Not the Grail, just a regular one - Bablokos!!! - page 61

 
Avals:

It cannot be 40 pips. Apparently on spread spread spreading counted without taking into account the real bid/ask

Yes, no spread, but I can post the balance. Just don't want to feed the non-stationarity specialists.

It's not clear how to trade.

 
faa1947:
Co-integration, at least one. After all, you claimed above that there isn't one. So prove that there isn't. You've done the math here and posted it. Come on, match it.

You haven't posted anything there. The whole model is laid out, not three obscure graphs. "There are subtleties."What kind of subtleties? What kind of rows are being studied?
 
Demi:

I thought it was based on the Spearman correlation coefficient.)

No, based on the "ideal synthetic tool" formula. Or as the author explained "to make the new graph the 'narrowest' one" :)
 
Demi:

you haven't put anything out there. The whole model is laid out, not three obscure graphs. "There are subtleties."What kind of subtleties? What kind of series are being studied?
There are graphs of series, residual from co-integration regression, proof of stationarity - the classics. Well the subtleties that follow make money.
 
faa1947:

Yes, no spread, but I can post the balance. Just don't want to feed the non-stationarity specialists.

It's not clear how to trade.


no, without a real spread the balance is useless
 
Avals:

No, based on the formula "ideal synthetic tool". Or as the author explained "so that the new graph turns out to be the 'narrowest'" :)

Things are much worse there.

I showed him that he had no correlation, and he replied, so what, I call it what I want - the man does not understand that you cannot apply established terms to your inventions.

 
faa1947:
There series graphs, residuals from co-integration regression, proof of stationarity - the classics. Well, the subtleties that follow bring in the money.

What is "cointegration regression"? "Series graphs" - are these series cointegrated? Please post a graph of these series, just standardise them.
 
Avals:

No, without a real spread the balance is useless.

It's not clear at all. Maybe the TA is suffocating, though.

You enter on the deviation and see that it is against the trend, and then it still turns out to be in the plus! What kind of nerves you need to have to trade.

 
"co-integration regression" - regression co-integrates with what?
 
faa1947:

Things are much worse there.

I showed him that he had no correlation, and he said, so what, I call it what I want - the man doesn't understand that you can't apply established terms to your inventions.


Well I think what his script finds will pass the cointegration test. Only on the same sample. I.e., the cointegration fit
Reason: