Demonstrating the cluster approach to market... - page 3

 
Mathemat >> :

4. I've been trying to find a common trading equivalent to the notion of state entropy - and I think I've found it. It's liquidity: the higher the entropy of state, the higher the liquidity.

5. One last thing. If there are any experts in statothermodynamics here, we could look for other thermodynamic potentials (like enthalpy, free energy, etc.) of our gas.

It is rather a matter of time. Before such a "stressed" (low-entropy) state "unleashes" to a wide flat (high-entropy state), a lot of water may leak out. So it is not known where to enter correctly. This needs research, but the topic is extremely fertile.

One specific question so far - what is liquidity in relation to the topic in question ?


In general, I think the analogy between physical processes and market processes is fundamentally wrong.

After all, physical processes occur without the intervention of the human psyche, without the intervention of the human will.

On the other hand, one is longing to apply already developed and tested physical/mathematical apparatus and physical/mathematical methods

to the analysis of market processes. I think we have to wait until another Newton/Leibniz, who discovers/invents

Infinitesimal calculus as applied to market processes. Then we will have psycho/mathematics, psychophysics, etc.


Although, of course, there is evidence to suggest that human mental processes,

can influence physical processes. However, this is too subtle a matter for us to discuss.

 

Well, in terms of your "tension", the higher it is, the lower the liquidity, I guess? That said, your hypothesis in red could be that the market is seeking to restore liquidity.

2 sab1uk: And how did we used to live without the floating spread and that awful 4-digit?!

2 ssd: physical analogies are perfectly acceptable as long as the implications of them don't contradict reality. I absolutely don't care how much psychedelic is on the market, as long as this entropy analogy helps me get in and out more accurately. By the way, econophysics has been around for years, and yes "quantum" is blooming and smelling.

 

The demand for currencies is mainly based on the demand for other assets and what they are denominated in. Of course, all investments and speculation are cyclical - one gets out of them sooner or later. The only question is when? and it all comes down to the right timing. Therefore, overbought is not the same as overbought and we should not lump them together. One and the same overbought condition can be caused by absolutely different investment processes or speculations and consequently has a different time duration. It is necessary to identify the overbought market more precisely, for example where the money ultimately flows to, because currencies are only a means to an end. In other words, the dynamics of assets such as commodities, derivatives and stock indices must be taken into account. If the overbought USD against other currencies is caused by symmetrical growth of SP stock index, it is one thing, if the money went into bonds, it is another thing and you can't expect it to come back soon, rather the opposite - you can continue buying. This is an intermarket analysis. The abstract overbought currencies is not a sufficient and independent signal.

Therefore for the cluster indicator we may write an auxiliary one - the dynamics of the major assets. Or even combine them: dollar index + major related assets; another indicator - eur index and its assets, etc. Or for each currency an analytical output - buying/selling what assets for it mainly occurs and for how long. Here it is also important to choose a precise reference point and recalculate it, instead of just taking an arbitrary value and averaging over some period. It is important to accurately identify the beginning of the observed process of buying currency and synchronise it with the beginning of buying or selling an asset. Then there is also the possibility of clearly identifying the start of the reverse process, or at least the end of the buying process. Although it is quite possible that the beginning of the corresponding session in this or that market will be appropriate. Of course, all this should be tested on the history

 

 
Above is an example from my experience. I calculate the relative value of currencies. In this case, it is assumed that 1.1.1999 the value of currencies is taken as a unit. oC is Swiss, oK is Canadian. IMHO, you can even do tehanalysis on such charts.
 
Mathemat писал(а) >>

...

2 sab1uk: How did we ever live without the floating spread and that awful 4-digit spread?!

...

Alexei sab1uk is right in this case. I completely share his opinion.

Yes, our life was bad. It was even worse when spreads were 40 pips. And it was even worse when we had no access to the market.

Z.U. Sabluk Skype if you want to work in this direction, two heads and four hands is better. than one (mine) stupid and crooked hands.

theses.

Tics only. (Ask-Bid)/2. Multicurrency analysis. Oscillator. Noise filtering.

 
Prival >> :

Alexei sabluk is right in this case. I fully share his opinion.

Yes, life was bad. It was even worse when the spreads were 40 pips. And it was even worse when we had no access to the market.

Z.U. Sabluk Skype if you want to work in this direction, two heads and four hands is better. than one (mine) stupid and crooked hands.

theses.

Tics only. (Ask-Bid)/2. Multicurrency analysis. Oscillator. Noise filtering.

That's right!!! Just don't filter out the noise :-)

 
Zhunko писал(а) >>

That's right!!! Just don't filter out the noise :-)

I can clearly and mathematically prove what is noise. If so, then why not filter it ?

 

For those who have not lost interest in cluster indicators, I present

Semen Semenych's indicators programmed directly, as they are called, without any tricks that complicate understanding.


CL8v.mq4 - shows all lines for 8 currencies cluster: "EUR","GBP","AUD","NZD","CAD","CHF","JPY","USD"

CL1i.mq4 - it is launched for a definite instrument, it shows the excess of the instrument base currency value

over the quote currency of the instrument.

The instrument's base and quote currencies value is determined, as it should be in cluster

indicators, based on the demand for the base/quoted currency from all other seven (7) currencies in the cluster.

The CL1i.mq4 indicator is intended for trading on any of 28 instruments, on which it is running.

Both indicators are "light" in the sense that for calculations they use real quotes of 7 (seven) instruments only: "EURUSD", "EURUSD" and "EURUSD".

EURUSD", "GBPUSD", "AUDUSD", "NZDUSD", "USDCAD", "USDCHF", "USDJPY".

The values of other 21 (twenty-one) instruments are calculated using the dollar crosses.



Condition for opening a position upwards:

double Ind_0 = iCustom(Symbol(),0, "CL1i",0,0);
double Ind_1 = iCustom(Symbol(),0, "CL1i",0,1);

if (Ind_0 > 0 && Ind_1 < 0)


Condition for closing an open position upwards:

if (Ind_0 <= 0)

-----------------------------------------------------------


Condition for opening a position downwards:

double Ind_0 = iCustom(Symbol(),0, "CL1i",0,0);
double Ind_1 = iCustom(Symbol(),0, "CL1i",0,1);

if (Ind_0 < 0 && Ind_1 > 0)


Condition for closing an open down position:

if (Ind_0 >= 0)

----------------------------------------------------------

Next week I'll try to test this so-called "frontal approach".

Files:
cl8v.mq4  15 kb
cl1i_2.mq4  10 kb
 
ssd писал(а) >>

Both indicators are "light" in the sense that the calculations use real quotes of only seven (7) instruments: "EURUSD, "GBPUSD, "AUDUSD, "NZDUSD, "USDCAD, "USDCHF, "USDJPY.

symbols: "EURUSD", "GBPUSD", "AUDUSD", "NZDUSD", "USDCAD", "USDCHF", "USDJPY".

on the minute timeframes synthetic pairs will have divergences with the natural ones