Quad test

Over the next few days I wanted to test the DMA against the MAFR versions.  You can see them streaming live by clicking the preview on the right.  There are two DMA versions on the left and two MAFR versions on the right.  The EAs on the top row are running on the 1 minute charts with 5 and 15 minute higher order timeframe references.  The bottom rows are also running on the 1 minute charts, but with 30m and 1h higher order timeframe references.

I’m also using the upper standard deviation as the TP, and the entry trigger is PN=0.2 (see mathematics).  I expect this setup to place many more trades, which should make it easier to gather data and calculate metrics.

Advertisements

4 Responses to Quad test

  1. gatornuke says:

    Oops!

    During testing i realized that positions were not being closed when the bias changes. After looking at the code i realized that where this is handled in the trading decisions subroutine, the statements are “else if” after a large conditional statement that is almost always true, therefore the subsequent “else” are not entered. I fixed this by making them “if” statements instead but i have yet to see a change in bias since i restarted the test this evening, so I don’t know that’s i’t working for sure (although it should be).

    For now I’ll just continue monitoring and hope that’s what was causing the horrific performance yesterday and today.

  2. zzuegg says:

    Hi,
    Interesting Strategy. You are placing quite a lot of trades. Just a few questions, Why do you open so much trades? Depending on your “reopen” Strategy (if you use some kind of averaging) it might change the picture on the prediction quality.
    Do not get me wrong, averaging can be a good tool, but i would be interested in the pure prediction quality.

    Another thing is that on your livestream i can hardly see any operations, maybe a detailed review, or a statement updated daily would be nice.

    Best Regards
    Michael

    • gatornuke says:

      You’re right, some of the EAs (especially the DMA versions) have been placing a lot of trades. I chose a low enought trigger to collect lots of data, while still remaining (theoretically) profitable according to calculations in the Mathematics section. That being said, the Japanese crisis has decimated the performance of these EAs, probably because they’re using stats as recent as last week when this crisis wasn’t around. In addition, I found a couple of bugs in the code preventing timely exit, so I might re-start the data collection next week as a result.

      Either way, a detailed review will be made available, although probably on a weekly basis rather than daily.

      Now, I told you how I’m getting a lot of trades due to the low trigger point, but I’m listening. Do you have any other suggestions to capture predictability?

  3. gatornuke says:

    We’re changing the strategy a bit for next week. Instead of allowing an unlimited number of trades whenever triggering conditions are met, we’ll allow only one trade per EA. When the triggering condition is met (we’ll keep using a PN value of 0.2 with max std TP) a trade will be entered. If a position is already opened, the TP and SL values will be modified instead. Positions will be closed when the TP is met or when PN is -0.05 (opposite of the trade direction).

    This new strategy should be a better barometer of the predictability of the EAs, and will also result in behavior consistent with zulutrade’s guidelines if and when I decide these EAs are good enough for commercial signal broadcast.

    I addition, you may have noticed that the cycle time for the mafr versions can be over 800 seconds. This is not very helpful when we’re trying to predict behavior in the 1m charts. I ordered a new machine with an intel i7 2600 and 16 GB of DDR3 1600 ram that should help with this, and i’m expecting it to arrive this week. I actually considered getting a dual opteron with 32GB of ram and 16 cores, but after some consideration, it appeared that the mainstream i7 would be better suited for this task than the Opteron system. As long as i don’t use up all 8 threads with the i7 system, and keep it to less than 8 simultaneous EAs, the per-core speed of the i7 system should deliver better performance than the opterons.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s

%d bloggers like this: