Your Market Beliefs - page 10
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
Simba,
1) It's all relative. I would say no.
2) Again, it's all relative. I'm not in the business of prediction.
3) It's a program called Statgraphics, I just got introduced to it a month ago.
There are many statistical software packages that may be more comprehensive that this, but this program is solid and very user friendly. A non-hardcore statistician like me can get the hang of it. I love how it has like 20+ distribution curves to select from, and lots of transformations available. Check it out. The training is also very good, the professor is the designer and trainer for the seminar.
Fractal in Nature
My research concludes for now that the current viewable universe is fractal in nature. Unending and infinite resolution. (See "Benoit Mandelbrot")
There is always a glitch in the matrix to exploit. The glitch itself is part of the fractal properties that paints the canvas for our senses to wonder.
Keep in mind... just as soon as too many start to see and predict the patterns... the patterns react and become unpredictable.
The observer creates by viewing.(See "What the Bleep", What the Bleep Do We Know!? & What the Bleep!? - Down the Rabbit Hole)
It is a secret law of the creator to never have the creation become bored.
Boredom is hell, stagnate, and unhappy.
The best we can do is is surf.
Try to position ourself for the best wave... but as soon as matrix code is cracked...low tide.... no more waves..
We can etch out a profit during this time... but with the weight of observance... it will all have to change soon.
In our short life span, we usually will not see these patterns but as I said... there are glitches...
This gives the illusion of a never ending, ever-changing fractal pattern.
We are currently on the edge of seeing the glitch. Time for a large pattern shift. For as soon as too many can see the glitch... all the pretty colors change...LOL
I hope I didn't lose anyone... I am simply saying.... as soon as we figure this out.... the rules will change.... But like a good gambler.... we keep placing our bets !!
Take care all !
JB
cointegration
, a set of 100 moving averages will do the trick
look at the EURUSD and USDCHF , if both above the 100 MAVG then sell both if both below then buy both ... , If you want a more complicate setup then use two MAVG instead of one and trade the cross...
I think this is the Alexey's article Simba mentioning about From chaos to trends in forex | Futures | Find Articles at BNET
Shanon Entropy is another interesting subject CB pays attention to determine the trend:
Stochastic Volatility
The other idea we will outline this time is stochastic volatility. If you use a nice charting tool such as those found in TradeStation etc and in the IG-Index Java trading platform, everyone will have noticed that as you look at the same FX pair at different time-frames, the series looks very noisy at some resolutions and much nicer and smoother at others. The smoothness does not just increase with timeframe -- I have seen the GBPJPY look pretty horrible at M1, M3, M5, M10 then look really nice at M15 but then go back to pretty nasty at M30, H1. This behaviour is due to a phenomenon called "volatility clustering". Folk who recommend trading using charts on multiple time frames are (whether they realise it or not) exploiting this property and it is a good idea. The amount of actual information, of course, increases as the resolution falls -- if you have the M1 data you can easily reconstruct the M3, M5, H1, H4, ..., etc data, but if you only have the H4 data you can't recover the information at finer time scales. However, the "mutual information" is quite different (we touched on this much earlier in this thread). If you compute the Shannon Mutual Information at different timescales, you will see that the curve has dips (actually peaks, but I use an algorithm that computes things upside-down for reasons that don't concern us here). What are we looking for here? We want to find something we can actually trade!! We want a genuine trend we can latch onto and avoid the dangerous shoals of false trends and ranging markets.
Think of this in terms of a communication system. The transmitter can send either a 1 or a 0 and the job of the recevier is to decide whether a 1 or a 0 was sent. (This is an accurate description of any modern digital comms system, eg, mobile phones). The radio channel adds noise and randomly affects the signal, sometimes changing 0's into 1's and vice versa. If the channel is noise free and we always receive a 1 when a 1 was sent (same for 0) then everything is great. You might think the worst case would be a very noisy channel that always changes 0's for 1's (and vv). But this is easy to handle too -- just invert everything at the receiver! The worst case is actually the channel where 1's are changed to 0's randomly so the receiver can no better decide what symbol was sent than by tossing a fair coin! In digital comms systems, we use powerful "error correcting codes" to code the information prior to transmission to combat the effects of the radio channel, whose properties are quite well known in advance.
So in our trading system, if we have our FX data at a given timescale whose mutual information is close to 0.5, we can no better decide whether to buy or sell (go long or short) than by tossing a coin. WE CANNOT TRADE THIS INSTRUMENT AT THIS TIMEFRAME!! However, if we find a timeframe where the mutual information is biased away from 0.5 (either up or down) then we are in with a statistical chance -- we have an edge and that is all we need. We do not need to win every battle so long as we win the war!! You will have seen from the equity curves of my robot that he is by no means always right, but that overall the equity curve is inexorably upwards!
I think this is the Alexey's article Simba mentioning about From chaos to trends in forex | Futures | Find Articles at BNET
Shanon Entropy is another interesting subject CB pays attention to determine the trend:
Stochastic Volatility
The other idea we will outline this time is stochastic volatility. If you use a nice charting tool such as those found in TradeStation etc and in the IG-Index Java trading platform, everyone will have noticed that as you look at the same FX pair at different time-frames, the series looks very noisy at some resolutions and much nicer and smoother at others. The smoothness does not just increase with timeframe -- I have seen the GBPJPY look pretty horrible at M1, M3, M5, M10 then look really nice at M15 but then go back to pretty nasty at M30, H1. This behaviour is due to a phenomenon called "volatility clustering". Folk who recommend trading using charts on multiple time frames are (whether they realise it or not) exploiting this property and it is a good idea. The amount of actual information, of course, increases as the resolution falls -- if you have the M1 data you can easily reconstruct the M3, M5, H1, H4, ..., etc data, but if you only have the H4 data you can't recover the information at finer time scales. However, the "mutual information" is quite different (we touched on this much earlier in this thread). If you compute the Shannon Mutual Information at different timescales, you will see that the curve has dips (actually peaks, but I use an algorithm that computes things upside-down for reasons that don't concern us here). What are we looking for here? We want to find something we can actually trade!! We want a genuine trend we can latch onto and avoid the dangerous shoals of false trends and ranging markets.
Think of this in terms of a communication system. The transmitter can send either a 1 or a 0 and the job of the recevier is to decide whether a 1 or a 0 was sent. (This is an accurate description of any modern digital comms system, eg, mobile phones). The radio channel adds noise and randomly affects the signal, sometimes changing 0's into 1's and vice versa. If the channel is noise free and we always receive a 1 when a 1 was sent (same for 0) then everything is great. You might think the worst case would be a very noisy channel that always changes 0's for 1's (and vv). But this is easy to handle too -- just invert everything at the receiver! The worst case is actually the channel where 1's are changed to 0's randomly so the receiver can no better decide what symbol was sent than by tossing a fair coin! In digital comms systems, we use powerful "error correcting codes" to code the information prior to transmission to combat the effects of the radio channel, whose properties are quite well known in advance.
So in our trading system, if we have our FX data at a given timescale whose mutual information is close to 0.5, we can no better decide whether to buy or sell (go long or short) than by tossing a coin. WE CANNOT TRADE THIS INSTRUMENT AT THIS TIMEFRAME!! However, if we find a timeframe where the mutual information is biased away from 0.5 (either up or down) then we are in with a statistical chance -- we have an edge and that is all we need. We do not need to win every battle so long as we win the war!! You will have seen from the equity curves of my robot that he is by no means always right, but that overall the equity curve is inexorably upwards!Mystified,
Thanks for the links..The article I was mentioning can be found here
The Trend Determining Method
Regards
Simba
Aleksey
Hi simba
So I read this Aleksey threads on FF
Intresting ..........so see if i got it:
he wrote at one point:
Maybe it is difficult to use this hybrid method wholly in the beginning so I advise you to start with the simple variant, namely to use this fixed basket:
EURUSD - the most forecastable pair because of its huge stream of money,
GBPCAD - the best pair among GBP-contained ones as less correlated with EURUSD,
CHFJPY - the only JPY cross is not subjected to "carry trade",
AUDNZD - the pair is determined local trade relations.
It is my favorite basket. Have successful research.
So this means that when i got a sell signal with whatever tradingsystem i use, because he gave not one...smile.....for instance on EURUSD
Instead of selling 4 lots ond EURUSD
I sell
1 lot EURUSD
1 lot GBPCAD
1lot CHFJPY
1lot AUDNZD
Am i right in this or?
yes of course this idea is great to have cointegrated basket but it seems that cointegration pairs are not so simple to filter........
Hi minime
can you give an example of your suggestion with the 100 MAVG?
And simba can you point me to the post where you did this fuller test with clahn?
Thanks
lodol
My research concludes for now that the current viewable universe is fractal in nature. Unending and infinite resolution. (See "Benoit Mandelbrot")
There is always a glitch in the matrix to exploit. The glitch itself is part of the fractal properties that paints the canvas for our senses to wonder.
Keep in mind... just as soon as too many start to see and predict the patterns... the patterns react and become unpredictable.
The observer creates by viewing.(See "What the Bleep", What the Bleep Do We Know!? & What the Bleep!? - Down the Rabbit Hole)
It is a secret law of the creator to never have the creation become bored.
Boredom is hell, stagnate, and unhappy.
The best we can do is is surf.
Try to position ourself for the best wave... but as soon as matrix code is cracked...low tide.... no more waves..
We can etch out a profit during this time... but with the weight of observance... it will all have to change soon.
In our short life span, we usually will not see these patterns but as I said... there are glitches...
This gives the illusion of a never ending, ever-changing fractal pattern.
We are currently on the edge of seeing the glitch. Time for a large pattern shift. For as soon as too many can see the glitch... all the pretty colors change...LOL
I hope I didn't lose anyone... I am simply saying.... as soon as we figure this out.... the rules will change.... But like a good gambler.... we keep placing our bets !!
Take care all !
JBWOW! Poetic Chaos
.
Hi simba
So I read this Aleksey threads on FF
Intresting ..........so see if i got it:
he wrote at one point:
Maybe it is difficult to use this hybrid method wholly in the beginning so I advise you to start with the simple variant, namely to use this fixed basket:
EURUSD - the most forecastable pair because of its huge stream of money,
GBPCAD - the best pair among GBP-contained ones as less correlated with EURUSD,
CHFJPY - the only JPY cross is not subjected to "carry trade",
AUDNZD - the pair is determined local trade relations.
It is my favorite basket. Have successful research.
So this means that when i got a sell signal with whatever tradingsystem i use, because he gave not one...smile.....for instance on EURUSD
Instead of selling 4 lots ond EURUSD
I sell
1 lot EURUSD
1 lot GBPCAD
1lot CHFJPY
1lot AUDNZD
Am i right in this or?
yes of course this idea is great to have cointegrated basket but it seems that cointegration pairs are not so simple to filter........
Hi minime
can you give an example of your suggestion with the 100 MAVG?
And simba can you point me to the post where you did this fuller test with clahn?
Thanks
lodolLodol,
No...that is not the way to do it.
1-Aleksey`s explanation was focused on 2 things,1-his system,which,you are right,he explained,very partially at FF(then he explained in full,or at least at a high level of disclosure, at Stocks and Commodities)..2-How to devise a hybrid basket,for ANY Trading system,in order to smotth the returns...they are 2 different things.
2-He explains very well how to do the basket...Let`s presume you have a System,ANY system..then you take the following currencies..AUD,CAD,CHF,EUR,GBP,JPY,NZD,USD...then you can have 7+6+5+4+3+2+1=28 pairs to trade with...once you get a signal in ,for example GBPUSD,YOU DELETE all other pairs containing either GBP or USD,and you look for signals on the rest...let`s presume you get a signal on EURCHF,same process..until you end with 4 pairs..for example buy GBPUSD,sell EURCHF,sell AUDCAD ,buy NZDJPY...if he has found only one pair as per his signals he just trades it...if he has 2,3,or 4 pairs,he knows his returns will be smoother...he has shown examples at his thread,sometimes trading GBPJPY as standalone or in combo with EURUSD..4 pairs is the optimal,but they are not necessary.
3-1 lot each...mmm...I don`t remember what/if he wrote on this specific,I just suggest that you normalize for volatility and value of position in order to have similar expected returns for each of the positions.
4-His system..at S&C he explains it...basically the entry is composed of a series of non linear equations on price,volume(I think he uses the volumes at dbfx..any ECN volumes will be ok) and open interest(he uses open interest of Futures markets as a proxy for cash markets)...The exit is done basically on a RBCI with parameters 8,10,40,44...easily replicable with the digital filter`s generator,smoothes it for 25 periods and then he applies 2 std volatility bands,and exits when the band is touched....He uses D1 data.
But this is off topic,even if probably very interesting for most traders,so,let`s stick to Mr Marketz topic .
Here are the links from FF both for Aleksey`s thread and for CB one for those wanting to learn more
The basket of Hybrids; 4 Pairs
EURUSD forecast & Chaos theory
Optimized Trend Trading
The latetst link is where both clahn and myself did a Dickey Fuller test on excel following CB suggestions...or so I believe
Regards
Simba
Hi minime
can you give an example of your suggestion with the 100 MAVG?
Thanks
lodolI wont take any credit for myself , I learned correlation from reading this thread
http://www.goldenmoneytree.com/forum/viewtopic.php?t=974&postdays=0&postorder=asc&start=75&sid=3eb47ccd7bf3c39dd7486fa03735c2ae
Sparxfxis the real deal he is a trader who makes money and there are few around to learn from these days
hi
hi thx simba
hi minemi
yes but thats about correlation not cointegration...
thx