The future of automated trading - page 7
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
Here's a comparison, not a bushcraft, but a CQG and Matlab bundle http://www.automatedtrader.net/news/algorithmic-trading-news/35299/cqg-connects-to-mathworks-matlab.
At first glance a great bundle. The task of the trading platform is to trade ... http://www.itg-direct.com/index.php?option=com_content&view=article&id=77&Itemid=96 nothing extra
The task of Matlab is to analyze and give commands to the terminal ... the task of the terminal is to trade
You'll laugh, but I don't know any forum dedicated to Matlab and CQG bundling. Where can I touch this great couple live and for free? However, I will not argue. If you like the bundle, use it :) But tell me about it later, OK?
I don't know either. It's just that in this thread Renat said something about using Matlab in trading. I tried to look for something similar. I just got curious. Unfortunately I'm not fluent in English. But I've dug these links. I understand that CQG is not a "dummy" and Matlab is also a decent program. As far as I understood the text in the first link (the link is talking about the connection between the two programs). Perhaps I'm wrong, but the idea is interesting ... If MQL creates a nice and comprehensive connection between terminal and matlab, it becomes a very interesting solution and can attract many new users...
When Matlab has full access to the trading account, the market environment and can trade on its own, then you can start asking questions with comparisons. In the meantime, it is a very good system, not directly related to trading.
Matlab is for analysis. And in this, it has already gone so far that the mcl will never catch up with it. You can imitate matrix calculations in the µl, but only imitate them. Something like PCA analysis for a matrix of, say, all S&P500 companies will not appear on µl simply by virtue of complexity. And for stock-market trading, these are the very basics, nothing advanced. Matlab can get market information from several sources at once, including Reuters, but where will MT get its information from?
Connecting it to any broker or exchange via API is also not difficult for professional programmers. The main thing is analysis, execution of trade orders is an easily solvable technical task. A google search for matlab API trading will bring enough information for those interested, there's also a link to CQG, Interactive Brokers, Oanda.
You'll laugh, but I don't know of any forums dedicated to the matlab and CQG bundle. Where can I feel this wonderful pair live and for free? However, I will not argue. If you like the bundle, use it :) Just tell me about it later, OK?
The notion that because of the increasing efficiency of the markets (due to robots, one would assume), it has become virtually impossible to make money comes from an attempt to justify your own powerlessness. As ten or twenty years ago those strategies that worked and gave good results are still working and earning as well as back then. Some of them are even available to the public, printed in books and everyone knows about them. The same strategies that did not work then, do not work now. Even a simple MACD strategy can significantly reduce the risk of an adverse price movement. It will not cause a stable growth, but it will help to avoid sharp drops and uncontrolled growths, which you can see in the symbols.
Examples of strategies that work, with proof of how they work, in the studio!
"Increasing market efficiency" is not an opinion, it is a fact proven more than once in the academic literature.
Once again the discussion has descended into a banana pitch. But that's not what this topic is about. The topic is about the prospects of automated trading, not about specific programmes.
The market is changing. It's not clear whether automated trading has good or bad future, but the future has already begun. The opportunities to make money in the market are also changing, simply because traders are changing, they are now silicon traders. What chance does a loner with matlab or mcl, whatever, with a weak internet connection, with commission eating up most of the profits, have against the big sharks with colocation on the exchange and much more advanced software? Where is the advantage that will help it survive?
Connecting it to any broker or exchange via API is not difficult for professional programmers either. The main thing is analysis, execution of trading orders is an easily solved technical problem. A google search on the words matlab API trading will bring enough information for those interested, there's also a link to CQG, Interactive Brokers, Oanda.
I am just talking about these very "professional programmers" - crutch on crutch they make. Each time, everyone does it for himself. And often on a trade api, where there are practically no historical datafeeds. To the question "where is the history?", the vendor usually answers "we provide the best trade api, but you collect the history yourself". So, these programmers have to collect and accumulate history at home, emulating the market environment.
And we provide a ready-to-use and working system for hundreds of thousands of traders. And the integration with other systems (transmission of complete market information with the history, receiving signals, etc.) is not just possible, it works.
The word "can (take)" is a couple of orders of magnitude weaker than "does (take)". Especially in the developer application. "Mighty" rarely produce working and massive results, limiting themselves to theoretical applications.
What are the chances of a single trader with Matlab or µl, with weak connection to the Internet, with the commission fee eating up most of the profits, against big sharks having a colocation exchange and much more advanced software? Where is the advantage that will help it survive?
In fairness it should be made clear that often "advanced big shark fast trading software" is just an ordinary scalper with 1-3 pages of business logic in code. No need to surround brokers' trading tactics with myths.
And when talking about "small trader vs fast broker" you should also clarify "we are talking about competition in scalping and lack of strategy". Go into the creation of long-term strategies and the primitiveness of "the battle is for every cent" reasoning becomes clear at once.
But traders often descend into scalping - it's easier, there is no strategy (although they convince themselves that it is a "strategy") and it brings money faster. Then the subject logically evolves into "they do not let us trade", "they do not give us pips", "the quality is wrong" and passes into "a disaster, the trader cannot trade with pips better than the broker's big shark".
The word "can (take)" is a couple of orders of magnitude weaker than "does (take)". Especially in the application of developers. "Mighty" rarely produces working and massive results, being limited to theoretical applications.
Getting only trading features with minimal datafeed with no history is fine. You don't need long history to trade, just like you don't need charts. It is needed for analysis and testing. Normal guys do not accumulate history, and buy it from Reuters, at the very least CRSP or SIRCA. This is not the case with IMF and SIRCA. There ticks from 150 markets of the world - 60 gigs of date per day.
Maybe it's all about perspective?
But traders often fall into scalping - it's easier, no strategy (although they convince themselves that it is a "strategy") and brings money faster. Then the subject logically evolves into "they do not let us trade", "they do not give us pips", "the quality is wrong" and turns into "a disaster, the trader cannot trade with pips better than the broker's big shark".