All Things Statistical - page 8
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Statistical Indicator Accuracy and Issues
Found some interesting studies from FF by Proximus that i could not resist taking a further look (errrm TSD & Mladen pls dont kill me for showing links from other forums
)
This 1st part showcases Proximus's study of the various indicator accuracies via his Genetic Algorithm. Good for general reference
Statistically testing every indicator with my Genetic Algoritm @ Forex Factory
M30 Timeframe, EURUSD
Simple MA Cross: 45%
Money Flow Index: 42%
Parabolic SAR: 39%
Ichimoku: 37%
RVI: 34%
MACD: 32%
Bullpower: 23%
BearPower: 22%
Single Simple MA: 16%
=============================================
On a different thread, Proximus does his experiments with creating a filter via a short term mean of a weekly/daily nature as assistance of gauging volatility, but concludes that it fails eventually as the data distribution isnt coming even close to a normal distribution....
Also, Proximus shows some interesting views, even if one might/might not agree with it, for instance like MA usage which some (not all) have managed to use effectively too... His views as belows:
"And very bad results, the best edge i got was ~0.3. And its no wonder as the MA sometimes goes flat around consolidation periods, its too smooth, and cannot adapt very well to sudden price surges, many HIGH/LOW's shoot out from the mean like a rocket, and the MA doesnt even move a pip in that direction"
"That would be nonsense, as this doesnt tell me anything about future events, in a time series like the market if i put on a trade at X index, what is the guarantee that X+1 index will behave as the mean suggested at X.If the trend is smooth then atleast i can estimate the likelyhood of it to continue"
"The regression doesnt really adapt to market shape thus its impossible to measure SD on it, also it doesnt tell me anything about volatility, the variance can be distributed infinite number of ways around that line"
"The second problem is that, after the filter is too complex, it can be always written as a simpler model, for example an ARI2MA with X,Y parameters is 99% indentical with a SMA [X], so its worthless to calculate the complex algo, when its almost identical with the simple one.There are exceptions, but basically the more complicated the filter becomes, there will be always a filter which will be identical with it and less complicated"
http://www.forexfactory.com/showthread.php?t=503680
Wintersky
Cheers
that's depressing...
that's depressing...
Hahahaha you mean the accuracy figures? Cheer up Pava! Life is exciting!@@@
Hahahaha you mean the accuracy figures? Cheer up Pava! Life is exciting!@@@
you're cherry picking right now with this statement of yours
There is another statical method mainly used by biostaticians and ecenometricians is least squares CUSUM:Statistical Process Control (SPC) is a scientific approach to quality improvement in which data are collected and used as evidence of the performance of a process, organisation or set of equipment. One of the SPC techniques, the cumulative sum (CUSUM) method, first developed by E.S. Page (1961), uses a series of cumulative sums of sample data for online process control
An Error Occurred Setting Your User Cookie
Dear Mladen his warning Method Not ImplementedPOST to /newreply.php not supported.
Posting to forum is complete mess , can you check with the developers.I get t
There is another statical method mainly used by biostaticians and ecenometricians is least squares CUSUM:Statistical Process Control (SPC) is a scientific approach to quality improvement in which data are collected and used as evidence of the performance of a process, organisation or set of equipment. One of the SPC techniques, the cumulative sum (CUSUM) method, first developed by E.S. Page (1961), uses a series of cumulative sums of sample data for online process control
An Error Occurred Setting Your User Cookie
An Error Occurred Setting Your User Cookie
(statistically insignificant)
Dear Mladen ,
I can not post a message properly, it is a complete mess can you delete my posts please!
Dear Mladen , I can not post a message properly, it is a complete mess can you delete my posts please!
ROFLMAO. It is okies nevar.... It is good sometimes to leave indelible marks of history for the future generations...
Thanks for the sharing Nevar. I think i posted CUSUM indicator on this forum before, but not Least Squares CUSUM. If it's Least Square CUSUM, then we are referring to the structural stability of the residuals around a straight line. Interesting definitely, but the main issue is the baseline level to compare against.
Wintersky
Cheers