Daily Speculations
 |
James Sogi
Philosopher, Juris Doctor, surfer, trader, investor, musician, black belt, sailor, semi-centenarian. He lives on the mountain in Kona, Hawaii, with his family. |
9/09/2005
Pendente Mercor by
Jim Sogi
The statistics reading gives bias and edge to initiate a trade and during a trade. That is critical. Once in, after manifestation of an idealized pattern, the continuing development of the bars will not necessarily continue to be significant as the random variable kicks in. Diebold analyzes the prospective error factor in a good forecast in the t+i as being pure noise. The accuracy of a prediction is best one bar ahead and added randomness creeps in further steps ahead. So as a trade progresses, the one bar ahead reading remains the most accurate. The debate is between the traders who insist on trading the initial idealized pattern through as per the original optimum stats at trade entry vs adjusting in the middle. The initial idealized trade is more like the Sports Illustrated Swimsuit issue, just too perfect, and as the date wears on, the makeup wears off, just like the trade. The trade that looked so good at the beginning, even as it get profitable and because it gets profitable, gets less likely to continue so in many cases. Counter intuitively, a trade that goes against, may get even better odds in trader's favor.
In real time, as a trade progresses, the original edge will deteriorate, but new developments will still give you a reading on the extent of your edge. This edge is a a guide to leverage. To quote Chairman Vic's talk at the Rose Garden Conservatory "size should be inversely proportionate to your edge"
The question is whether it is better to use idealized data patterns with high probabilities, or use real days data that may not be as s#xy, but still gives an edge. Ideally the parameters used will uncover the ideal patterns while incorporating the daily nitty gritty.
c
Jim Sogi, May 2005
(address is not clickable)