A probability theory problem - page 4
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
In the financial market (specifically Forex) there are both types of events - dependent and independent. And the dependent ones are just the main ones, while the independent ones are what you call noise.
Can I give you an example of dependent events in Forex?
I will not go into details, I will say in brief. Forex (with some assumption) works on the principle of "communicating vessels" in physics. If you put pressure somewhere, it will definitely come out somewhere else. (sorry for the somewhat "flat" model as an explanation). Starting with the fact that in forex there is no buying and selling, but only exchange (a fact that is a revelation to some). All addictions start with this. Many (I don't mean you personally) try to look at one column (of communicating vessels) to determine how (where and by how much) the level of this column will change without seeing (not noticing, not taking into account) all the others. Take any cross, for example if GBP/USD goes up, USD/JPY goes down, GBPJPY definitely goes up - what is not the dependence?
And there is a mathematical expression for this relationship. Once again, this only applies to Forex. The precious metals do not follow this equation :-) Unfortunately. But something tells me that you know all that, then maybe I did not understand the question.
I admit that my task is not set correctly. I cannot clearly articulate what is in my mind. BUT, the probability of guessing the next candle for certain past combinations is not always 1/2. I found the probability of guessing the next candle with probability 0.76 (the highest probability of all obtained) based on statistical data.
PS: If someone else is working in a similar direction - we can combine forces.
You would then at least clearly articulate the direction.
If you are interested in point 3, then read the branch 'Market condition - flat or trend? Which is dominating?" some of the work has been done. Interesting results there, a lot of thoughts 'between the lines', for the careful reader.
Despite the fact that you will find patterns in the area you are researching now, the patterns you find will not give any particular advantage in trading, because the dependencies are wider.
But the fact that you started with research is a good thing.
...Take any cross. For example, if GBP/USD goes up, USD/JPY goes down, then GBPJPY definitely goes up.
And there is a mathematical expression for this relationship.
GBPUSD * USDJPY / GBPJPY = 1 Always. So?
So GBPJPY is called a "synthetic" cross rate. So?
"Primitively/classically arbitrage" (using this formula), particularly in DC, is NOT.
Although yes - in this case this formula shows a "dependent event", but ... with the balls already pulled, and in no way future ones.
And another thing - what does "the trend of the fir-friend" have to do with the colour of a particular candle, and on an unknown timeframe?
1. And that's why GBPJPY is called a "synthetic" cross rate. So?
2. "Primitively/classically arbitrage" (using this formula), particularly in DC, is NOT.
3. Although yes - in this case this formula shows a "dependent event", but ... with already pulled, not future, balls.
And 4 - what does "the trend of the fir trend" have to do with the colour of a specific candle, and do not know what timeframe?
1) Nobody is prohibiting you to trade on it.
2. We are not talking about arbitrage
3. The question was about dependent events, not about how to apply them in trading.
4. No
GBPUSD * USDJPY / GBPJPY = 1 Always. So?
So GBPJPY is called a "synthetic" cross rate. So?
"Primitively/classically arbitrage" (using this formula), particularly in DC, is NEVER possible.
Very much possible, in my time I synthesized GBPNZD, because most brokerage companies simply did not have this pair, and where I did, the spread was so large...
The problem comes down to the correct calculation of lots included in the synthetic mojor cross.
And there is a mathematical expression for this relationship. Once again, this only applies to Forex.
These dependencies are understandable even for beginners, but unfortunately they are of no (or almost no) practical use, because there is no lag/reversal in the markets and none of the pairs can be used as a leading indicator.
The question is about dependent events within one pair. For example, if we observe 8 consecutive rising bars closing on the H1 chart. Can we say with a probability higher than 0.5 that the 9th bar will close down? Or in other words, does the probability of the 9th bar closing depend on the fact that the previous 8 bars in a row were rising? By the way, 8 and 9 bars of the same colour especially on an H1 chart is an extremely rare event. So, if someone uses similar statistics tonight on the pound, after 8 bars rising, would open in the short, he would have a floating loss of 50 points or more and would be able to close in the black by some miracle. I was investigating similar dependencies yesterday - today and came to the conclusion that there was nothing (almost) to catch. All I could squeeze out of the statistics was about 400p over 3 years. There are also a couple of hypotheses that need to be tested.
The conclusion: the probability of occurrence of the next bar practically does not depend on the history of previous ones, i.e. these events are practically independent.
Speaking about independent events, I meant similar events. Certainly, it is clear that if EURUSD rises and GBPUSD falls, EURGBP cross also rises. And its value can be named without opening the chart. :)
1. these dependencies are clear even to beginners, but unfortunately they are of little or no practical use because there is no lag/overrun in the markets and none of the pairs can be used as a leading indicator.
2. The issue of dependent events within the same pair. For example if we observe 8 consecutive rising bars closing on H1 chart. Can we say with a probability higher than 0.5 that the 9th bar will close down? Or in other words, does the probability of the 9th bar closing depend on the fact that the previous 8 bars in a row were rising? By the way, 8 and 9 bars of the same colour especially on an H1 chart is an extremely rare event. So, if someone uses similar statistics tonight on the pound, after 8 bars rising, would open in the short, he would have a floating loss of 50 points or more and would be able to close in the black by some miracle. I was investigating this kind of dependencies yesterday-today and came to the conclusion that there was nothing (almost) to catch. All I could squeeze out of the statistics was about 400p over 3 years. There are a couple more hypotheses that need to be tested.
3. The conclusion: the probability of occurrence of the next bar almost does not depend on the history of previous ones, i.e. these events are nearly independent.
Speaking about independent events, I meant similar ones. Certainly, it is clear that if EURUSD is rising and GBPUSD is falling, EURGBP cross is also rising. And its value can be named without opening the chart. :)
1. I gave pairs only as an example
2. Then I really misunderstood the question, because the way I see it, it's not a dependence, but a statistical pattern. You will only get the answer to this question by doing statistical research. Just off the top of my head I would say that yes it is possible with more than 50% probability, for example 60/40, but that doesn't rule out the remaining 40% as in your example today on the pound. For a more detailed answer we need to collect statistics: how many similar combinations (8 candles in a row) were in the studied area, how many of them ended with the opposite candle. But this is not all. In this case, you'll get an approximate probability of occurrence of a certain event - the opposite candle, but you won't have an answer to the question what the value of this candle will be, and the risk of getting on the opposite, but very small candle is a subject of a separate study.
What I mean by dependencies is different. Not the dependence of the next price on the previous one and your conclusions (3) are most likely correct here, but what the price (ratio) really depends on.
If your conclusions are confirmed by statistical studies, it would be useful to get acquainted with them.
3. if your conclusions are supported by statistical studies, it would be useful to read them.
Unfortunately, I have not kept them, as they have no practical value. But in the near future I will repeat a few runs from memory and post them with the EA.
The problem comes down to the correct calculation of the lots included in the synthetic cross of mojores.
I have tried to play around with "lot sizes" myself and have read/discussed the calculations in other conferences .... at best ... -3*spread :(
If only we could find "dependent events" in the formula EURUSD * GOLD * CocaCola * ???, then we could play with them, because, imho, no "Consortium"/DC would have time to level them. ;);););)
But it did not work too profitable to generate GOLD buy/sell signals "as a function" of EURUSD. :(
Not everything is as simple as balls in a poke :(
The problem comes down to the correct calculation of the lots included in the synthetic cross of mojores.
I have tried to play around with "lot sizes" myself and have read/discussed the calculations in other conferences .... at best ... -3*spread :(
If you are interested in synthesizing of artificial cross, write what pair should be obtained and in what brokerage company. We will solve it.
Recently, I synthesized an artificial futures cross from 6B and 6J, which I used to hedge GBPJPY. GBPJPY opened long to catch maximum positive swaps, synthesized futures cross 6B6J to short without swaps. But the idea did not work out, because the price difference (basis) between the futures and the underlying asset is slightly, but exceeds the swap profit. Everything has already been accounted for by the market. I tried to use the Expert Advisor to catch the delta of the basis and to play on it, but it is within the spread.
If only one could find "dependent events" in the formula EURUSD * GOLD * CACILA * ???, then one could play with them, because, imho, no "Consortium"/DC could compensate them. ;);););)
But it did not work too profitable to generate GOLD buy/sell signals "as a function" of EURUSD. :(
Not everything is as simple as balls in a bag :(