Have you ever tried running a fully automated EA before? Did you code it yourself, or buy it premade? How were the results?
What has been your experience with expert advisers?
- Thread starter farmerjohn1324
- Start date
fj, that was quick work.
What advice on starting out / setting up an EA system?
Links?
At first you need to develop a system. Depending on who you ask - that is the hardest part. You backtest should be solid, data of high quality and trades sample statistically significant. Thousands of trades, preferably across all the market types.
Then you need to figure out execution environment. I think Expert Advisor is normally mentioned in the context of Forex & MT software. But conceptually that could be any program that generates buy/sell signals. If you're using MT they have a way to code them so you don't start writing your own environment. I did write my own because commercially available ones are not sufficiently sophisticated, also, they wouldn't achieve the level of automation I wanted and wouldn't work with my data sources anyway. Mine is basically a Kotlin application that monitors the market and executes all my strategies. Generates reports, manages positions etc. I can command it a via Slack bot if required, but mostly don't touch it for weeks at a time.
Once execution is automated you might want to put it on a server so it just runs. And all tasks around it are automated. For instance with IB - handling daily software restarts, updating data (depending on your provider) etc.
There are probably a lot of platforms now where you can "rent" something like an EA and hook it up to your broker. I personally wouldn't use it as sufficient information to validate their strategy/performance is not typically provided.
In my journal I wrote a bit about details of what I do, but likely more on the strategy validation topics as those tend to be where people mostly fail, in my opinion.
While there are some details there on the automation itself, I didn't spend too much time describing it. Maybe I should. But again, I just believe technology is not the issue for most traders. If you have some sort of trick up your sleeve that you know makes you money thru reducing slippage or commissions and commercial software doesn't support it - then yes, technology might really help to improve returns or even enable trading certain strategies.
Val
Yes. I coded it myself in MQL5. I start it on Sunday evening and let it run until the end of the week. I haven't experienced any software crashes yet. I have never bought any premade EAs.
What's the results in % return?
You are aware that I was avoiding answering that question? The financial result is not dependent on the coding platform which is being used. So whether it is an EA, or coded with e.g. IB's API would make no difference. The financial results depend on the trading system, not on the software platform.
Having said that, the financial results of my EA are not good. I've told you previously that I coded this EA to get some experience with that MT5/MQL5 platform. Even though the financial results of this trading system are not good do I continue running it, to see whether at some point some weird software issue would emerge. You could call this a "tuition fee".
From June, 2019 to June 2020, my profit was 147%. Since June, 2020, my profit has been 69%. That's over 160% per year if this continues. I'm happy with my strategies. I use Tradestation Easy Language on daily bars only. No day-trading. I tweeted all my trades BEFORE I executed them. BTW, buy NAIL and SPXL Monday, 10/19/20.
CongratsFrom June, 2019 to June 2020, my profit was 147%. Since June, 2020, my profit has been 69%. That's over 160% per year if this continues. I'm happy with my strategies. I use Tradestation Easy Language on daily bars only. No day-trading. I tweeted all my trades BEFORE I executed them. BTW, buy NAIL and SPXL Monday, 10/19/20.
how to view your tweeted calls?