The fall of Outperforming Long Bots

Forums › ProRealTime English forum › General trading discussions › The fall of Outperforming Long Bots

  • This topic has 51 replies, 8 voices, and was last updated 8 hours ago by avatarJS.
Viewing 15 posts - 31 through 45 (of 52 total)
1 2 3 4
  • 04/27/2025 at 9:04 AM #246451
    LucasBest

    But, do you look for the Dominant cycle Period, as Ehlers does ? Or, as i guess, you test 9 custom cycle periods ?

    04/27/2025 at 9:10 AM #246452
    JS

    As mentioned, I stay in the “time domain”, meaning no “cycles (=frequency)” in my system…

    04/27/2025 at 9:34 AM #246453
    LucasBest

    If i am not wrong, Cycles are also a time related… Cycle depend on time period (inverse of frequency), has a phase (moment of the cycle) and an amplitude (energy, or DB for sounds, or price moves for charts). If you use EMA or any other recursive filter as a pass filter to analyze price, you are automatically using cycles ; or maybe i missed something ? 🙂

    04/27/2025 at 10:16 AM #246454
    JS

    Of course, it’s not forbidden to look at both the time domain and the frequency domain, but the key question remains: in which domain is the relevant information found?
    Take, for example, the simplest case — using a “moving average” as a filter…
    A “moving average” is an effective filter, in the time domain, for filtering out “random noise”, but it is a poor filter, in the frequency domain, for distinguishing one band of frequencies from others…
    Thus, this filter performs best in the time domain and worst in the frequency domain, making it logical to carry out the analysis in the time domain in this case…
    So the real question is: where is the relevant information — in the time domain or the frequency domain?
    There is no general consensus on the answer to this question… 🙂

    04/27/2025 at 10:45 AM #246455
    LucasBest

    Those who look for the frequency domain suppose that the frequency will remain little more allowing the trader to know WHEN it may reverse ?

    Those who look for the amplitude suppose that the amplitude will remain stable allowing the trader to know at which price it may reverse ?

    I am getting lost… 🙂

    04/27/2025 at 11:03 AM #246456
    JS

    Correct, in the frequency domain, people look at the movement of price in relation to a composite dominant wave and expect the price to continue following this dominant wave…
    In the time domain, the focus is on the price movement (amplitude) over time, and deviations in amplitude are seen as signs of changes over time…

    04/27/2025 at 11:19 AM #246458
    LucasBest
    JS wrote:

    In the time domain, the focus is on the price movement (amplitude) over time, and deviations in amplitude are seen as signs of changes over time…

    Isn’t it the purpose of RSI ?

    04/27/2025 at 11:26 AM #246459
    JS

    Correct, RSI and most other indicators “look” at changes in amplitude within the time domain…

    04/27/2025 at 11:43 AM #246460
    JS

    The problem with these kinds of indicators is that it is unclear which data set they should be used on and over which time period.
    This is also my biggest objection to these types of indicators — everything is too “general”.
    You cannot apply them reliably across all data sets, all timeframes, or during every period; everything is so poorly defined…

    1 user thanked author for this post.
    avatar coincatcha
    04/27/2025 at 11:57 AM #246461
    phoentzs

    Interesting discussion. But what’s wrong with a simple open-range breakout system with a trail to capture profits? Something like that has brought me very good profits over the past few months.

    1 user thanked author for this post.
    avatar coincatcha
    04/27/2025 at 12:02 PM #246462
    JS

    There is nothing wrong with that, and I am happy for you when it brings good profits…

    04/27/2025 at 12:07 PM #246463
    coincatcha

    ^^^

    Exactly why it is so important to look away from the indicator baskets we are given to first explore. It’s a trap. Dissecting price and time alone should contain enough building blocks to pursue an edge.

    This way the factory can be fed raw ingredients for quality control.

    There is enough variables in that alone to keep you busy and productive.

    Anytime an indicator is added as a filter (often rightly so), the frequency reduces leaving most bots in the slow bot basket. Sure there’s a viable edge, but scaling it in any reasonable amount of time becomes a journey.

    I’m sure this is why these fast bots get the most attention and why we are always looking.

    04/27/2025 at 12:14 PM #246464
    coincatcha

    JS does that mean your fast bot in its purist form has merit or works on other TF’s?

    04/27/2025 at 12:27 PM #246465
    JS

    During the testing of my system, it turned out that it performed best on the Dow Jones dataset. Using other datasets is possible, but resulted in poorer outcomes. This is entirely due to the characteristics of the different datasets.
    The 1-minute timeframe proved to tolerate little variation.

    1 user thanked author for this post.
    avatar coincatcha
    04/27/2025 at 12:35 PM #246466
    LucasBest

    Are you using moving average? or exponential moving average?

  • Author
    Posts
Viewing 15 posts - 31 through 45 (of 52 total)
1 2 3 4

Find exclusive trading pro-tools on ProRealCode Marketplace

Similar topics:
1 short and 1 long per day max
6 jan 24 prt not available – too long to respond
anybody else almost scared to go long?
bb trading one long one short on dax 9-17,30
best indicators/strategies to exit from a long position
buy a long every week if the high is higher than the last high
code to exit position after 30 bars if long still in profit
combining short/long strategies into 1 strategy issue
hiding expired and unwanted bots in automatic trading
long candle trading 5 min chart dax future

Create your free account now and post your request to benefit from the help of the community
Register or Login