It is really whatever you want. It might help to think about it as - to achieve a fully automated trading you need (a) very formally defined strategy (b) software to be able to execute it live. "a" is normally what is called an algo. You can have just "a" and trade of its' recommendations. Benefits of "a" is an ability to backtest and see in advance how it would perform in different market conditions. If you run it over 10-30 years of data over multiple market cycles you will get an accelerated education on what you will likely to face. What you normally don't know is what type of market we will be having in the future but it is likely on of the ones that already happened. You could define separate strategy which thrive under different market condition and doesn't loose too much while waiting for it, then combine them and aim to make money in most markets. Benefits of "b", as others pointed out is ability to not stare at the screen all day waiting for your setup and not having your emotions interfere, as computer will do exactly what you told it to do. Here a link to my journal where I talk in details about this stuff
2-4 signals a month?? So, what is the win-loss ratio? Does it make any sense to trade such strategy without huge amount of funds?
Cloud watchers can sometimes see more than people who are lost in routine. All depends on perspective. Automation is lazy peoples way of "working"..
Not so. Automation is damn hard yakka, much easier to trade manually & discretionally using algo as a prop, and, I think, probably better results, unless it's HFrequency.
I wonder if there are tools available to make setting up an automated system simple. If there were, it would be in USA only I imagine. I've tried but gave up, but using IB as a broker made it difficult in my circumstance.
I dont think there is universal but easy to use tool yet. Ib has decent api, to connect to custom software. Many programming langs are fine, at some point it might become easy to implement ideas... Especially if working for years and in narrow field. I built all tools from ground up. At detail level everything becomes easyer,but can take alot of time... Piecing all together , end result can be more complex. Also improving interactions of componenents can take alot of time.
I attempted via Sierrachart & IB. However there are/were incompatibility issues with those two which made it hugely time consuming. Besides, IB drove me crazy, forever updates and constantly being logged out by IB. I reckon it was one of my most stressful life experiences.