Econometrics: one step ahead forecast - page 20
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
Unfortunately, not so much.
woot.... That's a good place to start. And not to make some unknown prescott, with no answer to the question why and why...
I'll make a clear explanatory example, from which the meaning of this model will be clear. But it will take some time to create such a tutorial - I think, if nothing prevents me, I'll manage it by the weekend. And I'll post it in my thread as a visual aid.
woot.... That's a good place to start. And not to drive some unknown prescott without answering the question why and why...
I know exactly why and have explained it to you many times.
I will make a clear explanatory example, from which the meaning of this model will be clear. But it will take me some time to create such a manual - I think, if nothing prevents me, I will manage it by the weekend. And I'll post it in my thread as a visual aid.
If you don't mind, let me know when you put it up.
Yes, look through Yandex - query "A trader called Harry and his approach to the market" - first link click "copy". Will be from the yandex cache as the spider is now hanging
Or here http://hghltd.yandex.net/yandbtm?fmode=inject&url=http%3A%2F%2Fforex.kbpauk.ru%2Fprintthread.php%2FCat%2F0%2FBoard%2Ftrading%2Fmain%2F39461%2Ftype%2Fthread&text=%D0%93%D0%B0%D1%80%D1%80%D0%B8%20%D1%82%D1%80%D0%B5%D0%B9%D0%B4%D0%B5%D1%80&l10n=ru&mime=html&cht=1&sign=37c0d2a96b68df40eb5368b916539921&keyno=0
Logged in, but the pages themselves don't open.
Maybe you can retell it?
woot.... That's a good place to start. And not to drive in some unknown prescott without answering the question why and why...
I know exactly why and have explained it to you many times.
I will make a clear explanatory example, from which the meaning of this model will be clear. But it will take me some time to create such a manual - I think, if nothing prevents me, I will manage it by the weekend. And I'll post it in my thread, as a visual aid.
If you don't mind, let me know when you have it.
Give me a description of the prescott -- the formula to calculate it.
.
it's not in here.
https://ru.wikipedia.org/wiki/Прескотт,_Эдвард
Give a description of this prescott -- the formula by which the calculation is made.
HP is a well known filter, it is available in the code base. It's not about it at all. If it annoys you so much, you can throw it away, but the point won't change. And it is as follows.
We take the original cotier and extract a deterministic component from it. In my case I do it with HP. Then we model the remainder (noise). The idea is not mine, the technique is standard. It was first published 40 years ago by Box and Jenkins as an ARMA model. I take HP instead of regression - more efficient. I think it would be more effective if I took a wavelet.
But I'm not up to it at the moment. I'm trying to show the consistency of the prediction and the details are not very important. What is important is the TC design process itself, which can be numerically controlled at each step. One such transition from model #1 to model #2 was made based on analysis of regression coefficient estimation results. See what happens next. In my experience the model I use should not work, but it does!
Give a description of this prescott -- the formula by which the calculation is made.
.
It's not in here.
https://ru.wikipedia.org/wiki/Прескотт,_Эдвард
Look here.
But it's better in the codebase.
Logged in, but the pages themselves don't open.
Can you retell it?
branch summary:
Harry's methodology is based on this. As far back as Wyckoff argued that there is a group of highly capitalised organised operators in the market who "play one handed" and he called this group the Composite Man or simply the Market. It is this group that buys on the lows because they create those lows. Harry built a quantitative model that maximizes the income of this very group. Atits core, his model is similar to the "bottleneck problem" formulated by R. Bellman, and Harry's equations are likely to resemble those of dinamic programming .
The bottleneck problem in the classical formulation sounds like this. There is a final product A that requires raw material B, which is produced in system C. But to increase the capacity of system C you need to consume raw material B. This is the bottleneck. The goal is to maximize the output of A in a given time interval. The optimal strategy would look like this - up to a certain point, work on increasing the stock of raw materials, and at a certain point there is a "breakthrough" - that is, raw materials are sent to the main production. The analogy with the market is not difficult to draw.
Further. Harry's model has an essential duality - existence of two lines of a possible market way - the flat phase (accumulation of stocks on one of the sides) and the trend phase (initiation of a trend in this direction - analog of the main production output). Note that prior to this, the presence of trend and flat was only an observable phenomenon! Once this type of model is built, the division into phases becomes fundamental.
My problem - I can't set the problem correctly. Roughly speaking - what prevents CMs (Composite Mans) from pushing the price sky high after accumulation. What prevents costs from rising as the price moves, because to move the price up, you have to buy, i.e. increase the average price of your long stock and somewhere there is that optimum price above which profits start to fall.
In Harry's terminology - the base line is the level of accumulation of reserves, the projection line is the appropriate price to which CMs accelerate the trend. What matters is the duality, i.e. it is not known exactly, when the breakthrough (bifurcation point) will take place, but it is known that if the breakthrough takes place at point A, the price will go up to point B.
My view is that the essence of the model is not the division into insiders/cooks and suckers. That division is conditional. The essence of the model is that the potential for movement is determined by the losses of previous entrants. Harry seems to consider the flat phase as a period of accumulation of open positions, accumulation of some potential, and the trend as its realization. But there may be other realisations - they should reflect behavioural patterns: accumulation of positions, fixation of losses and profits.
A summary of the content of the thread:
For me it's undeniable that the basis for the market is the psychology of its participants, this is primary, not the supply-demand.
The main attraction of state-space models to me is the possibility of (allegedly) modelling "unobservable" random variables - in your case , they should reflect behavioural patterns of position accumulation, loss taking and profit taking .
But I have no idea how to put such things into a model.
In the state space model - "these are the states", which, along with other quotes, are taken into account when forecasting. There can be many states, e.g. interest rate, taxes, gross product, etc. These quantities enter the forecast calculation with their numerical values and noise.
I read your "Formalising ...." thread. It's very interesting. But we need to focus on something else - the predictability of the quotient. Not predictability, but predictability. This is the issue I keep trying to push through in this thread. Here I am calculating the forecast. Why should we trust that prediction? In TA this question is not raised at all and to the vast majority it seems wild. We have drawn support lines, we have taken into account the movement in the bullion, Elliot waves and even Elder waves - everything will be fine.
All the numerous approaches and methods are aimed at analysing the past, but we need to make a trading decision - to predict at least one step ahead. The analysis of trade orders is again the analysis of the past. We learn about the fact that they have already been accumulated, the market has already gone and we are trying to jump into a tram!
At this stage I see the following guarantees for the predictability of the quote:
A neatly designed TS, which by virtue of its measurable parts increases our probability of forecast execution
prediction error.
Take a look here.
But it's better in the codebase.
You got me wrong... It's all about interpretation...
.
.