Who keeps a record of price and time data for their market? I am talking about the duration of movements (time) and the high - low difference (price). Has anyone constructed a successful trading model based on analysis of this raw data? What sort of information would it be useful to analyse? For example, the mini Dow index futures (/YM@CBOT). Here is part of a spreadsheet. What is recorded? First print at the 09.30 opening, last print at the 16.00 close. Previous days close (to calculate an opening gap magnitude). Then each significant intraday swing, with the high/low print of the move and the minute in which this occurred. With this sample screenshot shown, we see that there was a trade either at the open or within 7 minutes of the open in 10 out of 11 days. All of these were for greater than 20 points. So it would seem that if one knows the direction to trade at the open, this is the fastest 20 points on offer - or $100/contract gross. Further refining to know whether to trade at the open, or at a time within the next seven minutes would be helpful. As would the direction - perhaps could be calculated with reference to the prior N days net change on the day, the opening gap, and other supporting parameters? Could this model be expanded to give us even more information - likely time windows for intraday turning points? Taking this data going back to the very start of the YM contract, or back 3 years, 5 years etc - how much data do you think should be recorded. How long do the "patterns" persist for? 2008 onwards was volatile compared to say 2003. Would there be merit in taking so much data? What about the most recent 6 months only? What do folks here think? Has anyone done this type of analysis?
From your screen shot, I'm having trouble understanding what leads you to the above observations. Can you take us through the process of how you get there, perhaps by identifying cell co-ordinates (e.g. âfrom column F on 6/6/11, we see that â¦â)?
totally useles IMO. the price can start to rally when ever the big traders want. There is no rule that says it must start then or then.
+++1 There are still too many amateur here think that by looking at the PA or some bs indicator, they can predict the future up or down. So far seemed no one really understand what is "market" mean for .. For big money, the buy or sell signal is definitely not due to some PA or indicator.
yeah, the only indicator is when their CEOs tell them to buy the shit out of the market.... But still then the Big Boys WAIT for the RIGHT TIME POINT, that is not a certain time you see on your clock, NO its a definition of price patterns inside other price patterns, where the best spot is to start the trade and push the price to start the rally..... What do you guys think ? Trading is fucking hardcore ? You must come up with more, than just a price & time statistic !!!
@ treath opener If you think you are tough, forget it !!! All the stuff what you may think out one day, was all made before, by guys with much more brain and balls than you.... Everything was already a lot of times there, but it is keeped secret and will be. This Trading Stuff is about making millions of millions. You must be better than the rest of that lousy lazy idiots, who think trading is easy. only then you might have a little chance to win. cheers
OP , Keeping stats of each swings' avg , min , max, duration and price distance traveled gives you a framework for filtering signals and applying probabilities to the market structure unique to your chosen market. Keep at it and learn your markets oscillatory tenancies to the point where it becomes very familiar territory.
As to the OP's original question/topic: I have been keeping and analyzing various spreadsheets with different metrics, and chopping up various values and statistics for about 6 months now. I have scripted various forms of data collection and mining to accelerate the process of finding useful relationships. I have tried various forms of aggregation, separation etc. I am sure there is something there of value, I still have yet to find it. An example of a few of the sheets I have. There are two examples here. I have many other types though. http://imgur.com/MB70Q
Got quite a few more metrics to look at. Often I've heard the answer being number based, and not chart based. Ok so I'm trying to look at my metrics with that outlook, my metrics are a combination of various vectors and timings associated with them. So one thing I am wondering, is the key simply the proper aggregation and compilation of various statistics? I.e. is it just a matter of "8 out of 10 times price did this?" -OR- Is it that once the data is organized in a specific matter, the path becomes obvious? Then of course there is the matter of actually having useful metrics to track in the first place which I suppose is part of the task.