From the December 01, 2010 issue of Futures Magazine • Subscribe!

Cotton: Following the thread

Price results from demand and supply forces, which generally remain in balance. Periodically, significant factors arise which cause an imbalance directly affecting price. We term these unexpected events affecting all real commodities as “shocks:” prices rise or decline dramatically in reaction. The shock is invariably followed by a return to levels near prior equilibrium. The astute longer-term trader may benefit from riding the wave forward and then using the shock to time probable market reversals (see “Using price shocks to time the market,” November 2009). We built a model for each commodity to determine when it went into shock. In “Shocking profits in new markets” (December 2009), we noted the last soybean signal generated was a “buy” in February 2009 at $8.72 per bu. Current price is $12.44 and we remain long beans.

We also described a useful trading heuristic in “Predicting price targets with the rule of seven” (May 2010). Measuring the first leg of any move from bottom to wave top projects four future wave tops. This can be employed on minor, intermediate and major bases. The upside waves tend to top at 1.75, 2.33, 3.5 and 7 times the ‘length’ of the first wave, or bottom at 1.4, 1.75, 2.33 and 3.5 times the downside first wave.

Cotton provides an excellent example of the combination of these two tools. Using the principles of shock, our cotton model examined the current monthly closing price vs. a three-month moving average of the prior year’s price and the price two years earlier. Examination of a long term chart in cotton clearly depicts the shocks. Cotton closed in June 1986, at 32.20¢, having fallen 47% in one year and 55% over two years. The bottom came at 29.50¢ in August 1986, when price closed at 35.28¢ per lb. followed by a surge to 75.81¢ (117%) within one year. Other historic bottoms were 28.20¢ (October 2001) and 36.70¢ (November 2008). Shock top examples were 97.77¢ (September 1980), 94.45¢ (May 1991), 114.65¢ (June 1995) and 84.80¢ (October 2003).

Examining the data, we found our key levels for the one-year comparison to be 34.8% for bottoms and 47.7% for tops; for the two-year, key levels were 36% and 54% respectively. A simple set of rules combines the two indicators and allows for immediate action if extreme price levels down or up (50% and 126%) are hit. The two-year model warns of the short trade and the one-year model signals the trade unless the extreme is hit. Essentially, this combined and refined our basic cattle and hog models described last year and recognized price tends to ‘run’ longer on the upside than the down.

Table A (Below) depicts the result applied to historical data keeping the model in the market at all times; 13 trades were made. The last long trade doubled up in March 2009, reinforcing the December 2004 buy signal. The final price in October 2003 is estimated at $1.23. The model had 100% winning trades for a total gain of 604.78 points or $302,390 profit. Applying $100 per contract slippage results in a net gain of $301,090 over 37 years — roughly $8,100 annually and the average trade lasted just under three years.

TABLE A

Trade

Date Open

Date Close

Open Price

Close Price

Net

Drawdown

Short

Dec-73

Dec-74

88.75

36.85

51.9

0

Long

Dec-74

Nov-76

36.85

79.11

42.26

0

Short

Nov-76

Nov-77

79.11

51.76

27.35

0

Long

Nov-77

Sep-80

51.76

89.12

37.36

0

Short

Sep-80

Jul-86

89.12

32.13

56.99

6

Long

Jul-86

Jul-87

32.13

79.3

47.17

0

Short

Jul-87

Feb-92

79.3

56.7

22.6

8.94

Long

Feb-92

Jun-95

56.7

86.06

29.36

5

Short

Jun-95

Oct-01

86.06

29.85

56.21

3.53

Long

Oct-01

Oct-03

29.85

76.73

46.88

0

Short

Oct-03

Dec-04

76.73

44.77

31.96

0

Long

Dec-04

Oct-10

44.77

123

78.23

0

Long

Mar-09

Oct-10

46.49

123

76.51

0

The worst closing month drawdown was 8.94 points. Only four of the 13 trades experienced any end-of-month drawdown. The aggressive trader could add a second contract upon trade initiation, scalping a 20-point profit every time.

Cotton went into shock and our two-year model produced a sell warning based on the September 2010 close. The October 2010 close will be the sell signal.

It also was possible to approximately predict the likely high to be achieved in this move using the Rule of 7. The major low of 28.20¢ was hit in October 2001. Intermediate monthly highs were 60.55¢ (March 2003) and 84.80¢ in October 2003. Table B (below) depicts application of the Rule of 7. The reader will observe the first target of the intermediate wave is 84.81¢ and the actual high is 84.80¢. The first target of the major wave is 127.25¢, and the high price to date is 130.50¢. When price traded above the second intermediate target of 103.58¢ in September 2010, the first target of the major wave became highly probable. A trader could remain long above 104¢ confident that price would likely obtain the 124¢—130¢ range. The shock nature of the trade provided the impetus and the Rule of 7 projected the target rewarding the long term trader’s patience with a gain of roughly 76 points since March 2009 and a combined gain exceeding 150 points since December 2004. A break above 131¢ could cause cotton to continue to run to 141¢ or as high as 160¢, but these prices are progressively less likely.

TABLE B

Intermediate
High: 60.55

Low:
28.2

Major
High:
84.8

Wave:

32.35

56.6

Ratio:

Goal:

Target:

Goal:

Target:

1.75

56.61

84.81

99.05

127.25

2.33

75.38

103.58

131.88

160.08

3.5

113.23

141.43

198.10

226.30

7

226.45

254.65

396.20

424.40

If history repeats, the monthly closing price in October 2010 should be a major high and provide the long term trader with a significant selling opportunity, while keeping in mind the risk associated with such a trade.


Dr. Arthur Field has a Ph.D. from Clemson and is a former fund manager for Fidelity International. He wrote “The Magic 8: The Only 8 Indicators You Need to Make Millions in the Markets”, available at www.themagic8.com. Email him at
themagiceight@hotmail.com.

Comments
comments powered by Disqus