Market theory - page 208

 

Primitivism cannot explain the fact that, the indicator allows to achieve stat advantage over the market for 15 years on EUR/Dollar, with a fixed lot 0.1 from 01 01 2000 to present time on TF D1 at SL=TP=1100 pps.

Yes, you can. With so many limitations (1. EUR/Dollar, 2. D1, 3. 15 years, 4. TP=1100) you can pick anything. Let's try to show us at least 10 more pairs with the same parameters (say, with fixed lot 0,1 from 01 01 2000 up to the present time on TF D1 at SL=TP=1100*K points, where K - a reasonable coefficient, so that TP and SL would be bound to the relative values of the pair volatility).

 
Yousufkhodja Sultonov:
You were happy to expose me. No one is in a position to expose me as I'm sure I'm trying to convey the true patterns of the market. As long as you don't understand it, that's another matter. In time, everyone will start to understand and develop my thoughts. In the meantime, I welcome constructive criticism. Do not speak in riddles, speak openly, criticize, I will not be offended. Ask me if I don't understand things - I will explain.

Dear Yusuf, I will prepare a file in a moment. Similar to the one that mt gives out when exporting quotes. I.e. several columns <type DTYYYYMMDD>,<type TIME>,<OPEN>,<HIGH>,<LOW>,<CLOSE>,<type VOL>. I am asking you to show what the indicators will draw, if these data are fed to them. There are 5000 bars here. Up to 4500 the level is 1.2, then 1.3.

Files:
EURUSD5model.txt  284 kb
 
Mikhael Isakov:

Dear Yusuf, I will prepare a file in a moment. Similar to the one that mt gives out when exporting quotes. I.e. several columns <type DTYYYYMMDD>,<type TIME>,<OPEN>,<HIGH>,<LOW>,<CLOSE>,<type VOL>. I ask you to show what the indicators will draw if these data are fed to them.

It is better to show on any pair of any section of history on any TF of the terminal MT4. Let's feed the terminal data directly. What is the problem? What do you need your data for? Moreover, they are not clear to me at all.
 
Yousufkhodja Sultonov:
Better show on any pair any section of history on any TF of the MT4 terminal. We'll feed the terminal data directly. What is the problem? Why do you need your data?
Because I'm telling you as a physicist: to understand the essence of the phenomena you need to go to the limiting situations. I have attached the file. I ask you for a response of your algorithm to it.
 
Yousufkhodja Sultonov:

Look at how on the M1 TF, the normally calm, virtual market prices P (Bears) (red line) until the last point showed, 3 hours before the news release, where the current price of CD could fall, and after the news release did not waver. This means that the current price should return, which it did. At the time of the news release, the Bears were handed the price and then it was handed to the Bulls, who carried the price upwards:


How I understand you, Yusuf. I used to draw similar curves myself for several years. And even called them virtual prices. And had all these hopes of building the perfect filter. And I could draw better filters than yours. One nuance: they all lag. Naturally. That's what I want to show you - if you are sincerely mistaken - on the example of your filter.
 
Mikhael Isakov:
Then, as a physicist I say to you: to understand the essence of phenomena you have to go to the limiting situations. I have attached a file. I ask for your algorithm's response to it.
The algorithm is dynamic, so it will not show anything on static data, or rather it will show complete calm. I will not even try it.
 
Yousufkhodja Sultonov:

Now, tell me, Dear, what other indicators are capable of achieving a stat advantage over the market at TP=SL.

No problem. I myself can show you a manual trade that will have an "advantage over the market". In the sense that the deposit will grow. I can even offer you a duel for the public's amusement. A dime account for 10 to 50 quid, me and you. I'll do some trading, and so will you. On your theory. And we'll see the results. Without robots, so you can't say they're dumb and the theory is true.
 
Yousufkhodja Sultonov:
I'm not even going to try.

This is highly suspicious. You promised to accept constructive suggestions. For Plato is my friend, but Truth is dearer.

And yes, the notion of "advantage over the market" is not related to the idea of TA=SL. The condition TP=SL is needed only for ease of assimilation by the public of the idea of an advantage. But otherwise, if the algorithm has an advantage, it is only important that the ratio of TP to SL should be constant. It does not have to be equal to one. You may choose TP=0.1 SL. In this case the SL will be triggered (without taking into account the spread) 10 times less frequently, but it will also take 10 times more money :) However, if the algorithm makes sense, the deposit will grow. And if anything, in order to avoid some random fluctuations (which can take out the SL, while you really "guessed" the market correctly, but it turns out that it was more important HOW the price moved, not WHERE it went: namely, by slightly twitching backwards in advance), it's better to make the SL a couple or three times larger than the TP.

 
Mikhael Isakov:
How I understand you, Yusuf. I have been drawing similar curves myself for several years. And even called them virtual prices. And I had all these hopes of building a perfect filter. And I could draw better filters than yours. One nuance: they all lag. Naturally. That is what I want to show you - if you are sincerely deluded - on example of your filter.
Look at the screenshot above, 3 hours before the event the market prices showed a possible turmoil in the market, which subsequently happened. And you say lagged. We are speaking in different languages. I am not talking about any filter. Re-read the thread and see how many times I have shown how market prices anticipate future events with their sudden movements.
 
Yousufkhodja Sultonov:
Look at the screenshot above, 3 hours before the events the market prices showed possible unrest in the market, which subsequently happened. And you say they are lagging.
Show me the previous part of the picture. Maybe, there was an event 12 hours or a day before it happened - and you saw it. For you are late :-). Maybe even a month before. I do not know how much data from the past history you have taken into account. But you are late by half of that amount.
Reason: