Formalising common approaches to trading - page 20

 
Urain:

If you try to formalise these very patterns of manipulation then you don't have the right information, reliable statistics, whereas "nasty" patterns obviously have statistics.

So what should be implemented? Something that you have no idea about or something that has a base?

one technique implemented, how many there are left... probably the north80 can do it.

 
sever31:

It means that these nasty patterns need to be taken apart piece by piece. formalised, reduced to a mathematical, dynamically changing model of price behaviour. programmed, automatically recognising it on the chart, its hybrids, collecting statistics on it. I think we can see a lot of interesting things.

I am parsing and checking - the patterns are really ugly, because they occur on different TFs at different times, it feels like if I play always on one TF I will still have to see all patterns. And what is interesting, we all know that market is fractal and wave analysis describes the market structure, only one thing bothers me - how do I know that, did I get it myself? ;)
 
BLACK_BOX:

Perhaps if we are trying to formalise, we should start with a paradigm from which to build.

I can propose the following paradigm out of the many contradictory ones:

Market and information. The market is a mechanism for determining the current value of an asset. In the modern economic society this task is more than actual, there is a constant growth of requirements for the accuracy of estimation in time and in price. The value of an asset is determined by a body of information about it. Usually information is determined relative to other objects (or over an empty set of such objects, an "axiom"). Information has a horizon of influence and a function of degree of influence over time, "buy the rumour, sell the fact".

I agree. This has been scattered in different posts (exogenous/endogenous processes). Let it be in one :)

Trader. In common parlance, a Trader, is nothing other than an appraiser. The profession appraiser at all times was necessary. The task of each trader is to assess the real (fair) value of the goods. He voices his estimation with the help of his positions, the weight of his opinion is determined by the size of the position. The aggregate of all traders' opinions, weighted by the size of their positions, ultimately determines the current value of the asset. Good appraisers get bonuses - profit from trades, bad appraisers get penalties - losses, bad appraisers eventually leave the trade, they are replaced by new ones.

This is one of the trader's ways. More relevant to investors, as I understand it

Efficiency. Over time, the rate at which information affects the market is steadily increasing. This is due to the growth of technology and the globalisation of the economy. Any valuation strategy, (even more so if it is effective) is in turn information which is absorbed by the market over a period of time, or "there are no permanently profitable strategies, i.e. grails". Any historical pattern is also information which is absorbed by the market in a certain period of time, thus destroying any hint of stationarity.

agree

 
The purpose of markets is to enable people to trade. A market performs its function well if it serves as many people as possible. That is, the purpose of the market is to maximise turnover. The entire industry that lives off the % of it contributes to this. Therefore, a good market goes for volume. There are enough people who want to get a stop loss at the extremums and the market goes there :) In this regard methods that analyze accumulation zones of orders, traded levels, etc. are interesting. There are also appropriate tools such as market profile, volumetric profile, etc. Generally, these methods are a direct consequence of the trading organization, and the basis of their presentation has been described in the first posts of this thread. But in essence, the ideas behind them are too big and disjointed to be classified as a separate class. imha
 

It is worth discussing and systematizing the presentation of historical data, as it itself becomes a source of stereotypes, or market inefficiencies if one likes it better :)

As correctly noted by BLACK_BOX, detailed information about trades is T&S, but such information is available only for stock markets (or operating on the same principle) and it is quite voluminous. In addition, there are more complete views, which not only show the executed trades and their volume, but also the state of the stack at a certain depth (usually only the best bid and ask with their volumes). This information may of course be useful in some markets, but most markets deal and focus on compressed data of past trades. There are not many ways to compress this information and they are determined by the initial "fields", and there are 3 of them - price, time, volume. Further they discretize changes of one of the fields with a fixed step and summarize the rest on this interval or find extrema and initial and final points. Bar representation - discretize by the 2nd parameter (time). Equivolume plots - discretize by the 3rd parameter (volume). X/Zero, Renko, Kagi, etc. - discretise by price and discard volumetric and time. Volumes profile - also discretise by price, but with volume summed in each count, etc. You could more precisely describe and list all the ways of representation that exist and even that don't yet exist :) For more complex price transformations with averaging, modulation, etc. there are many indicators. But the main purpose of price or its derivatives is to show the alien actions. And these "aliens" must be a quite cohesive collection in their actions))) (group). Therefore, generating super complex indices by applying super approximations and quantum physics))) is an empty undertaking.

 

Avals:
Касательно назначения рынков - это обеспечить людей возможностью торговать.

If I may disagree, the ability to trade is a consequence of the existence of the market, the purpose of the market is the efficient allocation of assets to serve the capitalist economy. Or another contradictory analogy, the purpose of a factory is to enable people to work (and not at all to produce goods).

.... In this regard, methods that analyze order accumulation zones, traded levels, etc. are interesting. There are also appropriate tools such as market profile, volumetric profile, etc. Generally, these methods are a direct consequence of the trading organization, and the basis of their presentation has been described in the first posts of this thread. But in essence, the ideas behind them are too big and disjointed to be classified as a separate class. imha

I agree. The market file is a visual representation of the T&S tape at price levels and, in my view, is the most informative, by far, means of determining the distribution of volumes and the structure of bidders' interests.

Thus, I think we have come to the starting point: in order to formalise trading approaches based on endogenous information, it is, imho, necessary and sufficient to formalise them in relation to the market file as well as the T&S tape to monitor current activity.

 
sever31:

let's agree that those manipulations (no matter who is responsible for them) are typical price chart areas where most market participants are losing, because the logic of price behavior is opposed to profitable trading by an average person with his trading psychology, fear and greed. in short, we are losing on these patterns. it describes why and where and in my opinion quite successfully.

It means that these nasty patterns need to be taken apart piece by piece. they need to be formalised, brought into a mathematical, dynamically changing model of price behaviour. they need to be programmed, automatically recognise it on the chart and its hybrids and collect statistics on it. I think we can see a lot of interesting things.

I don't know why, but we are using head/shoulders and wedge with harami, of course we can't smoke with such armament)

I agree, but what is striking is the power of these manipulators' algorithm, which defeats any strategy. Apparently ticks grouped into bars provide more useful information to them than to us. Why don't tick charts exist? Are they difficult to organise? Or do they exist at some level or DC? It would be interesting to look at the true face of the market and try to guess the logic of the big players or their community.
 
yosuf:
... Why are there no ticking charts? Are they difficult to organise? Or do they exist at some level or DC? It would be interesting to look at the true face of the market and try to guess the logic of the big players or their community.
Google "Time and Sales" and "Time and Sales History". Or start reading the thread from here and get the answer.
 
BLACK_BOX:
Google "Time and Sales" as well as "Time and Sales History". Or start reading the thread from here and you will get the answer.
I have read it in part, but I want to know why is there no tick chart, although all information for its organization seems to be available? If I were to ask, who has tried to create such a chart with the help of programming?
 
yosuf:
Thank you for reading this partially, but my question is why there is no tick chart, even though all the information for organizing it seems to be available? Or has anyone tried to create such a graph using programming?
There are tick charts on many platforms except MT4.