I've just completed THREE more trades. I mixed the trials for each of the 3 separate trades together randomly and continued to add trials until I reached a consensus. After about 150 trials for each of the three predictions (which took almost 3 weeks to complete), all three of them showed a similar pattern, so the decision was made to stop adding trials and risk capital on all three predictions.

TWO of the three were successful and ONE was not successful. The unfortunate news is the one trade that was wrong was WAY wrong!! Here are the details:

Trade 14 - S&P $5000 Trade 15 - YEN $5000 Trade 16 - Cattle (-$16,060)

Total Profit to date: $21,265 # trades correct: 11/16 = 69%

I suppose in retrospect, to limit and control the down-side risk to the trades, I should have used a money management STOP LOSS. That's where I can exit the trade immediately if the price reaches a predetermined level where losses reach a certain percentage of total capital. The problem with stop losses is sometimes the market reaches your stop loss point, talking you out of the trade at your predetermined loss and then turn around and close the period at a profit. Those cases would result in a NEGATIVE profit, but I would be correct in predicting the final outcome of the trade. If I had used a $5000 stop loss in this case, the three trades would have resulted in a combined profit of $5000 rather than a loss of $6060. However, if I had used a stop loss for ALL 16 trades to date, it is hard to say for sure that some of the profitable trades would not have resulted in the stop loss getting inadvertently hit resulting in a LOSS rather than a PROFIT.

My goal is to find a method of predicting trade outcomes that is 100% successful regardless of how much work is involved. The idea starting with trade 12 has been to continue to ADD individual ARV trials until a statistically significant consensus has been reached. It has worked 4 out of 5 times until the big $16,000 loss with trade 16. The question now is: "Why was Trade 16 incorrect?"

There are two possibilities:

1. I did not reach REAL significance with the 150 trials that made up Trade 16. Rather than considering ALL of the 150 trials, I consider only a sub-set of the trials - trials that occurred during certain times (lunar phase, solar wind speed, local sidereal time, and a few other proprietary indicators). In the case of all three trades, that subset of trials clearly predicted the outcomes that I went with. It is possible that due to the limited number of trials resulting from the filtering, that I am not considering the TRUE statistics - i.e.: what seems significant may in fact still be due to chance.

2. It may be that 100% success at using ARV to predict future outcomes is impossible. For some unknown reason, even if I added a million more trials to trade 16, and amassed a HUGE statistical argument for the price of live cattle to go UP last week, It could still go against my prediction and go DOWN. Maybe the 'PROBABLE' future outcome was UP, but something less probable derailed that possibility, and the market went the opposite way.

At this point, I'm still very much in favor of possibility #1 - that I made some errors in evaluating the consensus of the trials for trade 16. I do believe that it is possible to amass enough consensus that I could be 100% correct. Logically, it would make sense that if there is significant evidence that a random future event is predictable (and there is!), then ALL random future events are predictable. So far, there does not seem to be any measurable difference between the degree of predictability and the type of event, or probability of that event occurring. The answer may be doing even MORE trials!!!

What do YOU think?

Best regards, Greg K

...