Numerical Price Prediction Challenge

Discussion in 'Journals' started by expiated, Jun 9, 2018.

  1. expiated

    expiated

    From the perspective of the time zone here in Los Angeles, possibly the best time to begin trading is around 11:00 p.m. (the London session), though sometimes the volatility/liquidity begins picking up as early as 10:00 p.m. or 10:30 p.m.

    Usually, I can count on the most movement during the New York session from 7:00 a.m. to 11:00 a.m. (when things usually begin slowing down) but the action during these hours is a lot more hectic than during the European session. Also, things sometimes begin to pick up as early as 5:00 a.m., or remain active until the closing bell at 1:30 p.m. and every once in a while beyond that.

    On the other hand, it's not uncommon for things to be pretty quiet and uneventful from 1:00 p.m. to 10:00 p.m., which I believe corresponds with the Tokyo session. Every once in a while, things pick up around my dinnertime at 6:00 p.m. (is this when the market opens in Sydney?), but when this happens, I think things usually die down again from around 8:00 p.m. to 10:30 p.m.

    The London session is often kind of dead by 2:00 a.m., but the markets often reverse direction during the hours between 3:00 a.m. to 5:00 a.m. (setting up for the New York session?) so if I fall asleep and my take-profit targets are not hit by these early morning hours, it can often get me in trouble.

    As for the best stop loss and take-profit settings, I now have a set of statistically generated indicators that plot these levels for me automatically based on historical data.

    And as for how much time typically transpires before my take-profit targets are hit, there is no norm. It might happen in as little as twenty minutes, or take as long as two days.
     
    #81     Nov 9, 2018
  2. expiated

    expiated

    A number of currency pairs began reversing their day-to-day trends today, stopping me out of two of my positions and forcing me to abandon a third. I nonetheless managed to come out ahead anyway, except that I still remain in two trades (long USDJPY and short NZDJPY).

    ScreenHunter_2525 Nov. 09 00.00.jpg
     
    Last edited: Nov 9, 2018
    #82     Nov 9, 2018
  3. expiated

    expiated

    I just shorted EURGBP from around 0.8714. This trade is premature in that I do not yet have permission from my system to execute it. But it's time to go to bed and I do not wish to stay up waiting for the signal.

    EURGBPM30.png

    My justification for going ahead and sending the order include:
    1. All but my shortest-term trendlines are bearish.
    2. The rate has pulled back so that it is currently just below my "automated" stop loss (i.e., statistical resistance) at 0.8716, which means the odds of it climbing much higher, or possibly any higher, are hypothetically relatively small, so that the chances of my being stopped out before the pair ultimately hits my take-profit target at 0.8682 (if I give myself 30-pips worth of breathing room) is almost nonexistent.
    3. The asset has been falling ever since October 31st and has yet to evidence even the slightest hint of letting up.
     
    #83     Nov 9, 2018
  4. expiated

    expiated

    I got out of this trade with a slight profit ($67.81).

    ScreenHunter_2527 Nov. 09 08.10.jpg
     
    #84     Nov 9, 2018
  5. expiated

    expiated

    When I got up this morning NZDJPY was headed in the wrong direction, so I baled with a -$108.03 loss. The market makers appear to be forcing USDJPY in the "wrong" direction as well, as I believe they often do on Fridays (setting up for the next week?) but I am unwilling to give up on this trade just yet.

    ScreenHunter_2528 Nov. 09 08.20.jpg
     
    #85     Nov 9, 2018
  6. expiated

    expiated

    I noted yesterday in a post entered in my Triple-zone Strategy thread that I had settled on: (1) an extremely narrow band of moving averages to signal/confirm reversals in the general, overall, intraday trend; (2) a specific moving average for gauging the depth of pullbacks in this trend to also serve as a trigger for entering positions when rates exit such pullbacks; (3) a specific moving average for tracking the overall short-term trend; and (4) another moving average for monitoring fluctuations within the short-term trend.

    I also mentioned a set of simple moving average envelopes for defining typical price range during periods of low, medium, and high liquidity/volatility levels, as well as another envelope bordering what is usually the maximum day range.

    These parameters were generated from a four-hour chart. However, in dropping down to lower timeframes, an unexpected development led to what looks like a radical increase in accuracy, such that the “triple zones” previously found useful have been rendered moot, along with my Dynamic Probability system’s dynamic price range envelopes.

    I commented that next week’s activity should mark the end of the beginning and ought not to include any other major unknowns. Indeed, this potentially final metamorphosis of my trading methodology has led me right back to the fundamental premises on which everything began—the contention that, based on logic, it would seem (to me) that there ought to be a single moving average within a given timeframe that represents the true direction of price better than all others.

    NPP.png

    So in returning to my Numerical Price Prediction approach to forecast modeling, I am using a (brown) standard moving average to track fluctuations in the short-term trend, along with three proprietary moving averages—one to convey the general, overall direction of the (blue) intraday trend, another (red) for triggering position entries, and the fourth moving average (purple) for triggering position exits.

    Statistical support and resistance levels (typical price ranges) are defined by three moving average envelopes—envelopes that are no longer dynamic in nature, but are instead associated with (locked to) two of the aforementioned proprietary moving averages. (The typical day range [standard] envelopes remain unchanged.)

    The corresponding lower panel indicator, which I think of as a price anomaly channel, conveys these maximum levels of intraday price range fluctuation in an extremely user-friendly format.

    All of this should constitute everything I need to continue trading like I was doing on Friday, but theoretically with even greater effectiveness. But of course, time will be the judge of that.
     
    #86     Dec 8, 2018
  7. expiated

    expiated

    Because I believe that what I do works, but have not seen it being done by anyone else, I’m always keeping an eye out for something similar. And given that I like the look of IAlwaysWin’s charts, share his contention that it truly is possible to attain a daily success rate north of 90%, and had some extra time on my hands, I was curious to compare one of the images he posted to the setup I will be using next week.
    ScreenHunter_2763 Dec. 08 15.25.jpg
    It didn’t quite work out because my platform is not currently accessing five-minute data that far back, so I had to compare a 15-minute setup to whatever lower timeframe he was using. Nonetheless, I discovered there was only one aspect of how I’m trading that appeared to be even distantly related to his tools.

    Generally speaking, his graphics seemed much more focused in on details than the broader data/information with which I’m concerning myself, all of which I deleted from my chart since it appeared totally irrelevant.

    The parabolic SAR-like indicator on his chart seems vastly superior to the “factory equipped” version I have, which makes me wonder how parabolic SAR is generated, and whether perhaps an everyday trader could improve on the code that creates it, as I feel I have done with my proprietary moving averages. I fiddled around with the Step and Maximum parameters a bit so that the indicator at least remotely approximated the signals on the other chart, but in the end, the only thing I got out of this little experiment other than a smidgen of fun was some affirmation regarding how I monitor temporary fluctuations in the short-term trend.
     
    #87     Dec 8, 2018
  8. expiated

    expiated

    Oops! In returning from my "triple-zone" system back to my Numeric Price Prediction forecast model, I neglected to include a lower-period simple moving average whose significance was overlooked when designing my "final" setup. This led to a couple of misinterpretations of what I was looking at, hence a daily success rate of only 71%. But this was still high enough to result in a profitable day of trading.

    ScreenHunter_2776 Dec. 10 09.39.jpg
     
    #88     Dec 10, 2018
  9. IAlwaysWin

    IAlwaysWin

    This strategy has evolved into one that relies on distance from VWAP for signals. I thought I had the holy grail back then but found out that using volume profiles gives you a leg up above other retail traders. I have a 100% success rate on 5 minute binaries now :) It truly is a powerful setup.

    EVOLUTION.PNG
     
    Last edited: Dec 11, 2018
    #89     Dec 11, 2018
  10. expiated

    expiated

    That’s impressive! I looked at 5-minute binaries way back when and was unable to get even an inkling of how I could trade them successfully. They remain way outside my realm of possibility. For a while I was visiting Anne-Marie Baiynd’s live trade room and she was ALWAYS talking about VWAP, but I could never exactly figure out what the heck she was talking about or doing with it. In any case, it’s great to see someone else who views trading in as positive a light (even more so) as I do, so here’s to hoping you continue to enjoy your success!
     
    #90     Dec 11, 2018
    IAlwaysWin likes this.