The goal of this thread is to keep track of ideas, and to receive community feedback; with the purpose of exploring the value and role of statistics with respect to day trading in the financial Markets. I will be building statistical models and combining them to HOPEFULLY build a statistic based discretionary trading model.
To me statistical analysis of the financial markets can be used to accurately represent the collective experience that is normally developed over years and years of trading.
In all honesty, I wasn't entirely sure on where to begin with this thread, but the only way to reach your goal is to just start, it doesn't matter where, just get going.
Two words that I'm certain will get thrown around a lot in this thread are statistics, and probability. So for reference I'm including their definitions in the first post, to try and ensure they are used correctly.
Statistics: the analysis of numerical data in large quantities.
Probability: The extent to which something is probable. (USE THIS WORD WITH CAUTION, when I use It I am referring to “the likeliness of the outcome of the next single or series of events”).
I will try to keep the thread organized to the best of my ability, but be forewarned, I have horrendous attention deficit disorder. And not in the way that “everybody is a little A.D.D.”, but in the way that without medication I cease to function with any beneficial degree. But regardless, I will try to hyperlink important and beneficial posts to the first page of the thread.
All of my data related contributions to this thread will be derived from market data pertaining to the ES (E-mini s&p 500 futures contract).
I would like to start with a few questions:
How do you view or interpret statistics in the market?
What questions can we answer using statistics?
What questions are impossible to answer using statistics?
Is it possible to trade from statistics alone?
What do you think? am I chasing my tail on this one?
Last edited by addchild; December 27th, 2011 at 02:31 PM.
Reason: removed a redundant word.
The following user says Thank You to addchild for this post:
Not that I am an expert on the matter, but seems like the thread so far is just looking for a place to start?
I would say Taylor has work in developing a trading book that holds various metrics about price movement. In organizing the book, you wind up with metrics that can easily be aggregated in various ways if one so desires, i.e. averages, ranges, distributions etc.......
I think in keeping statistics about price movement, one has to pick REFERENCE points in order to be able to keep metrics.
For example, one could ask the simple question....."How far does price typically move?" This question is meaningless without a reference point. To make the above question more meaningful, one could re-phrase the above question with respect to a reference point. A'la Taylor, that question would be, "How far does price typically move beyond the prior day's Hi/Lows?
So not that I am saying Taylor is the answer, but I think there are definitely some useful ideas there, at least to begin thinking about the topic.
Point being, stats require a point of reference from which to measure. Attached is a Taylor book of ES.
The following 3 users say Thank You to forrestang for this post:
Ok, so to date I have been importing price action daily into excel. But what I would like, is to have excel automatically import and update price data. I have searched, and tried out a few things, nothing is satisfactory.
Any ideas on programs that will automatically import to excel? (openoffice calc?) Ideally it would be free, or atleast have a 20 day demo or something, but I am open to spending money if its a quality solution.
Hi addchild - I was wondering how far along you got on your statistical trading endeavor? I'm interested in the subject matter and particularly about using non-parametric stats and ohlc data to have an expectation about what "should" happen and add an overlay of volume profile/discretion for entering and managing a trade with a 1 - 5 day holding period. If there are any resources or threads that you found helpful I would love to read them.