The Demand-Speed Index: Capital Flow, Economic and Fundamentals Indications. 6/6/04
DRAFT FOR REVIEW AND DISCUSSION Only, CAUTION: READ DISCLAIMERS BELOW.
FOR EDUCATION OR EVALUATION PURPOSES ONLY. NOT INTENDED AS TRADE RECOMMENDATIONS.
For the week beginning: June 6, 2004 Beta Test Version 03.9
Testing at over 84% accuracy (Primary Signals), use technical signal confirmations to reach the 95% accuracy goal.
Signals are between 0 and 7 days EARLY. Often, expect one wave AGAINST the move before triggering.
To enter, use stop entries just out of hourly (chart) range, or wait for hourly technical signals to trigger the trade.
Then, chase the trade with close stops throughout the move. While "Position Trades" are not intended, if you find
yourself in one, use wider daily chart Sup/Res or Elliott wave stops to capture the longer term move.
Watch your stops, severe market events or breaking news can and do occur without indication in this index.
Trading is risky! .05% stops suggested.
THE ORDER THAT FOLLOWS IS:
***Commodity Symbol(s), followed by:
A. Expected direction this week, if any is strongly indicated by the demand curve.
B. Probability estimate of that direction, based on type of demand curve, slope, variance from mean, and/or demand signal.
C. Crossover of demand curve and direction, if any. (For your information only, not always a reliable signal.)
D. Net slope market direction indication. (For your information only, not always a reliable signal.)

***ADOLLAR D. LONG?, Slope data only, No Strong Signal B.~61% C. X +Up
***BPOUND A. LONG Signal B.~84%
***CDOLLAR D. SHORT?, Slope data only, No Strong Signal B.~61%
***USDOLR A. SHORT Signal B. ~74%
***EUROcur D. LONG?, Slope data only, No Strong Signal B.~61%
***JYEN D. LONG?, Slope data only, No Strong Signal B.~61%
***PESOMQ D. LONG?, Slope data only, No Strong Signal B.~61%
***SFRANC D. LONG?, Slope data only, No Strong Signal B.~61%
***CRUDEny D. SHORT Signal B. ~70% C. X -Dn
***HEATOILny D. SHORT?, Slope data only, No Strong Signal B.~61%
***UNLEADny D. Flat?, Slope data only, No Strong Signal B.~61%
***NATGASny D. SHORT?, Slope data only, No Strong Signal B.~61%
***EUROdolr D. LONG?, Slope data only, No Strong Signal B.~61% C. X +Up
***5YRNOT A. LONG Signal B. ~74% C. X +Up
***2YRNOTE D. LONG?, Slope data only, No Strong Signal B.~61%
***10YRNOT A. SHORT Signal B.~84% C. X -Dn
***TBOND30 A. SHORT Signal B.~84%
***SOYOIL D. LONG?, Slope data only, No Strong Signal B.~61%
***CORN D. LONG?, Slope data only, No Strong Signal B.~61% C. X +Up
***W-KCBT D. LONG?, Slope data only, No Strong Signal B.~61%
***W-MN D. LONG?, Slope data only, No Strong Signal B.~61%
***OATS D. LONG?, Slope data only, No Strong Signal B.~61%
***RICE D. LONG?, Slope data only, No Strong Signal B.~61%
***SOYB D. SHORT?, Slope data only, No Strong Signal B.~61% C. X -Dn
***SOYML D. SHORT?, Slope data only, No Strong Signal B.~61%
***W-CBOT D. LONG?, Slope data only, No Strong Signal B.~61% C. X +Up
***?ESsp-OUT D. LONG?, Slope data only, No Strong Signal B.~61% Caution!
***?NQ-OUT D. LONG?, Slope data only, No Strong Signal B.~61% C. X -Dn Caution!
***?DJ-OUT A. SHORT Signal B. ~74% C. X +Up Caution!
***FEEDCTL D. SHORT?, Slope data only, No Strong Signal B.~61%
***LCATTLE D. LONG?, Slope data only, No Strong Signal B.~61%
***LEANHG D. LONG?, Slope data only, No Strong Signal B.~61%
***PBELLIE D. LONG?, Slope data only, No Strong Signal B.~61% Caution!
***GOLD D. LONG?, Slope data only, No Strong Signal B.~61% C. X +Up
***COPPER D. SHORT?, Slope data only, No Strong Signal B.~61%
***PLATINM D. SHORT?, Slope data only, No Strong Signal B.~61%
***SILVER D. LONG?, Slope data only, No Strong Signal B.~61% C. X +Up
***COCOA ny A. LONG Signal B.~84% Caution!
***COTTONny D. SHORT?, Slope data only, No Strong Signal B.~61% Caution!
***OJny D. LONG?, Slope data only, No Strong Signal B.~61% Caution!
***COFFEEny A. LONG Signal B.~84% Caution!
***RLUMBRny D. LONG?, Slope data only, No Strong Signal B.~61% Caution!
***SUGAR ny A. LONG Signal B.~84% Caution!
BACKGROUND
Accuracy concerns intrigued us in 2003. We realized that adding additional studies to an already long string of studies
might not solve the problems, yet 70% accuracy was unacceptable. We suspected the fundamental information was not accurately
factored since we had no econometric models for the purpose. We suspected that measuring the economic flows by day, week and
commodity in an economic "index" might help, but the impact of worldwide flows are such that they must be also be considered. To reach a
demand-flow measure, we created an economic index using trading volumes and marginal demand measures for each commodity,
considering magnitude, time, risk, historical demand, and political stability factors. (This model changes often.) In early 2004,
we correlated the included factors with the outcomes, fortunately resulting in a simplification of the formulae and processes involved.
Marginal speed and demand considerations are the only focus, price and supply are EXCLUDED from this study.
METHODS SUMMARY
1. Measure the capital flowing into/out of each commodity in major markets adjusted for politics/economy, as above.
2. Complete the necessary statistical measures (needed for steps 3 and 4). Then create the C1-Demand Change index.
3. Regress/correlate these economic flow levels with the actual price changes of the commodity (in the US).
4. Create Marginal-speed-demand charts on a percentile base and software to interpret the charts, resulting in the recommendations above.
5. When economic flow levels reach statistically extreme levels/targets, expect the trend to change/continue within 0 to 7 days.
5a. With some exceptions,financial indices do not correlate as well: i.e.. DJ, ES, SP, etc., but the signals often help.
5b. Low volume/volatile/NY markets correlate well but can still surprise us because of their rapid reversals: i.e.. PB, NG, OJ, LB, KC, etc.
6. When economic flow levels continue and remain within statistical bounds/targets, expect the trend to continue.
7. Excluding 5a.items, limited back testing suggests 84% or higher primary accuracy. We continue adding or adjusting the accessible factors.
8. Seek technical confirmation of these identified trend changes before trading, thus pushing accuracy into the targeted 95% + range.
9. Based on the statistics, we evaluate the probability of being correct, creating "Primary", "Secondary", etc. observations weekly.
10. This speculation is one-to-seven-day trading. Many trades can go longer, occasionally blooming into positions, but
position trading is not directly intended.
11. The trends are statistically reanalyzed and recalculated by commodity, by week. Each commodity (of 43) and time frame is unique.
The standard deviations, variances, slopes, confidence intervals, etc. change for each week and each commodity.
(I am a private, individual, trader and make no representations as to qualifications in creating this or any other system.)
May the fortunate trades be with you. Good Luck.
DISCLAIMER: There is risk of loss in all commodities trading. Please
consult a knowledgeable broker before you trade for the first
time. Losses can exceed your account size and/or margin requirements.
Commodities trading can be extremely risky and is not for everyone. Some
option and commodity strategies have unlimited risk. Educate yourself on
such risks and rewards prior to trading. We, and anyone associated with
us, do not guarantee profits or predetermined loss points, and are not held
monetarily responsible for the trading losses of others, clients or
otherwise. Past results are not indicative of potential future returns.
The report(s) may contain recommendations in which trading or performance
results are mentioned. These results are hypothetical, as they are not the
result of trades which have been placed for execution, and which do not take
into account market conditions and other factors. The use of stops or other
loss prevention techniques may not function as expected.