How are you with a market mindset? - page 5

 
Martin Cheguevara:

Let's start with the basics... the fundamentals so to speak...

Can anyone here even tell the difference between a chart of quoted price data and a chart of something incomprehensible?)

The answer is no one.

No one can tell me in the world where the "bad" charts end and the "right" ones begin.)

No one can tell me with 80-90% confidence which charts are pricey and which are not of that kind - I'm 200% sure of that)

And it follows that no one can tell me the pricing formula)

And then, in my opinion, it is worthwhile to talk about any formulas).

Do you want to make sure or do you agree with my statement?)

I'm not committing to anything...it's just my opinion.

Although in fact it is quite justified.

If everyone agrees...then rightly consider the matter closed for myself =)
I can tell a random graph from a real one. Not by eye, of course, but by analyzing distributions.
 
Vladimir:

And I want to make sure, and I disagree for the case of currency pairs. Their quotes and charts have a natural property of their own, only not individual for each pair, but a group one. If we take 3 or more currencies V1, V2,...Vn (not pairs, but currencies), then the pair rates Vi to Vj, equal by definition to the purchasing power ratio Pi/Pj (for example GBP/JPY, USD/CHF, EUR/USD...), will together be calculated through each other using obvious formulas (the same Pi/Pj). This is a fundamental property of forex rates, deviations from it will be studs, and execution means correctness of quotes in this sense.

Out of 8 currencies USD EUR GBP CHF JPY AUD CAD NZD due to specifics of rate definition (fraction, division) one can be assigned any value of purchasing power, the other 7 Pi can then be changed independently of each other, and all of the 28 currency pair rates will be directly calculated by the seven independent Pi. Checked it repeatedly, using this fact to account for driving forces in trading (usually called cross deviations from calculated values). That is, 28 different rates actually have exactly 7 degrees of freedom.

Here it would be interesting to know what methods of formal data analysis are able to reveal these exact patterns in the history of rates.

For example, machine learning neural networks - can they? If they can't even identify the dimensionality of the problem, then what can they do?

Do more widely known methods, such as autocorrelation and factor analysis, albeit on plots of almost linear changes in Pi, recognise this hard factual relationship?

If anyone is familiar with the question, please advise which ways you can identify these actual patterns from the known history of 28 exchange rates (or at least the USD EUR GBP triplet)... Suppose not even the regularities themselves, but only the number of independent degrees of freedom. After all, this is the key property, the dimensionality of the problem.

Cointegration to your aid...

 
Alexey Volchanskiy:

Don't stop the theorists from figuring out who has the longest ))

ahahahahah)))))))

 
Maxim Romanov:
I can tell a random graph from a real one. Not by eye, of course, but by analyzing distributions.

1718 1782 1704 1635 1696 1604 1669 1720 1618 1565 1620 1708 1661 1624 1573 1502 1613 1560 1498 1525 1635 1623 1619 1619 1545 1493 1468 1406 1478 1414 1524 1541 1607 1636 1701 1648 1597 1588 1549 1556 1534 1528 1537 1567 1660 1672 1616 1631 1604 1664 1638 1628 1688 1777 1880 1832 1749 1705 1802 1856 1831 1913 1983 2026 2105 2173 2282 2397 2444 2530 2601 2542 2442 2480 2533 2605 2549 2460 2585 2659 2587 2611 2557 2474 2469 2576 2486 2524 2465 2515 2519 2523


220 221 221 221 221 220 219 219 220 220 220 220 220 222 222 223 224 229 228 227 227 228 230 229 228 228 229 228 228 227 228 228 227 225 224 225 226 225 230 229 229 227 227 229 229 227 226 223 223 220 218 219 216 218 219 223 221 217 216 219 218 219 218 221 220 221 222 222 223 222 224 223 222 221 221 222 222 222 221 222 218 218 219 219 221 220 221 219 221 222 221 221 222 220 220 220 220 220 219 220 220 220 220 220 220 220 221 218 219 219 221 221 221 221 221 221 221 224

Well define)

 
Олег avtomat:

This has happened before... It's a familiar scam... And the methods are thrifty... Watch your hands. And he's already claiming to be a master, it's in your subcortex, and as if no proof is needed, you've already believed the lie... believe it? Do you really believe it?

We have already discussed this with you.

Show us the result and then we'll talk.

 
Martin Cheguevara:

We have already discussed this with you.

Show us the result and then we will talk.

Before you can demand something from your opponent, you have to do it yourself. That is, show a master class in trading).

 
khorosh:

Before you can demand something from your opponent, you have to do it yourself. I.e. show a master class in trading).

You're absolutely right!)


There's a 98% chance my system will work that way.

If there was a free server I would have thrown it in there and stupidly given the result.

I don't know how to do it with MT4...

Enlighten me.

PS: I'm not a scalper, i'm not a statistician.

My system is statistically dependent and stable. So it is acceptable to increase deposit and risk linearly.

And in no way has any martingale, netting or similar nonsense.

so on the opening prices of hourly bars the test is enough for me.

The quotes were downloaded from Dukaccopy.

I can't provide anything else.

 
Vladimir:

And I want to make sure, and I disagree for the case of currency pairs. Their quotes and charts have a natural property of their own, only not individual for each pair, but a group one. If we take 3 or more currencies V1, V2,...Vn (not pairs, but currencies), then the pair rates Vi to Vj, equal by definition to the purchasing power ratio Pi/Pj (for example GBP/JPY, USD/CHF, EUR/USD...), will together be calculated through each other using obvious formulas (the same Pi/Pj). This is a fundamental property of forex rates, deviations from this will be the studs, and execution means correct quotes in this sense.

Out of 8 currencies USD EUR GBP CHF JPY AUD CAD NZD due to the specifics of rate determination (fraction, division) one can be assigned any value of purchasing power, the remaining 7 Pi can then be changed independently of each other, and all of the 28 currency pair rates will be directly calculated from the seven independent Pi. I have checked this repeatedly, using this fact to account for driving forces in trading (usually referred to as cross deviations from calculated values). That is, 28 different rates actually have exactly 7 degrees of freedom.

Here it would be interesting to know what methods of formal data analysis are able to reveal these exact patterns in the history of rates.

For example, machine learning neural networks - can they? If they can't even identify the dimensionality of the problem, then what can they do?

Do more widely known methods, such as autocorrelation and factor analysis, albeit on plots of almost linear changes in Pi, recognise this hard factual relationship?

If anyone is familiar with the question, please advise what methods can be used to identify these actual patterns from the known history of 28 exchange rates (or at least the USD EUR GBP triplet)... Suppose not even the regularities themselves, but only the number of independent degrees of freedom. After all, this is the key property, the dimensionality of the problem.

Ifthe Gibbs-Rosebohmtriangle is expanded to an N-dimensional simplex then it follows from it:

The dollar index is a double value calculated by a formula kindly provided to meby Neutron

,

where USD/YYY is all direct quotes, like USD/CHF, XXX/USD is all inverse quotes, like EUR/USD.

And indices of all other currencies. And the chain paths vary.


 
Martin Cheguevara:

You're absolutely right!)


There's a 98% chance my system will work that way.

If there was a free server I would have thrown it in there and just given the result.

I don't know how to do it with MT4...

Enlighten me.

PS: I'm not a scalper, i'm not a statistician.

My system is statistically dependent and stable. So it is acceptable to increase deposit and risk linearly.

And in no way has any martingale, netting or similar nonsense.

so on the opening prices of hourly bars the test is enough for me.

the quotes were downloaded from Dukaccopy.

Well another tester's grail. Every first one here can draw them, every second one is not even worth talking about.
Oh, and about the free VPS. There are a lot of them now. Yes, you need to meet some conditions, but for free you have to sacrifice something. And there are quite cheap and stable options for payment.
 
Martin Cheguevara:

Co-integration will help you...

For what? I don't need help, as I said, I've been using the above distinction between forex quotes and "just charts of obscure things" for years.

I'm wondering if the dimensionality is defined by formal analysis tools. In particular, even for the simplest case of EURUSD EURGBP GBPUSD - do methods of, for example, machine learning, the branch of which has more than a thousand pages, analyzing tens and hundreds of signs, reveal that these three time series are tightly and precisely related by the simple equation EURGBP = EURUSD / GBPUSD?

Are there examples where IO diagnoses the presence of such dependencies? Or is this a super task for MO? It seems to be a thousand pages long, one can detect such a key ability of the method...