I would like to share the link - page 7

 
tara:


Of course, not enough. I don't have any of that in me, I don't sweat much, I weigh 64 kg (already 30 years old), I started using 0.5 glasses when reading a year ago. Active heterosexual.

What exactly is missing?

French

;)))

 

Даже не пытайтесь понять и расслабьтесь. Про успех фонда Renaissance, в который берут только физ-мат технарей со званиями (и никаких экономистов!) Вы, наверно, не слышали.

I'm not. But I see you are. Why do you ask? The Renaissance Fund is successful, and indeed the maths works there. But you consider the fact that in order to implement some things, especially on intraday noise, the funds are pouring in substantial money. And they obviously are not trying to solve serious problems so that non-stationarity could be quickly removed and the spread of equalizing coefficients could fit in +-1SD. You should look at things more realistically, earn money, look for something new, and not do all that rubbish stomping on the spot.

French ;)))

And you, I see, lack a sense of humour? I make my own jokes, don't I?

Of course I don't have enough. There's nothing wrong with me, I don't sweat much, I weigh 64 kg (30 years), I started wearing 0.5 glasses when reading a year ago. Active heterosexual. And what exactly is missing?

I would not claim, because it is a personal matter. The market seems to be lacking a stable income, if that's your approach to the matter. Well, sports or what;)

After comrade avtomat it's even nasty to write about serious things.

 
Yuupi_Como:

to realise some things, especially on intraday noise, funds are pouring in substantial money

Did you come up with this on your own? A dozen Mbit/s is enough to work intraday and successfully catch what you need, believe me.
 
Yuupi_Como:

... ...they just won't settle down. This is some kind of incessant mathematical masturbation, pardon my French.

Yuupi_Como:

After comrade avtomat it's even disgusting to write about serious things.

Apparently you have too delicate a soul... if the mention of your French puts you in a state of nastiness that prevents you from writing about serious things...

But put aside your hypertrophied nastiness! (By the way, something similar has been flashed here recently)

Teach those nasty math junkies some serious stuff!

 

Это вы сами придумали? Чтоб работать интрадей и успешно ловить то что надо, достаточно десятка Мбит/с, уж поверьте.

So go ahead and catch it, who's stopping you. Apparently we're talking about different scales.

Apparently you're too delicate of a soul... if the mention of your French puts you in a state of nastiness that prevents you from writing about things serious...
But put aside your hypertrophied nastiness! (By the way, something similar has been flashed here recently)
Teach those nasty math junkies some serious stuff!

I don't give a fuck (fuck if you like) about anyone's learning, including yours)))) I just don't like trolling humour. It's just a bit simple.

 
Yuupi_Como:

So go ahead and catch it, who's stopping you. Apparently we're talking about different scales.

I don't give a fuck (fuck if you like) about anyone's learning, including yours)))) I just don't like trolling humour. It's just a bit simple.

You speak French like a Frenchman, Buonaparte, no less!
 
LeoV:

You see, there is no experimental evidence (only in your words) as to how this unit root test is related to the non-stationarity of financial markets. Directly or indirectly, squared or rooted....)))) You have decided that related.

Not me and not just me, but a great many people.

Looking at your result, I can say with almost 100% certainty that what you got on history, you will not be able to predict in the future. The whole trick of the non-stationarity of financial markets is that by taking a piece of history you can always find an algorithm that can make the maximum (or not the maximum) amount of money on that history. Thus, the impression of a certain stationarity of the market is created. But this myth is quickly dispelled, because it is not guaranteed that the algorithm that you have found on the history will work on future, unknown data.

For some reason a lot of people don't pay attention to my posts and attribute the exact opposite to me. What's the matter, I can't figure out.

Essentially. I assert:

1. Simply testing on a sample as well as testing out of sample (the forward test) does practically nothing.

2. A judgement about the future (probabilistic) can only be made to the extent that non-stationarity of the market has been taken into account in the samples mentioned. If our TS does not consider non-stationarity at all - then the future is not much different from a casino
 

Curious article about fat tails.

see attachment

 

Interesting article for proponents of the link between price increases and trading volumes.

The article very significantly clarifies that this relationship should not be sought in the form of correlation, but in the form of Granger causality.

Files:
 
Found a website with lots of books on time series available for download. A search for Time series yielded over 10 sheets. Lots of totally fresh books. Had a look at some - quite curious. Would have seen 10 or even 5 years ago. I copied the beginning of the list and put it in an attachment.
Files:
ref.zip  21 kb