How can you create/find a truly profitable EA? - page 4
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
The Event Happens: A catastrophic incident occurs.Human Journalistic Delay: A human journalist or agency editor must physically witness it, verify it, and manually type out a text headline (e.g., "Breaking: Explosion reported at..."). This step alone takes anywhere from 30 seconds to several minutes.
To be clear, the journalistic delay applies regardless of the method of reading, watching, and/or listening. That imposes no additional delay on algorithmic trading.
Your EA only checks the web feed at set intervals...
How would you know? My EA isn't public. No, my EA does not check news feeds. It simply detects price reactions to whatever in milliseconds. When price flies around, it gets scalped. If you're worried about heavy slippage, use the appropriate working/pending orders.
institutional algorithmic traders with multi-million dollar direct feeds (like Bloomberg Tomorrows or Reuters Alpha) have already emptied the order book.
If you're going to rate every aspect of retail trading as contrasted to every aspect of professional trading, you're in for a lot of dismay.
I'm not here to promote commercial news wires, but the most popular retail-priced news wire (not for an entire pro trading floor) having a AAA latency rating sells for 367 USD per year. It requires a ping every 10 seconds to avoid termination of the connection, and news is delivered in real-time.
A well written algo is faster than any human... period.
To be clear, the journalistic delay applies regardless of the method of reading, watching, and/or listening. That imposes no additional delay on algorithmic trading.
How would you know? My EA isn't public. No, my EA does not check news feeds. It simply detects price reactions to whatever in milliseconds. When price flies around, it gets scalped. If you're worried about heavy slippage, use the appropriate working/pending orders.
If you're going to rate every aspect of retail trading as contrasted to every aspect of professional trading, you're in for a lot of dismay.
I'm not here to promote commercial news wires, but the most popular retail-priced news wire (not for an entire pro trading floor) having a AAA latency rating sells for 367 USD per year. It requires a ping every 10 seconds to avoid termination of the connection, and news is delivered in real-time.
A well written algo is faster than any human... period.
I'm a beginner in algo trading, so I really enjoyed reading this forum thread. One more thought: I agree that you can't build a universal robot, but isn't it possible to pick a strategy and optimize it — tune the parameters for the current market conditions — and then, every few months for example, review and retune those parameters for better performance? Wouldn't that work?
I'm a beginner in algo trading, so I really enjoyed reading this forum thread. One more thought: I agree that you can't build a universal robot, but isn't it possible to pick a strategy and optimize it — tune the parameters for the current market conditions — and then, every few months for example, review and retune those parameters for better performance? Wouldn't that work?
As you progress from a beginning coder to intermediate and so on, you will likely develop tactics for reorganizing price data in a way in which market phase changes are no longer such a shock to the system. I prefer to construct custom charts wherein every bar contains the same amount of certain data, e.g., Renko bricks, equal tick volume bars, or range bars. In this way, the source data that I'm looking at (and my algo is analyzing) is sensibly smoothed─thereby eliminating the shocking chaos. In such custom charts, there are merely shorter price runs versus longer price runs. There are no flat bars nor giant bars because time is removed from chart construction.
I find that the problem with continuously optimizing is that it involves chasing a moving target. It relies upon a recent price pattern repeating itself, and hoping that price pattern exits in the most recently selected optimization data sample. Having analyzed the aforementioned custom charts with a nearest neighbor utility, I can tell you that most recurring price patterns can and do go back many years─and even decades if the historic data exists in MT5.
Even a manual trader can suffer some losses, and then adjust indicator parameters to look better on more recent data. Again, the problem therein is that any reasonable degree of certainty in statistical analysis, probabilities, and data forecasting goes out the window.