Forget random quotes - page 5

 
yosuf:
So, are you in favour of volumes or not? Can anyone show a graph of the relationship between price and volumes?

Well for example, there was a real drama with people, both in price and volume. :)

Yusuf, what dependence may there be? You need a clear formula, but it does not exist. Some people find formulas for certain situations (with a certain accuracy). But this is very far from what you do.

PS. Volume is important, but it is very difficult to work with it. And it has a higher volatility than the price itself.

 
Mathemat:
Where was this discussion, show me. I don't care about your sniggering and smiley faces.
Well, whatever means whatever.
 
I went to Selpo today. On the receipt they gave me the prediction: "NO EVENT EVER happens". Seems appropriate.
 
HideYourRichess:

Well, for example, there was real drama with people, both in price and volume. :)

Yusuf, what dependencies can there be? You need a clear formula - there is none. Some people find formulas for certain situations (with a certain accuracy). But this is very far from what you do.

PS. Volume is important, but it is very difficult to work with it. And it has higher volatility than the price itself.

1. First, for the sake of principle, you have to renounce time;

2) Are the volumes time-dependent or not?

3.How are they calculated and provided?

4.Are they updated with every tick?

5. Are the incoming volumes added up within a bar?

6. Do they come in chunks or as a whole?

 
MikeM:
I went to Selpo today. On the receipt they gave me the prediction: "NO EVENT EVER happens". Seems like a good point.
We are forced to look for patterns in accidents, pardon the pun, as in the case of gas laws.
 
HideYourRichess:

Yusuf, what kind of dependencies can there be? You need a clear formula - there isn't one. Some people find formulas for certain situations (with a certain accuracy). But this is very far from what you are doing.

PS. Volume is important, but it is very difficult to work with it. And it has higher volatility than the price itself.


Oh yes, looking for patterns on such a scale, and without taking into account the direction of trades in volumes... Good luck :)

Linear regression is enough.

 
yosuf:

1. First, for the sake of principle, you should renounce time;

2. Are the volumes time-dependent or not?

3.How are they calculated and provided?

4.Are they updated with every tick?

5. Are the incoming volumes added up within a bar?

6. Do they come in chunks or whole?

1. Yes of course, why should I disown time? My principle is simple: price does not depend on time, but it changes with time.

2. volume changes over time.

3. My calculation is very simple, it is "traded volume" - the number of concluded contracts to buy/sell, if you like.

4. Yes, it is updated with every tick.

5. Yes, aggregated within the selected time period.

6. er... each tick, if it's Time&Sale, is one trade, with price and volume (actually there are subtleties there, but they're interesting usually for HFT, but otherwise you can count as written). If it's bid/ask, then there are volumes at each level of the "cup" - same there a tick is a bid, with price and volume.

 
anonymous:


Oh yes, looking for patterns on such a scale, and without taking into account the direction of trades in volumes... Good luck :)

Linear regression is enough.

On what scale?
 
HideYourRichess:
On what scale?

A second or less.
 
anonymous:

A second or less.
If only the data were accurate, seconds wouldn't be a problem.