Crops & Shoulders
The primary function of a spread is to regulate supply and demand
through time. The pricing mechanism can encourage consumption
sooner rather than later or ensure supply for when needed most.
Consider soybeans. It can vary some by latitude, but the bulk of
the US crop is planted May/June and harvested October/November.
The USDA crop marketing year runs September-August. Because so
few newly harvested soybeans are available for delivery against
September CBOT futures, November is considered the first and most
representative new-crop contract. In South America, Brazil
harvests mostly April/May and Argentina (the larger exporter)
May/June. Thus, CBOT May futures now most purely represent US
old-crop soybeans because new South American supplies cannot
transit in time for delivery --- making May/November Soybeans a
pure US old-crop/new-crop spread.
The function of such a spread becomes obvious. If physical
supplies are burdensome, then the contract for delivery sooner
rather than later is more likely to trade at a discount in order
to encourage more rather than less consumption and sooner rather
than later. In contrast, when physical supplies are tight, that
more nearby contract is more likely to trade at a premium in
order to discourage consumption or delay it until the next crop
is at least assured if not available.
Normally, old-crop soybeans weaken from the beginning of the new
calendar year into February due to producer selling in the new
tax year, logistical difficulties, an approaching seasonal
decline in domestic soymeal consumption, and anticipated export
competition from new South American harvests. Old-crop/new-crop
spreads tend to follow lower. By early March, however, the
market has usually discounted these pressures. With the crop
marketing year now half over, far more than half last year's
production has already been consumed. The remaining old-crop
supplies must stretch until the next harvest --- still but a
twinkle in producers' eyes. In the Midwest breadbasket, corn is
planted first and, all else equal, preferably. Thus, soybeans
often need to rally enough to "buy acreage" away from corn and to
place a risk premium into price to compensate for the possibility
of erratic weather.
This spring rally is usually led by old-crop, supplies of which
will only get tighter. Higher old-crop prices tend to encourage
acreage, thereby placing some pressure on new-crop prices. Thus,
old-crop has usually outperformed new-crop from early March into
May. For example,
the
Long May/Short November Soybeans
spread has closed more favorably toward old-crop May on about April 18
than on about February 26 in 12 of the last 15 years --- suffering
daily closing drawdown greater than 5.00 cents/bushel in only 4
years. (The CBOT requires minimum margin of $338. Brokers will
accept spread orders.)
So far this year, large old-crop supplies and expectations for
sharply reduced soybean acreage in favor corn (to meet ethanol
demand) have conspired to drive this spread to as low so far as
-38.25 cents/bushel. That was the biggest discount in
May/November spreads since March-May 1983 (which, by the way,
immediately preceded a major weather market that drove August
Soybeans from under $6/bushel in late June to above $9 prior to
expiry). The only two years in history in which discounts were
wider (between -60 and -70 cents/bushel) were 1980-81, when
soaring US interest rates sharply increased carrying charges.
With nearby soybeans at about $7.80/bushel, maximum CBOT storage
premiums at about 4.50 cents/bushel/month, and the prime rate at
8.25%, full carry for this spread is now 63.00 cents/bushel
($7.80/bu x 0.0925/12 mo = 6.00 cents/bu/mo; 6.00 + 4.50 = 10.50
cents/bu/mo; 10.50 cents/bu/mo x 6 mo = 63.00). Thus, at current
prices, the discount has a theoretical maximum of -63.00. Given
that such discounts rarely exceed 80% of full carry, the more
realistic maximum tradable discount would be -50.00. Thus, risk
would appear to be minimal with potential for reward unlimited.
Not to suggest it will do so this year, but note that the
May/November 2004 version traded at a premium of almost +$3.00
before expiry.
The USDA Forum is scheduled for March 1-2, from whence will come
estimates for corn and soybean acreage. Will and, if so, when
will bullish speculators forced to liquidate March futures prior
to First Notice Day to avoid deliveries reinstate those positions
in May or July? For chart aficionados who prefer to buy
breakouts from trading ranges, this spread has been unable to
narrow within -34.00 for six weeks now.
More adventuresome traders might consider natural gas. Retail consumption
expands during winter's heating season and again during summer's cooling
season. Periods in between each, when consumption contracts, are known as
"shoulder months." The heating season ends with March, and the
cooling season --- when gas is used to generate electricity to run
air conditioning --- begins again with June. Thus, April and May
constitute spring's shoulder months.
But that is also a period of transition. With total national
inventories normally near annual lows, selling is muted but
distributors in hot-weather regions need to accumulate inventory
to meet retail demand during peak cooling season. But timing can
be everything. Natural gas is a market whose price structure
tends to place greater premium into months of higher consumption.
For example, even as consumption usually is low in April, those
inventories need to be growing into May. Thus, as soon as April
becomes lead contract, cold-weather distributors with excess
supply may sell into it whereas hot-weather distributors are
beginning to price supply for delivery in May and beyond. So May
futures usually outperform April during the first several days of
March. In fact,
the
Long May/Short April Natural Gas
spread has closed more favorably toward May on about March 11 than on about
February 28 in 14 of the last 15 years --- suffering no daily
closing drawdown whatsoever in 9 of those years, and none greater
than $0.018/mmBtu. (The minimum increment of $0.001/mmBtu is
worth $10.00. NYMEX requires minimum margin of $1,350 in this
volatile market. Work closely with your broker on placing any
orders because execution can be, ahem, "difficult.")
This year's spread has generally favored May from a discount most
of last year all the way into early January, when it reached its
so far peak premium of +$0.110/mmBtu. The spread then narrowed
back toward even money but held late in the month at 0.020. By
mid February it had widened back to as much as 0.093 before
pulling back modestly. The spread has been coiling between
January's high and its low, with February's low so far at 0.045.
The April contract will be front month upon March's expiry on
February 26.
The CRB Index appears
to be trying to break upward again.
If you have not already done so, you are encouraged to
visit
spread charts.
(Both for future convenience and for reference to past
Commentaries, you can bookmark the WSC Index.)
Please remember: These are NOT trading recommendations. They are intended only as potential
ideas based on the market's own performance in the past, but past performance is not necessarily
indicative of future results. Futures trading involves substantial risk of loss.
If you have any questions or comments,
e-mail
me or call me directly at 541-484-7256.
Trade 'em
Jerry Toepke
| 