I wrote a bot and gave it 25k to trade

Discussion in 'Automated Trading' started by _terminus_, Feb 23, 2020.

  1. henry76

    henry76

    Point taken about % , but I think somewhere you said you closed positions on open, implies position held through close, and therefore held for some time , which even in relatively small mkts means a lot of liquidity and diluted effect of "followers" .
     
    #191     Sep 9, 2020
  2. _terminus_

    _terminus_

    The system can hold positions for many days and it can close them at any time, not on daily open only.

     
    #192     Sep 10, 2020
  3. d08

    d08

    Banks usually don't deal with anything like this. Many of the employees in second-tier banks come from schools that teach that timing the market does not work, only investing or interest does.
    It's especially bad if you're someone off the street without a well-known name to connect you.

    Best to find business-minded people who are ambitious and open-minded, they are around.
     
    #193     Sep 10, 2020
    _terminus_ likes this.
  4. _terminus_

    _terminus_

    Well, as they say, opportunities come when you are ready. In this respect I guess I'm only half ready. I'm still waiting to see if the returns will allow the system to go on its own, or if at some point I'll have to look for some kind of partnership.

    What's your take on Collective 2 and similar services?

     
    #194     Sep 10, 2020
  5. d08

    d08

    I'm not a big fan. I've mostly traded capacity constrained strategies, so any competition is a no-go. Slippage is another issue -- I don't know if the signal subscribers were trading 1 contract or 100 contracts. It makes a big difference.

    I've seen a few systems over the years that did really well and were solid, they always closed up and eventually used outside money. The ones remaining are mediocre to bad.

    As sleazy as it sounds, I'd only sell something on C2 that worked temporarily to eventually implode.
     
    #195     Sep 10, 2020
    virtusa and _terminus_ like this.
  6. _terminus_

    _terminus_

    +++ Monthly update: +6.11% in September with 4 trades +++

    The (relatively) long term, long strategy is doing fine. +65.52% since inception as calculated by me, +76.70% as calculated by IB. The recent flat line in the lower graph represents the pause when the system stopped trading during the September meltdown.

    In a sense, the two strategies that I was able to come up with so far can be seen as mere drawdown reducers. They just buy, with leverage, when the index is rising and go flat when it's going down.

    Still working on a viable, automatable shorting strategy.

    results.jpg

    portf.jpg
     
    #196     Oct 1, 2020
    d08 likes this.
  7. d08

    d08

    Impressive stats.
     
    #197     Oct 1, 2020
    _terminus_ likes this.
  8. traider

    traider

    Can you share backtest results for the past 10 to 20 years?
     
    #198     Oct 1, 2020
  9. _terminus_

    _terminus_

    Not that far, but here they are for the last 6-7 years for the last strategy (MNQ):

    bt.jpg

    Y axis = %
    X axis = no. of trades

    Note: the graph reports the series of trades as they would have occurred, but it doesn't take into account the dates. For this reason it shows no flat sections, as the trades are presented one after the other.
     
    #199     Oct 1, 2020
    ValeryN likes this.
  10. They

    They

    @_terminus_

    Have you checked out https://www.worldcupadvisor.com/ (Also known as leader/follower software)

    To my understanding you are charging a subscription fee based on the amount of contracts needed to trade the system. If a follower wants to have more contracts traded then they will have to pay for another subscription.

    If your system is scalable and good enough - low capital requirement, low drawdown, above average returns - brokers using the leader/follower software will pimp your system for you.
     
    #200     Oct 1, 2020