If I operate two pairs that are correlated to 100%, I'm actually doubling the risk. So, if I want to operate both, I have to split by half the size of each operation.
If I operate two pairs that are correlated to 100%, I'm actually doubling the risk. So, if I want to operate both, I have to split by half the size of each operation.
The probability is always 50 and 50%, or you win or lose, either with one or more pairs at the same time.
Divide by 3 to balance the risk. You still can't know if the pairs will be correlated in the future, those data are from the past. You can assume that there is correlation as long as the fundamental base doesn't change.
What exactly do you mean, bet size, you mean the risk percentage?
I choose EUR/USD, GBP/USD and USD/JPY. I take the returns in pips of the last 50 periods in H4. I make a scatter chart of each one against the others. It is clear that if a signal appears in Cable (GBP/USD)
Finally I find someone talking about this in the forum! This is the level of CFA financial analysis. It has to do with portfolio management, measuring and comparing peer returns over an X period. So you will know the risk, correlation and coefficient. So you can adjust the risk and lot size according to the expected return, using the efficient border. I read the theory, but I still don't know how to apply it. Take a look at chapters 7 through 9, the rest isn't much good for the moment.
So, if you follow the trend, you go short in EUR, GBP and AUD, and long in JPY, CHF and CAD. Six operations, but it's actually a giant position. So wouldn't it be better to operate just a couple?
Depending on whether Risk On or Risk Off mode is activated, you end up with a huge exposure, whether long or short, in USD. And that�s just what I�m trying to avoid. To get rid of that problem, I�m looking to operate a crossover. The strongest currency against the weakest. Then I remove those two currencies and repeat the process for another pair, taking into account the current correlation in the short term (actually it�s co-integration).
So your system wouldn't work if you only operate EUR/USD and EUR/GBP, which are almost uncorrelated?