Thursday, 11 July 2013

Shortening the Time to Consistent Profitability

In my post here, I wrote about the strengths of Automated Trading. I want to take this further in today's post.

Traders seem to be able to learn the mechanics of trading quite quickly. They can find a good trading picture or copy someone else's and get to CP in SIM in a very reasonable time. The big leap seems to be getting from CP in SIM to CP with real money.

Automated Trading is a way to make that transition much more quickly. If someone has worked out a trading picture to trade and made it into an algo, backtested and forward walked it and then successfully SIM tested it, then the only way they can sabotage their success is to either turn the algo off or to override its signals. Two failings that are easier to cure than having to make discretionary trading decisions ten times a day.

What is interesting is that you don't need to be a programmer to create your algo. On May 22 I started using BLOODHOUND and am live trading. Flo loves the'hound. I can setup more interesting rules that I didn't have the competence to do in easylanguage.

Additionally, I'm using Logik Ultimate Renko bars. These take even more of the noise out than the range bars. Renko bars are complicated and have many pluses and minuses. I'll post separately on this topic but for the time being I can say that all renko bars are not created equal. Far from it. I tried about a dozen of them before I found Logik.

I haven't recommended much in the blog in the past but these two add-ins to Ninjatrader are together a quantum leap for me.

Full Disclosure: I'm now an affiliate of both Bloodhound and PureLogik. My aim is to help both these good products get into the hands of more traders on their road to CP. All financial recognition from these vendors will go to Macmillan Cancer.

The chart and stats below tells the story. It's a very simple reversal algo. I am running a number of different algos on a portfolio of markets.


8 comments:

  1. Looking at those stats it seems you've given up your requirement of a high win rate. 46%?!

    ReplyDelete
  2. Anon 15:53, Everything is a trade off. I'm running a portfolio with 3 different algos over a number of markets with win rates as low as the one I've shown and as high as in the 70's. The other useful metric is the average trade as that, together with other metrics, tells me how robust my system is and how much slippage I can stand in a volatile market. I'll tie all this up with a second part of the Bost about The Business Plan.

    ReplyDelete
  3. Hi,
    Even with a sub 50% win rate, the figures still look very impressive. How many contracts is this from?
    US hours only, or 24x5?
    Cheers,
    Ross.

    ReplyDelete
  4. Wow BLOODHOUND is amazing!!! I can take all of your methodology I’ve learned and plug it into BLOODHOUND then tweak the logic to get more of a discretionary trader result without the human part best of both worlds.

    ReplyDelete
  5. jenrique42


    as I can see in the T / S the " SELL AT THE ASK, AND BUY AT THE BID "

    ReplyDelete
  6. Any updates on Bloodhound? I'm still learning all of the Solver, Function and Logic Nodes looks very promising.

    ReplyDelete
  7. Hersh, doing a lot of model building and testing. Going well. I'll post as soon as I have something that's interesting to share.

    ReplyDelete