Mostly that's because these folks don't understand much about statistics or how to do Out of Sample testing, not because the method has a defect. To even start on the road to systems what will perform the same in real life you need to: 1) Develop metrics for market topology 2) Divide all available historical data into numerous segments with statistically representative coverage in each of all market topology classes. 3) Develop or train the system on one group of datasets called Training 4) Optimize or validate the proposed system on another group called Test. When satisfactory Freeze everything 5) Test Frozen on the final group called Validation or OOS (Out of Sample) 6) Compare performance metrics of the system between OOS and Test 7) Remix the segments between the three groups and repeat the entire process 8) Repeat 7) until all data has been used for each of the three logical groups and a set of common rules and definition parameters is both the same and works acceptably well for all repetitions. 9) Compare the differences in performance metrics between all of these. If they are within abut 6.5%, in for example W/L %, you have a system that will perform the same in live trading as long as the market topology of the live market has been represented in the process. If the market shows a significant new topology characteristics and performance metrics also vary stop trading immediately. If the performance metrics are similar but topology changes trade with extreme caution as the market is in a new âspaceâ. If topology and performance metrics are similar and stable to development trade away!
Yes, it can.... it requires working n-dimensional hyperspace and non-linear mathematical processes. Solutions to problems almost always occur in a higher order of dimensionality then that in which the problem is stated.... otherwise the solution would be obvious and it would not be a problem or question.
One of the best (ignored by daytraders) indicators is basic fundamentals. Pull up a multi-year chart of USD for instance, mark turning points & locate financial news at the time. The old adage is - no risk, no champagne
Wow ! Thanks Jack ! I will peruse it some more when my head quits hurting from reading it the first time.
It works well and consistently especially with a drove of indicators. Nice thing about indicators ... they don't just reveal the market ... but they also reveal the predictable actions of the big fish who decide on similar indicators.
The mathematical problem of the markets is precisely solved using the mathematics required and stated explicitely by the deductive hypothesis set (HS) of the market and its parametric measurement set (PM) of the market. I'll add another sentence. The mathematics of the market is Boolean Algebra and the parametric set dimensions applied to the P and V of the market are binary vectors. A vector has dimensions of magnitude and direction. Attached is a double chart pertinent to the prior question asked of me for another measure of the market. Here you see the Dow Jones used as a leading indicator of the S&P. The specific problem solved for trading the markets, is to determine whether a trend (a price change segment over time in one direction) is "continuing" or "changing".
Premium is by and large a late indicator useful for confirming good trades. On occasion it will also be the beginning of a good trade. FWIW you should be in front not behind premium spikes. Cheers
Website is down: We have termporarily suspended our website. Please contact us directly at (212) 808-0500. Thank you.
It is difficult to explain. I did try. Premium spikes are something I do not use for anything. In terms of Larry Harris and his hierarchy of trader and investor types, I am front running the smart money. So in his book where he uses 32 boxes to describe these folks, put me down as a person who is parasitic and technical and frontrunning. Recently I posted with respect to comparing Premium in two ways. I posted a chart that did NOT have any Premium spikes on it. The bars show the relationship of smart money's "bias" to the premium when smart money is trading the YM or the big contract. There were no minis when this was originally dealt with a while back. Since the Premium (as I indicated, is an FA, derived indicator), it is fairly steady during a day. But as I stated it does drift. Bots and algos take advantage of this when it is out of line by a certain amount. If you are in a pro trading room and they use algos or bots, you can watch the agitation of the players as they watch the time coming up when their algos and bots are going to trigger. It is almost akin to the noise formerly heard in the background during the days of phone trading before the minis were invented. None of the drift is shown on the charts I posted. I monitor to trade in front of the algos and that is another matter for first and second derivatives as well. A person has to differentiate drift from a "trending day". That is not part of the ET range of considerations as yet.