Which mash-ups are the best to cross? - page 3
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
They don't have to keep up with it, the waving must have a purpose. Depending on the period it may serve as a support - resistance line, as well as a direction and strength indicator. It is not about the sleeves, but about how to prepare them.
I don't question the value of support and resistance lines, and I use them as a filter.
But I do not use them "directly", I just recycle them as I see fit.
The question of the topicstarter is the direct use of pens for prices.
You can still drive a steam locomotive or a Zhiguli today, but it is hardly efficient or comfortable.
I'm not going to argue. And will you share the link?
I will. Catch!
Sergei, what lag?! Nothing's lagging. MA is a filter. If you highlight a low frequency, it will have the appearance of lagging, but that doesn't mean it's lagging. It just doesn't have the high frequencies in its spectrum. That's why it's low frequency. That's its essence.
There is a contradiction in the phrase "MA lags".
There would be a contradiction if there was no lag!
In this case, any extrapolation of smooth MA one step ahead will allow predicting the future, including for random VR, and the latter is impossible due to violation of the law of causality. Thus, lagging in any MA is inevitable (unless, of course, we believe in the absence of Magic).A lag is useful in a way. If there is no lag, there will be too many false signals to enter the market.
The devil is in the details. "To a certain extent" and "too many false signals" are intuitive and imprecise definitions.
To rely on them as a foundation for building a mechanical autonomous trading system is to take excessive risks.
Do you trade with your hands?
If you have ever designed and commissioned a critical product, you will understand that this approach does not work.
The requirements there are very stringent for failures and "floating" parameters.
comrades
someone post the meme for the fifth and the expert based on it
...Thus,lagging in any MA is inevitable (unless of course we believe in the absence of Magic).
Mathematically it is absolutely correct, but in terms of trading it is possible to build a tick diagram inside a minute bar, it will also have a lag, but there is a question of what kind of lag is acceptable. It all comes down to the TS.
There is a great deal of slyness in your statement.
To be correct, if you work with a particular BP, you need to make a forecast one step ahead. Making a forecast for any number of steps makes no sense, since forecast accuracy decreases exponentially as the forecast horizon increases (in context - as the number of forecast counts increases). If you need to make a forecast for a longer interval, you should (from a mathematical point of view) switch to BP with step between samples equal to the required interval and make a one-sample forecast one day ahead.
So sitting on ticks and predicting minutes is not optimal and therefore there is no contradiction between what I said about inevitable lag and your example about ticks.
Neutron:
.... forecast accuracy falls exponentially as the forecast horizon increases ...
Why expotential? It seems to be proportional to the square of the drop.
See.
Let one step forward prediction accuracy q=0.1, then, for i=2 Q=0.1*0.1=q^2.... i=n Q=q^n
We have an exponential function with base q of the form Q(x)=q^x. Let's move to another basis. ln(Q)=x*ln(q) => Q(x)=exp(x*ln(q)) as required! We have an exponential dependence on a negative argument (ln(0.1)<0).
There is a fair amount of slyness in your statement.
To be correct, when working with a particular BP, the forecast should be made one step ahead. It makes no sense to make a forecast for an arbitrary number of steps, because forecast accuracy decreases exponentially with increasing forecast horizon (in context - with increasing number of forecast counts). If you want to have a longer forecast interval, then mathematically we should go to BP with step between samples equal to the needed interval and make one-sample forward forecast.
So sitting on ticks and predicting minutes is not optimal and therefore there is no contradiction between what I said about inevitable lag and your example about ticks.