On the unequal probability of a price move up or down - page 120

 
Vitaly Muzichenko:

This is the outermost bar on the chart, it is selected in the terminal settings, I do not even know how many, but the default is probably 5000

P.S. I lied, 20 000.

I don't care how many bars you have in your chart.

the very beginning of the overlap ?

 
Renat Akhtyamov:

Show an overlapping chart, the number of bars does not matter.

the very beginning of the overlap ?

Once again: The beginning is exactly the same as the middle and the end.

 
Vitaly Muzichenko:

Once again: The beginning is exactly the same as the middle and the end.

Show me a picture.

 
Renat Akhtyamov:

Show me a chart with an overlay, the number of bars does not interest me

the very beginning of the overlay ?

I when I started this, and it was 2015, I did overlays with a reference point, took 350 bars of history, i.e. a floating window, then realised it was self-defeating.

 
Renat Akhtyamov:

Show me the picture.

Here's the picture. Failed history for some reason, but that's okay


P.S. The bars in the story are 2152, Time M30.

 
Vitaly Muzichenko:

Here's the picture. Failure of history for some reason, but that's okay

ok

Thank you

 

While I was drinking a coffee and a bun, I got an idea how to decrease my position in the market with this technique

Forum on trading, automated trading systems and testing trading strategies

About Unequal Probability of Price Going Up or Down

khorosh, 2020.01.23 19:31

I finally stopped on one of my indicator variants for pair trading. It's been about half an hour since I entered it and I'm already making a profit.

13.00 is profit.


The idea is to open two opposite positions and note the equity/balance at that moment. If we see plus, we close and look for another signal. We wait until minus on the losing leg, so that our profit on closing the minus leg is more than on entering it.

Example: Profit on the euro pair +50, on the pound -72. As a result, we cut the EUR and wait till the correction of the pound is -40 and close it. As a result we have +10 from the trade.

If the minus grows - let's average.

In this case, we exit the market earlier than waiting till 2 pairs reach the total profit. The system is not new, but everything new is long forgotten old.)

 
Vitaly Muzichenko:

While having coffee and a bun, I got an idea out of nowhere of how to decrease positions in the market using this methodology


The idea is to open two opposite positions and calculate the equity/balance at that moment. If we see plus, we close it and look for another signal. We wait until minus on the losing leg, so that our profit on closing minus is more than on entering.

Example: Profit on the euro pair +50, on the pound -72. As a result, we cut the EUR and wait till the correction of the pound is -40 and close it. As a result we have +10 from the trade.

If the minus grows - let's average.

In this case, we exit the market earlier than waiting till 2 pairs reach the total profit. The system is not new, but everything new is long forgotten old.)

should work fine )) except if - Example: On the EUR pair a profit of +150, on the GBP -72, you can close the EUR and wait for the GBP to rise.
 
Vitaly Muzichenko:

Once again: The beginning is exactly the same as the middle and the end.

Rated 100. ++)

 

Grigori.S.B:

If you want to invent statistical arbitrage, correlation has nothing to do with it. Two highly correlated instruments can diverge for as long as you like. What matters here is cointegration, not correlation. Read for examplehere. And compare returns at high correlation and cointegration. Correlation and cointegration are not related and independent concepts.


khorosh:

Have you read the thread in full? I have already written there about it.

In fact, the true cointegration is possible only for the instruments the ranges of oscillations of which intersect, otherwise, like for example for the pound and the euro, the cointegration can be observed only for the transformed data that is a function of the initial quotes. In fact, for this purpose it is necessary to detrend. Whether the cointegration is stable or not, depends on the quality of detrending. For quality detrending, it must be dynamic. But of course, even high quality detrending does not guarantee the maintenance of cointegration, this is primarily a property of the tools. But, if the detrending is bad, the cointegration gets worse.