8 Comments

  1. i think that's a really good basis. i think along the same lines starting with as little as possible. the advantage of more trades is that it increases confidence. but after a certain number of trades the number of trades starts to matter less and less. also assessing the entry and exit strategy separately is useful. i don't write bots for forex, but for crypto. one advantage here is free data and a direct connection to the exchange (no brokers). the protocols are much simpler too so it's easy for programmers to get started. if you don't use an existing platform then managing the data is a big job by itself, but it adds flexibility for me. disadvantage of these exchanges is that often things like stoploss orders don't exist or they can't be moved. so it's a doubled edged sword. but, i haven't used my bots much (always tinkering), so i can't really say how well this stuff works yet

  2. This is very similar to what I do for trading except for I'm on crypto with what I do. if I'm in a range-bound environment for example I can set it to not do any trades after a certain price and not to start any trades if it goes below a certain price and have a stop loss. I think really the reason it works so good for me is it takes a lot of me out of it when things are happening I still make all the decisions but it takes me away from the actual trade which keeps my emotions out of it. very glad to hear you do a video on this so I can hear your little snippets on how you would adjust it and what not big thank you.

  3. Hey Mark. I tested out 1 of your comments about earning interest with a short position.. I'm a swing trader and currently holding a short on USD CAD.. but my broker (IG) is still charging me overnight interest. I also tried holding a high interest currency against a low interest currency.. still they're charging me interest. Could you explain to us how to do this. Many thanks .

  4. It would seem most quants go for the 100 trades at a 100th of the size for better odds at getting it right, for instance with Bayesian statistics, and so that's how most high frequency trading bots are modelled, perhaps even using stuff like TensorFlow. Anyway some great ideas there. I'm dabbling a bit in it myself, but more in automating stop calculations, and the risk management and probability aspect of it.

Leave a Reply to FX Cancel reply

Your email address will not be published.


*