteristics but to a lesser extent. It can be inferred from this seasonal study that expiring positions are not immediately rolled into the next contract month via a spread or switch order.

A comment on the scale gradients of the Silver and Gold graphs are in order. It looks like sizable swings occur in the open interest indexes from month to month. But unlike some of the other seasonal graphs where a 5% differential exists between the index readings the distance between the Silver numbers is 3%, and only 2% on the Gold graph. The overview is that the 16% range from high to low in Silver and (only) 6% high-low range in Gold means that seasonal open interest analysis is not as crucial as in, say, Heating Oil with its large 40% range between the open interest extremes.

Volume variations in these two metal futures are more extreme than monthly open interest fluctuations, with volume indexes often changing 40% from one month to the next in Silver. A distinct pattern exists. Volume peaks in the months prior to futures expiration months. As discussed in the T-Bill comment, a technician establishing the categories for high, average, and low volume must realize these distinct monthly volume traits.

V. New York Markets: Coffee, Sugar, Cotton, and Heating Oil

Of all the commodities analyzed in the previous study in the 1991 CRB Year Book, the change in the Coffee graph is the most pronounced. In the 10 years ending in 1988, Coffee moved within a maximum range of 14% in open interest and even less in volume. In the most recent 10 year study ending in 1996, the open interest range remained the same, but the lowest index readings of both open interest (89%) and volume (80%) occurred in the month of December. A definite year-end effect has developed.

Sugar showed no change in its tendency to build open interest from an October low. But like most of the U.S. futures, the month of December is now a decidedly lower volume month.

Cotton now has a sharper seasonal open interest peak in the month of September whereas the 1988 study showed a three month plateau of high open interest in September thru November. And December is the lowest volume month (80%) both then and now.

Heating Oil futures are the salvation of the futures industry in the otherwise lackluster month of December. Open interest is an above average 112% and volume is robust at 120% of the average month for the year. A futures technician will need to factor the distinct seasonal increase in both volume and open interest into any technical analysis of the internal make-up of Heating Oil futures.

VI. Stock Index Futures: S&P 500

Insight into the seasonality of volume and open interest in the S&P 500 futures is new to this CRB study. The graph definitely shows a volume surge into the end of the first calendar quarter (March), summer doldrums (July/August), and steady but below average volume in the last quarter of the year.

A more interesting phenomenon is the clear tendency for open interest to escalate to higher peaks as the year progresses. This is not simply due to the growth curve in open interest over the 10 year period in the study. Each calendar year tends to exhibit the same seasonal increase. For example, total open interest in S&P futures in 1993 increased from 186,253 on March 15, to 206,384 on December 15, a 10.8% expansion. In 1996, open interest over this same time frame increased from 187,367 to 242,828 contracts, a 29.6% expansion.

A possible answer is that fund managers, under increasing pressure to report favorable year-on-year return comparisons, are hedging (some long, some short) as the calendar year-end approaches. In any event, a technical S&P trader must revise high, average, and low volume parameters every quarter to insure they are appropriate. Since open interest changes are the important technical input, the expectation is that larger actual daily changes will occur as the year progresses and incrementally each quarter as the expirations approach.

Summary and Conclusion

Futures technicians have taken seasonal variations in volume and open interest in the traditional agricultural and livestock futures into consideration for many years. This updated analysis shows that the shape of the seasonal graphs of these commodities have not changed. Including the possibility of a seasonal influence on volume and open interest enhances the results of an investigation into the health of a price trend on any futures chart.

Proper analysis of the financial instrument futures requires a knowledge of the contract specifications. These include whether the contract is cash or physical delivery, when delivery takes place, and when options on the futures expire.

Most noticeable in this updated study was the larger month-to-month changes in volume than recorded in the previous CRB study in 1988. This is due to the increase in options volume and especially the expanded definition of volume by the Chicago Mercantile Exchange. The additional volume in CME futures due to the inclusion of in-the-money exercised options and futures deliveries must be removed for proper day-to-day comparisons.

Another ramification of the unequal monthly volumes: the setting of high, average, and low volume categories must take into consideration any distinct seasonal volume peaks and troughs which occur over a very short time period.

The CRB publishes a weekly chart magazine, CRB Futures Perspective. The task of analyzing seasonal open interest considerations is made easier by the inclusion of a plot of average open interest for the previous six years on the daily charts. This provides a quick and easy visual comparison to current open interest movements.

CRB Yearbook CD. Contains all articles since the 1965 Edition!

Editor’s Note: Additional insights into interpretation of volume and open interest are available in Mr. Shaleen’s book, Volume and Open Interest: Classic Trading Strategies for 24-Hour Markets. IRWIN Professional Publishing, 1997.

Kenneth H. Shaleen is President of CHARTWATCH, an international research firm to the futures industry. CHARTWATCH distributes a weekly technical publication and produces a daily telephone market update covering the financial instrument futures. Mr. Shaleen has instructed technical analysis courses for the Chicago Mercantile Exchange since 1978. These courses are also conducted in Singapore, Malaysia, Hong Kong, London, and numerous other locations throughout the world. Mr. Shaleen is the author of Volume and Open Interest: Cutting Edge Trading Strategies in the Futures Markets (Probus 1991); Technical Analysis & Option Strategies (Probus 1992); Volume and Open Interest: Classic Trading Strategies for 24-Hour Markets (IRWIN Professional Publishing 1997); three Chicago Mercantile Exchange course books as well as numerous articles. Mr. Shaleen can be contacted at CHARTWATCH, Fulton House, 345 Canal Street, Chicago, IL 60606, 312-454-1130; or at his website. www.chartwatch.com