For a long time we have known that different algos work better on different days. Putting that into Market Profile speak, an algo usually performs differently on a Trend Day ( or a more trending day) than on a Normal Day (when the market is rotating).
I have long been setting up algos to only trade at a certain times of the trading day when I believed that certain activity that suited the algo would more likely happen.
But there is a more hands on approach. The object of the algo is for it to quickly identify a trading picture and act on it without me being required to be alert and glued to the screen. Much of the time I'm in my office doing research or putting on option positions etc. I love what I do and am never bored in this business.
I was often frustrated when I saw Flo take a trade that I "knew" was going to be a loser. When I first started trading with algos I often over rode the signals until I realised that I was wrecking the math of what I had created.
Taking one step back, what if a trader can identify the type of day that was happening, or just the type of market that was happening, and turned on the appropriate algo that was suited to that type of market. That would require a little attention but nowhere nearly as much as discretionary trading. Having the algo identify these activity changes was something I have tried to do and am still trying to do with different degrees of success. However, if a trend day starts I can easily tell Flo to only take trades in the direction of the trend. If I have a quiet day I can turn on a Flo that trades well in a rotational market.
Makes more money and takes away the frustration. Of course when I'm not there, Flo does her thing and gets more "average" (profitable but bigger drawdowns) results which is fine as that's the price of freedom.
Today's chart is a typical DAX trade. The market was called to open gap up - ES was trading higher overnight - so I turned on the Flo that trades this very usual trade. As soon as Flo entered the trade I went to my charts and quickly found the target and entered the order. No rush, no sweat. I also knew where the price had to go for me to be wrong and a stop was placed there.I then turned the automation off. There is no law that you have to use the same algo all day. The markets often trade in three different modes - the opening hour or two, the middle and the closing hour or so. The reason for this is that it is at the open and the close that the bigger market participants have to create and balance positions.
As you can see, entry was at 8254.0 completely automatically in accordance with the logic in the Bloodhound FloBot. Exit was on a discretionary but rule based basis (trading plan) at 8274.0. The profit of 20 DAX points or 500 Euros per contract meant that my morning trading could be done for the day. You can see where Flo would have exited the trade on an automated basis, at 8270.0, 100 Euros less profit. By trading as a hybrid trader, it is possible to consistently beat an already profitable algo by concentrating on manual trade management, a lot easier bar to meet than trading as a fully discretionary trader. The chart below is using the PureLogik Universal Renko Bar
No comments:
Post a Comment