UsdJpy DAR
GbpUsd Statistical Probability With Price Action 109 replies
EurJpy Statistical Probability Trading With Price Action 46 replies
GbpJpy Statistical Probability With Price Action 22 replies
Statistical Validation on trading 24 replies
DislikedFor those of you that haven't been following the thread,
I was inspired by goodthing's work and decided to do my own stats for GBP USD using 8 years of historical data.
Here are some results (finally) based on weekly highs and lows:
In the past 8 years consisting of 410, 5 day trading weeks
GBP USD has retraced to
61.8% : 254 times, for a 62% probability retracement to this level WITH respect to Bullish/Bearish movement during the same week or the following week
50% : 292 times, for a 71% probability
38.2% : 335 times, for an 82% probability
23.6 : 378 times, for a 92% probability
For only 32 weeks out of 410 weeks has GBP USD not retraced to any fib level during the same or following week and simply continued trending towards an extension for an 8% probability OR just traded in a very narrow range not going up or down..
I will have more interesting things to know about the last 8 years soon.
Some notes regarding probability:
Probability ONLY means that an event is likely to occur, it does not necessarily mean that if there is a 92% probability of something happening then then it's sure as hell gonna happen as opposed to an event with 8% probability, but that is the most likely event based on past data.
GOODTHINGS: Thank you for the eye-opening experience in this thread.
P.S I'd initially tried to apply logic for retracement before the week's end in Excel but it was waaaay too much headbanging and seems irrelevant to trading realistically, but if someone is an Excel guru reading this I'd appreciate some help to do it anyway.Ignored
DislikedI have no doubt a credible strategy may be developed based on historical retracements, after all people keep telling us/me that trading is ALL about high probability.. so what could be better than actual statistical probabilities which you've pointed the results are related to a pattern of human emotion which seems to be fairly consistent.
In order to develop a strategy the questions asked from the data must be more refined.. like IF Movement = Bearish AND Daily XX EMA Slope = Bearish THEN IF Ret. (weekly)= 61.8%-+10pips AND Price FAILS Beyond THEN Sell .. Hell a lot of thinking must go into this and also much price action discretion must be applied
Bemac is doing a great job but is trading very low timeframes for me. For me, TP's would be based on dynamic fibs from an MA applied to median/closing price.
Who is this Jr97? I don't see him anywhereIgnored
DislikedI'm working on a PHP/SQL page which you can upload exported charts history and it will tell you the retracements and whatnot for any time period you select, average daily, weekly, monthly range, how many times the range hit 100, 120, 150, etc etc.. just been really busy with work and learning MQL. Will post it in here when I am finished.Ignored
DislikedFor those of you that haven't been following the thread,
I was inspired by goodthing's work and decided to do my own stats for GBP USD using 8 years of historical data.
Here are some results (finally) based on weekly highs and lows:
In the past 8 years consisting of 410, 5 day trading weeks
GBP USD has retraced to
61.8% : 254 times, for a 62% probability retracement to this level WITH respect to Bullish/Bearish movement during the same week or the following week
50% : 292 times, for a 71% probability
38.2% : 335 times, for an 82% probability
23.6 : 378 times, for a 92% probability
For only 32 weeks out of 410 weeks has GBP USD not retraced to any fib level during the same or following week and simply continued trending towards an extension for an 8% probability OR just traded in a very narrow range not going up or down..
I will have more interesting things to know about the last 8 years soon.
Some notes regarding probability:
Probability ONLY means that an event is likely to occur, it does not necessarily mean that if there is a 92% probability of something happening then then it's sure as hell gonna happen as opposed to an event with 8% probability, but that is the most likely event based on past data.
GOODTHINGS: Thank you for the eye-opening experience in this thread.
P.S I'd initially tried to apply logic for retracement before the week's end in Excel but it was waaaay too much headbanging and seems irrelevant to trading realistically, but if someone is an Excel guru reading this I'd appreciate some help to do it anyway.Ignored
DislikedWhen feeding in daily data, the problem was that there's no time signature ..Ignored