From theory to practice - page 1701
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
What are you going to do with it...
1. you have a flat - a section with a conditionally constant MO. A trend has started - the series rises or falls, leaving the upper or lower limit of the horizontal channel. Does the IR change or remains constant? Calculate the IR in the sliding window and select the optimal size for the indicator sensitivity change.
2. you, "physicist", I wrote to you two years ago - there is no indicator of a breakdown. And there can't be. And if there were - all your research would be unnecessary, because you could generate profits with the simplest TA models.
3. I forgot to ask you.
1. I know all that without you.
2. There is such an indicator and I use it, but sometimes it also "misses" trends that my TS doesn't need. I would be much worse without it, but I always want to achieve perfection, therefore I ask those who suffered for it to work with Hurst and autocorrelation.
3. to see the difference between you and him, just read his post like this:
Forum on trading, automated trading systems and testing trading strategies
From Theory to Practice
Vladimir, 2018.03.03 15:40
Maybe the point is that you're looking for one "bell"? One for every time of day and day of the week. Take a look:
https://www.mql5.com/ru/forum/221552/page19#comment_6168925 activity changes during the day
https://www.mql5.com/ru/forum/193378/page49#comment_5239746 for Mondays, Tuesdays, ...Fridays
Why not make this "bell" a function of two more variables, the day and hour numbers, since you're not looking for a scalar (a number) anyway, but a distribution? Or parametrize the bell you are looking for, so that the parameters become functions (at least tabulated) of the day and hour numbers. After all, as I understand it, you don't need all of it, a description of behaviour around "heavy tails" is enough...
I have just actually solved this problem and it has yielded a qualitative improvement in my TS.
Have you ever even thought about it, set yourself or anyone else such a task? Yeah, you just can't understand the depth of it, let alone articulate it.
Okay, there's nothing to talk about. Your best friend QuantumBob is taking a break, so talk to him.
1. I know all this without you.
"The question was specific - how do you determine the trend with MO and variance? The answer is, you can't." (с). Gave him the answer."I know all that without you." (с).
Curtain.
If you knew, hillbilly, who you were talking to in that tone of voice, you'd have yourself banned forever (if you had a conscience, of course).
It's people like you that make me disgusted to appear on this forum.
First time I completely agree with you)
watch out, hot Finnish guys...
-----
there are trends in the market, and they are detected and traded by standard deviations.
Everyone is massively confused by the simultaneous presence in the market of several trends on different horizons, and the cycles-trends relationship, when one wants something monotonous. But miracles do not happen :-)
-----
Catching a 3-day bank trend/cycle on a one-day diurnal (omitting a lot of details):
on H1 ( or M30 if you want better sampling) we mark minimal last knee of the zigzag bigger than a confidence interval (visually or instrumentally, whatever)
from the far top to the current moment, but not more than 24 bars from the near one, draw a channel of the standard deviation.
When the price crosses the border of the channel, open your blog with economics/politics news for the last 48 hours, and read the foundation. You have 10 minutes to make a decision.
-----
no strategy works without a highlighted bold
What to continue, O Vladimir?
Give you a first-year university textbook?
Or reveal the hidden secret of information under the codename "Google trend in mathematical statistics"?
P.S. Approximate the time series with a linear function, take the slope of the straight line, introduce the trend-flat evaluation criteria and the golden key is in your pocket.
P.S. Pulling mathematical statistics on the ramblings of a 19th-century journalist without a mathematical education - I'll pass.
In this post there is already an answer. Approximation is a method of approximation theory (approximation) of functions, and in terms of MO and variance the slope angle can not be expressed. That's what I was talking about. You have to refer to the original data ("time series") and work with methods other than probability theory and matstatistics. The slope angle is also not from there, from geometry or matanalysis. Incidentally, the notion that characterises rate jumps (continuity modulus) was the key to Jackson's first theorem precisely in the theory of approximation of functions.
About Charles Dow separately. He is the founder of The Wall Street Journal which became one of the most respected financial publications in the world. Inventor of the Dow Jones index.
Wall Street and "bullshit" - yes... That's a big swing, Dimitri. The magazine has continued to be published for a century and a half now.
Actually, people here are interested in the very same nonsense that this very Wall Street was already doing in the 19th century. And you? Is it really Hinchin's theorem?
Approximation is a method of the theory of approximation (approximation) of functions, and you can't express the slope angle in terms of MO and variance. That's what I was talking about. You have to refer to the original data ("time series") and work with methods other than probability theory and matstatistics. The slope angle is also not from there, from geometry or matanalysis. By the way, the notion characterizing rate jumps (continuity modulus) was the key to Jackson's first theorem precisely in the theory of approximation of functions.
I myself have already understood that I wrote too complicated - one has to think. At the top of this page it is easier to write.
Go on, then. Let me remind you that the question is how to define a trend in terms of properties of probability distributions. Since you've already chosen these terms.
- show me how in terms of MO and variance varying with time you can define what a trend is. Show it. Give that definition. So that it is clear that it gives the same classification as known. For example, like Charles Dow: in case of ascending trend, the next peak on the graph must be higher than the previous ones, in case of descending trend, the next declines on the graph must be lower than the previous ones.
Express the same in terms of MO and variance changing with time.
What does the Dow have to do with anything?
I don't get it at all.
https://en.wikipedia.org/wiki/Stochastic_drift
1. I know all this without you.
2. There is such an indicator and I use it, but sometimes it "misses" trends that my TS does not need. I would be much worse without it, but I always want to achieve perfection, therefore I ask those who suffered for it to work with Hurst and autocorrelation.
3. To grasp the difference between you and him, just read his post like this:
This is a problem I have only just actually solved and it has produced a qualitative improvement in my TS.
Have you ever even thought about it, set yourself or anyone such a task? Yeah, you just can't understand the depth of it, let alone articulate it.
Okay, there's nothing to talk about. Your best friend QuantumBob is taking a break, so talk to him.
Sash, I get the feeling from your conversation that you're looking for a difference in the views of physicists, not in market quotes.
Physicists have the same relation to markets as plumbers have to musical instruments.
Sash, I get the feeling from your communication that you are looking for a difference in the views of physicists, not in the market quotes.
Physicists have as much to do with markets as plumbers have to do with musical instruments.
Now it does not matter to me whether a person is a physicist or not. Vladimir can hardly be called a physicist. But whoever has seen his CV, probably immediately sat down on his versatile knowledge and interests (if he wants - he will introduce himself).
In fact, I not once or twice re-read this thread in search of the Truth. And have come to a conclusion, that there is nothing here to read, except Vladimir, Koldun, Asaulenko and, partly, Bas.
Alas, Vova (you may do what you like, but Uladzimir is a name difficult to pronounce, Vova is much clearer) you are not included in this list, from my point of view. Some graphs, some channels, some turbidity and ripples... Pardon me... Where are the ideas? Where are the soul-stirring studies and stats? None.
And yet the Grail is closer than ever. We need a point in this epic. So you could talk about the golden ratio or some other music for the market. It will be rewarded.