Forex Trading 2018- TA/FA- Setups and discussion, page-4630

Cheers mate, i’m not a fan of paper trading - for right or wrong I think it feels like a false environment and doesn’t account for the psychology which brings scrutiny and that critical self assessment on trade performance, so I just treat those accounts as games of monopoly so a real struggle for me, but I appreciate the sentiment, generally I find ‘pilot accounts’ the better apprach for me to test new things with money on the line and happy to lose a little to learn.

Will check out ricks posts and also heaps of good stuff on front pages too, keep me busy for at least a few weeks

Re Algos, massive interst to me - three of my mates are programmers / devs (and two of those trade stocks), and another of my mates is a massive TA bug - so algos is actually where I tried to start about a year ago, albeit started with ‘manually executing’ algos as soon as I got the ‘alert’, but never made it to the stage of backtesting because I had huge a run of losses and blew that account so got discouraged and concluded I need to do at least a year at the coal face to understand the basics before diving into algos, but didn’t have the time to sacrifice STT performance back then. So me and my mates still hold that longer term objective of algo tradiabd certainly have the coding skillsets on hand. Not sure if it’s if interest to people here but we already started to map out a few alternative data sources to try and make a more complex algo using AI (one of my mates works in this space and trades with me so salivates at the thought haha). An example might look something like this;

1) run queries on stock screener for whatever your setup is i.e. normal algo entry/exit criteria

2) run qeuries on alternate setups (even though these setups aren’t traded by step 1, it’s for statistical purposes) i.e. you trade breakouts maybe, but run a ‘second algo’ on a trend following setup to assess when multiple patterns align (the more the better IMO to feed the AI)

3) scrape data from key sites or other APIs which contribute to non-conventional TA, i.e. ‘hot’ sources of tips and buy/sell sentiment (broker positions data)

4) run AI through all of this to both backtest and screen in realtime the %probability of succes on going short or long on any given pair

So the two options is either have a core strategy underpinned by other strategies (so your algo setup might be 60% probability, and 8/10 of your supporting setups are showing) 60%+ so ‘green light’, or conversely your setup might show 70% success but other setups only 1/10 is 60%+, so the trade has less weighting. The other option is simply AI spitting out probably by crunching all lf this and adapting it’s own signal criteria based on live market performance, so you’re only ever buying or selling based on skynets instructions (with optioyof course to drill down into data and add your own ‘hard’ conditions to the AI as needed.

The latter obviously would take much longer to refine! Anyway, I can dribble for far too long sometimes, must be that second coffee hitting me - but yes have given a lot of thought to this algo area and remains a mid- long term interest and is why I want to get the fundamentals right in my mind before pushing this too aggressively. That said, doors are open if anyone on here wants to ‘JV’ an on exchange of knowledge/strategy/ideas to potentially co-build something with me and a few of my mates.

PS - does anyone here invest in algo ‘managed funds’ (one I know is AlgoTechs), I can never convince myself whether they are legit or just scams, there’s so many and such mixed reviews