![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
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
It doesn't work like that... I read this forum solely because of your, Matemat, Neutron and Prival's posts (extremely impressed by Faa's and his views, leaving aside some questions, and Reshetov), will you go away-I have one less interesting person to read? No way... Let's develop the topic once we've started. That will not do. Read then the forum hundreds, no need to deprive them of good, high-quality, interesting thoughts and disputes. So you can't go anywhere from here.) No...
So, now there will be no one to fight with on the forum. :)
By making some assumptions about the "physics" of the quoting process with the help of simulation modelling, I tried to obtain a resulting quotation that was close to the real one in terms of its properties. "Reality" was defined by the following characteristics:
Actually, I modeled trade of more than 100 banks and for each bank there is its own behavior in fulfilling "obligations" (conventionally speaking). The feedback (i.e. decision making based on already obtained quotient) and speculation have not been done yet, but I already obtained interesting results (probably they are mine, and just for that reason - interesting :o). For simplicity of modelling, the bank expresses the interests of many, i.e. I model one bank, rather than a bunch of unique participants expressing their interests through the bank.
I did something similar. However, the aim was more modest - to get clearer estimates for "how it could be" processes, if the market process corresponded to the one being modelled. And yes, the result is something interesting, although it cannot be used directly.
A madman alone..." // Shakespeare
HideYourRichess 11.02.2012 09:43
By making some assumptions about the "physics" of the quoting process with the help of simulation modelling, I tried to obtain a resulting quotation that was close to the real one in terms of its properties. "Reality" was defined by the following characteristics:
Actually, I modeled trade of more than 100 banks and for each bank there is its own behavior in fulfilling "obligations" (conventionally speaking). The feedback (i.e. decision making based on already obtained quotient) and speculation have not been done yet, but I already obtained interesting results (probably they are mine, and just for that reason - interesting :o). For simplicity of modelling, the bank expresses the interests of many, i.e. I model one bank and not a bunch of unique participants who express their interests through the bank.
As I see it, the problem is divided into two parts: the TS itself (1), which we will consider, for the time being, a black box, and (2) the input on which this trading system is developed and tested.
(2) It makes no sense to develop and test the TS on real data (a horrible trespass for this forum). We need an artificial input signal (kotir) that has known properties, such as different types of noise, trend, rebound and maybe what is mentioned above. What we feed to the TS input depends on the TS itself - what we are trying to check in it and it may be more important.
(1) At the TS one should check its behaviour for:
a shift in the deterministic parameters' coefficient
shift of coefficients in stochastic parameters
error of specification of a deterministic parameter
specification error of stochastic parameters
error of coefficient of deterministic parameters
error of the coefficient of stochastic parameter
the nature of variation of the error variance
the nature of the error accumulation along the forecast horizon.
That's enough.
We then test the artificial input and black box (TS) for adequacy (scaling) to the real quote. If we have proved that all or part of what is called here is generated by the real quotient and not invented by us, or we can prove that something is not available or can be neglected, then we have a CU with a capital letter.
The way I see it, the problem is divided into ...
My perceptions are somewhat different. I think state identification is much more important than forward tests and stuff.
"...Please leave
A madman alone..." // Shakespeare
Tyndi rindi ryushka
Piggy is sick.
Slap! mustard on her back.
Once and cured Piglet // A.Kozhevnikov
Then I didn't see any difference between our perceptions. You would also agree if you read my post carefully
faa1947 is slowly but surely recruiting fighters with market non-stationarity. It would seem that my branch, which has nothing to do with the issue of stationarity, has also become a springboard for recruitment:)
Tyndi rindi ryushka
Piggy is sick.
Slap! mustard on her back.
One and cured Piglet // A.Kozhevnikov
You can't do without a mustard pack, you need an alcohol compress) // Malakhov+
faa1947 is slowly but surely recruiting fighters with market non-stationarity. It would seem that my branch, which has nothing to do with the issue of stationarity, has also become a recruiting ground:)
Not me. Convincing the market.
And your branch, like no other, is connected to nonstationarity as the last and most important point of TC: either we are stationary (stable) or we all lose when nonstationarity becomes fatal. And we resent "Forward test was so beautiful!!!! and depo is empty".