From theory to practice - page 395
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Why does it go to infinity?
With ABS(return)=const=0.0001 sigma = (N*0.0001)/ Root(N) = 0.0001* Root(N)
sigma = (SUM(ABS(return)/CORN(N).
At ABS(return)=const=0.0001 sigma = (N*0.0001)/SUM(N) = 0.0001*SUM(N)
Ahhhh... Well, right. But, in a certain time window (at finite T) it will be some kind of constant. And we are working exactly at finite T.
Incoherent babble.
There is a formula which shows that the RMS of random walk increases in proportion to the root of time.
You are trying to invent a variation of this formula.
But it has nothing to do with the window (sample), it has a constantly growing "window" from point t=0, at which the process value also = 0.
And secondly, this formula is of course incorrect for the market, as it is not a long-run SB.
And the RMS in the window is calculated by the standard method, as the root of the mean squared deviation from the SMA (or whatever you have instead of the SMA).
Tested it on other pairs. Everything is fine. This month let's have a look at equity and....
Get your pockets ready Vladimir!!!
Ahhhh... Well, right. But, in a certain time window (at finite T) it will be some kind of constant. And we are working exactly at finite T.
For finite T such a sigma at 1 tick per second and 4-bit quoting will depend on T in this way:
Does such a sigma make sense, why define it at all? Or the question about the meaning is superfluous ?
For finite T such a sigma at 1 tick per second and 4 digit quoting will depend on T in this way:
Does such a sigma make sense, why define it at all? Or is the question of meaning superfluous ?
We do not define it. On the contrary, we reasonably refuse and use it:
The standard deviation of the price from the mean in the sliding window = 4 hours is calculated using the formula:
sigma = Root((SUM(ABS(return))/T)*(SUM(ABS(return))/N)*14400)
where T is the current system run time since the beginning of the trading week.
Well, that's me working in the window = 4 hours.
Of course, everyone is free to choose a window that suits his/her trading style.
Of course, I would still try window = 24 hours (bas is right here - it is where the tick Poisson flow sits)
I recall Doc's answer to the question, would the average tick increment rates for a year, for example, be the same?
The number of ticks per year is also different for each symbol and year. It's all too unstable to expect any consonant in the end. But I counted speed as(absolute value of all ticks returnees in a row)/(all time). If as you wanted to first build a second timeframe, then it should give a constant number of bars for the same time intervals. Whether the speed will be constant in this case - I do not know, it should be checked.
I liked the advice here in the thread not to look at the time. The ticks are not random values; the returnee of the next tick may also be predicted based on returnees of previous ticks. In Forex it's a stupid idea, unprofitable because of the spread. But an interesting conclusion is that our time (of an observer) is non-linear relative to the Forex time. From the point of view of Forex - for it every tick comes in a constant time interval. And for us - the time of forex accelerates at lunchtime and slows down at night (judging by your graphs).
p.s. It's still just an interesting logical conclusion, and probably not accurate. But I couldn't find an argument against it.
It should not be forgotten that changing the observation rate of a physical process does not in any way change the process itself. It seems to be clear to the hedgehog, though not to everyone.)
In this case we have a strong overloading, up to useful signal degeneration in noise, and loss of information about the initial process, so cheating with reading of signal values at convenient times is an obvious self-deception and fantasy, which has no relation to reality, not to mention the illiteracy of such an approach from the scientific point of view.
But, as the classics say, whatever a child needs, he can't cry. ))
Does such a sigma make sense, why define it at all? Or is the question of meaning superfluous ?
Moreover, it is not a sigma relative to the SMA, but to the process value at the starting point of the window) not sure that is what the author of the thread is looking for)))
As a continuation of the previous answer - there is no second timeframe in mt5, it's difficult to do it yourself. There is a minute timeframe. I will use it for rough estimation.
I will use it for rough estimation. Atacha script shows average price speed per m1 bars.
eurusd 2015: 370886 bars, price for them passed 50.74050 pips in absolute value. velocity = 0.000136811 pips/minute
eurusd 2016: 373151, 39.02563, 0.000104581
eurusd 2017 year: 370355, 32.62879, 0.000088101
It has no constant. I suppose if you create an s1 timeframe, there's no constant there either.