Absolute courses - page 4

 
Dear grell, there is only one option. It would be strange to expect even two, let alone the sea. Probably, due to your young age, you are unfamiliar with the basics of differential and integral calculus. You have tried to write down (rightly or wrongly) the same thing that Avals has already written down. I asked to write down something else entirely. The connections between d(ED), d(PD), d(EP), and dE, dP, dD, I hope nobody minds such designations (ED is EURUSD, and so on).
 
grell:

EURUSD(i)=(EUR(i+1)+dEUR)/(USD(i+1)+dUSD)

EURGBP(i)=(EUR(i+1)+dEUR)/(GBP(i+1)+dGBP)

GBPUSD(i)=(GBP(i+!)+dGBP)/(USD(i+1)+dUSD)

There are plenty of options, but I repeat, this is stomping around.


Take USD(i+1)=1, calculate from the exchange rate all the others, and then try to find the increments.
 
Dr.F.:
Dear grell, there is only one option. It would be strange to expect even two, let alone the sea. Probably, due to your young age, you are unfamiliar with the basics of differential and integral calculus. You have tried to write down (rightly or wrongly) the same thing that Avals has already written down. I asked to write down something else entirely. The relations between d(ED), d(PD), d(EP), and dE, dP, dD, I hope nobody minds such designations (ED is EURUSD, and so on).

No, there are plenty of variants, the question is which one you will abandon the idea on.
 
Dr.F.:
Dear grell, there is only one option. It would be strange to expect even two, let alone the sea. Probably, due to your young age, you are unfamiliar with the basics of differential and integral calculus. You have tried to write down (rightly or wrongly) the same thing that Avals has already written down. I asked to write down something else entirely. The relations between d(ED), d(PD), d(EP), and dE, dP, dD, I hope nobody minds such designations (ED is EURUSD, and so on).

Age! =Experience. And it pisses you off!
 
Dr.F.:
Colleague Avals, thank you for the attempt at reason in the midst of obscurantism. Unfortunately, I asked to write down a completely different ratio. you don't have "EURUSD increments" in yours, nor GBPUSD, or EURGBP increments, and therefore have nothing to relate to dEUR, dUSD, and dGBP increments as you labeled them.



EURUSD+dEURUSD

--------------------- = EURGBP+dEURGBP

GBPUSD+dGBPUSD

EURUSD+dEURUSD=EURUSD+GBPUSD*dEURGBP+EURGBP*dGBPUSD+dEURGBP*dGBPUSD

GBPUSD*dEURGBP + EURGBP*dGBPUSD + dEURGBP*dGBPUSD - dEURUSD=0

 
grell:

Age! =Experience. And it pisses you off!

That's what I'm saying.... Dr.F. You are doing nothing. Believe in the experience of many people who at one time began, as you do, to master the basics of time series analysis and asked the same questions. Over the course of dozens of years. Thousands of people. And the answer is always the same.
 
Joperniiteatr:
6Aftar may be referring to this, the arithmetic is... without diffs. Above shows the calculation through absolute values. And you can work with delta increments or with oscetal values, where the difference is replaced by a division minus one.

It is more correct to calculate not in increments, but in coefficients, because the yield and the volatility depend on the current exchange rate. More details why this is the case, for example in derivation of the Black-Scholes formula

In general,

EURUSD*dEURUSD

--------------------- = EURGBP*dEURGBP

GBPUSD*dGBPUSD

dEURUSD/dGBPUSD=dEURGBP

 
Joperniiteatr:
6Maybe the author means this, the arithmetic is... without diffs. Heh-heh-heh. Above showed the calculation via absolute values. And you can work with delta increments or with oscillator values, where the difference is replaced by a division minus one.

All this is useless, the author himself understands, that until an absolute scale is set, no results will be obtained here. But there is no such scale, simply because there are no eternal, unchangeable values which can be taken as a yardstick. Gold used to be a surrogate, but then it was abandoned, for better or for worse.
 

Avals, think of it this way: ED(E,D)=E/D. d(ED) =[d(ED)/d(E)]i*dE+ [d(ED)/d(D)]i*dE, where[d(ED)/d(E)]i is the derivative of (E/D) at a particular bar i. Take the derivative. By the rules of partial differentiation.

 

The point is this. Here dE with index i means E[i+1]-E[i], where indices i and i+1 mean bar numbers. Please write down similar ratios for the other two "sides of the triangle" and go from absolute to relative increments.