Thursday, 21 November 2013

Algo Expectations

Trading fully automated is quite different from both discretionary trading and hybrid trading. In this case, I am referring to the amount of profit per contract per month.

While as a discretionary trader I might make, say, $2,000 per contract per month, That would be say an Average Trade of $100.

As a Hybrid trader I may be able to replicate that same performance and maybe even better it.

However, as a fully automated trader, I would be happy to have an Average Trade of $25 per contract per month.

The advantage of the fully automated trading is the way that I can leverage my skills. As a discretionary trader, I am trading one market. My profits are limited by the amount of risk per contract I am willing to take and the number of hours I am willing or able to spend fully focused in front of my workstation.

As a hybrid trader I can be a bit more leveraged and trade two markets at the same time.

But as a fully automated trader I can leverage myself and trade as many markets and contracts as I have capital for.

I had 5 markets that I was trading fully automatically today. Gold was the best performer and the ES was the least best. The other markets were the Euro (6E), Light Crude and DAX. By having that diversification I achieve a smooth upwards equity curve. Funnily enough, none of the markets lost money today.

Nowadays, I trade a mixture of discretionary, hybrid and fully auto. I run longer term positions using option strategies.

There are two schools of thought: put all your eggs in one basket and really watch the basket, and diversify. The choice is partly dependent on available capital.


  1. EL,
    I have been watching many of the Bloodhound videos. I've asked Bloodhound about backtesting once I have a system (and would like to quantify the results). They said it would be best to talk to Ninja Trader about that functionality. My Question is: 1) Are you able to back test using an excel spreadsheet by using data exported by NT, 2) or is there another way that you backtest? Thank you so much. Love your posts!

    1. Anon 15:07, Bloodhound has a backtest capability, SiRaven, that Ninjatrader utilises. Having said that, SiRaven does not look inside the bar so you will need to either use the Ninja playback facility to test or to have a saddle written, which I did, that looks at the data tick by tick so the backtesting is pretty accurate.

  2. how does the backtesting results compare to what you can get in MC or TS ?

    1. MC allows inside the bar testing on regular bars only so that impacts backtesting validity. TS is not as good as MC. The techniques you use for backtesting are as important as the platform.