Current Events and Path Forward

As you (the one visitor a day I occasionally get) could imagine, I’m currently in the middle of Experiment 6 (E6).  I’ve been conducting this test all of last week and will conclude it at the end of this week.  I previously noticed that a number of EAs would only hold their positions open for an hour or two, even though the forecasts are 4 hours in the future.  I decided to test the prior setups versus 2-hour forecasts setups, and use the extra computational margin gained to run more histories.  This way I halved the forecast length and doubled the number of histories, to keep the total number of random samples equal.  The results so far indicate that the 2-hour forecasts do have a significant edge over the 4 hour configuration.  It’s still premature, and I won’t post results until the end of E6, but that’s the indication so far.

After E6, E7 will test a variety of entry and exit configurations for the DMA variant in a 2hr forecast mode (if E6 results remain consistent until the end of the week).  I’m currently testing an entry of PN>0.18 and exit of PN<.03 or TP/SL at 2 SD, whichever comes first.  E7 will test entries of 0.2 and 0.5 and exits of PN<0 with 2SD TP/SL, exit on TP/SL or opposite entry, and exit on opposite entry only (no TP/SL).  I haven’t had a chance to explore different entry/exit setups since I’ve been comparing different EAs against each other, but my recent work is indicating that DMA 1-5-15 2hr 2SD gives the best results.  I’m excited to run E7 and see how my DMA EA behaves under these conditions.  Even though it is profitable, I’ve noticed it often leaves a lot on the table by exiting too quickly, so these alternate conditions will test a less jittery EA against itself.

After E7, I think i’ll have narrowed down the EA’s profitable and safe operational parameters enough to start exploring other currencies.  I’ll be using the results of E7 to configure the DMA EA in E8 to test its behaviour in EURUSD, USDJPY, USDCHF, USDCAD, GBPUSD, and AUDUSD.  This selection is somewhat arbitrary and may change before E8.  Ideally, I’d like to test everything under the sun, but i have limited computational resources.

As a side note, I’l like to mention that I don’t plan on posting experimental updates as independent blog posts anymore, since experimental results are all documented as pages under the “Results” section.  I am however, planning to continue broadcasting the live video on livestream.com/gatornuke, although I admit this broadcast has not been uninterrupted.

Advertisements

8 Responses to Current Events and Path Forward

  1. Roman says:

    I like your approach in general but I am unsure about your experiment settings. As far as I see, you are testing on the data being produced in the particular weeks, in which case you would test different parameter settings on different data sets. I guess, I’m basically interested in the distribution (or variance) of the testing results. Another potentially important issue might be overfitting, which may likely occur with all these different parameter settings. Do you plan to address this one?

    I’m looking forward to more results,
    best regards

    • gatornuke says:

      That’s right, as every experiment is being conducted in real time, they all invariably occur on a different data set. However, the orthogonality and blocking of the setups ensures an adequate comparison is made between them on that given data set. I’m analyzing the data from these factorial experiments with Welch’s t-test , approximating degrees of freedom using the Welch–Satterthwaite equation, and then obtaining the p-values. But back to what you’re saying, this is still being done on different data sets every time. The assumption I have to make is that the behavior of these setups relative to each other over the two or three week’s 1-minute charts will remain consistent thereafter.

      As for overfitting, I’m not particularly concerned about that. Typically overfitting occurs when you have a strictly mechanical EA (fast EMA > slow EMA = buy, etc). In this case, tweaking the MA period may give you good results over a given test period, but may not be the same afterwards. I’m not changing the way the EA calculates the Monte Carlo projections, but rather I’m comparing different methods with different sampling functions against each other, and comparing different entry and exit thresholds. I can see why at first glance It may seem like I’m doing the same thing, but the difference is that I’m evaulating which results to use rather than tweaking how those results are being calculated.

      Thanks for your comments.

      • Roman says:

        Thanks for your answers.
        But the different entry and exit thresholds are part of the EA, right? Isn’t it that it still breaks down to the question of how good your 1-week data represents the entire distribution of market states (given the assumption that it remains somewhat stable over some next period of time)? As far as testing concerns, you may at some point run into multiple hypothesis testing problems, but given your background you are probably aware of that.

  2. gatornuke says:

    Yes, it’s not perfect, but I have to make the best of the limited time and computing resources I have available. Besides, even if i was able to run all the experiments concurrently and on the same data set, there is no guarantee that the market will continue to behave like that in the future. I have to assume that changes in behavior will be gradual enough to be captured and incorporated by the sampling functions, rather than an overnight shift in market dynamics. That’s not to say that we couldn’t have a catastrophic collapse of the euro tomorrow from a Greek, Spanish and Portuguese default, but in case of a cataclysm like that, I would stop execution and allow at least a few months for the new market dynamics to be established and incorporated into the EAs.

  3. Roman says:

    Thanks for your information. I’d like to say that I really did not want to pound your project in any way. Far from it – it looks both methodology interesting and promising, so I hope you can keep up your enthusiasm and I am looking forward to seeing the final outcome.

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: