OmniTrader Professional Forum
OmniTrader Pro General Discussion
Reducing Signal Noise in OmniTrader (Video)
2 replies, 318 viewings
|
OmniTrader Professional Forum
OmniTrader Pro General Discussion
Reducing Signal Noise in OmniTrader (Video)
|
Last Activity 12/19/2025 3:08 PM 2 replies, 318 viewings |
|
|
|
| Printer friendly version |
jsigalow Member  Posts: 37 Joined: 3/4/2006 Location: Poughkeepsie,N.Y.  User Profile |
I’ve recently been evaluating an external plug-in within OmniTrader that focuses on reducing signal noise rather than introducing additional indicators or visual complexity. The approach is statistical in nature and operates as a filter layer, leaving existing strategy logic unchanged while attempting to remove lower-quality signals. A short walkthrough video has been published by the developer, demonstrating how the plug-in behaves within the Strategy flow chart and on charts: https://youtu.be/qCzhSQXpOrc?si=bOF8t8nIZiq0p7vy For transparency, I was provided access to evaluate the tool. I’m sharing this here to discuss the methodology and to hear how others approach noise reduction within their own systems — whether through filters, confirmations, or trade frequency control. Interested to hear perspectives from those who have explored similar techniques. Regards Jerry | ||
| jsigalow Member Posts: 37 Joined: 3/4/2006 Location: Poughkeepsie,N.Y. User Profile |
Following on from the post I shared last week regarding the NeQsis plug-in and its use as a noise-reduction layer within OmniTrader, I wanted to share a brief follow-up that may be of interest — particularly for those running reversal-based strategies. The developer has now published a walkthrough showing how the same NeQsis filter methodology has been applied specifically to the 2008 Reversal Strategy, turning it into a fully mechanical, rule-based strategy without altering the original reversal logic itself. The focus remains on filtering trade quality rather than increasing signal frequency, with testing carried out across the S&P-100 universe. The video walks through how the filter is added within the Strategy Flow Chart and how it behaves on charts. One thing I found interesting during the evaluation was how the strategy behaved once it moved from back-tested data into forward-tested, unseen data. Rather than falling apart — which is often the risk with reversal-based systems — the trade behaviour remained structured, with fewer but more selective trades. The walkthrough touches on this transition and explains why the filter layer changes the behaviour, rather than simply presenting results. Sharing this for discussion rather than promotion, as I’m interested in how others here approach noise reduction or mechanical filtering when working with legacy reversal strategies. https://youtu.be/BKcfFVNL6Ds Regards Jerry | ||
| jsigalow Member Posts: 37 Joined: 3/4/2006 Location: Poughkeepsie,N.Y. User Profile |
Just a short follow-up focused specifically on VBX-3 with NeQsis Extreme added as a filter layer. For clarity, this is not a new strategy and not a rewrite of VBX-3. The original VBX-3 logic is left completely intact. The only change was the addition of NeQsis Extreme as a quality gate — effectively asking: “Is this pullback or stretch occurring in a condition that historically deserves risk?” What immediately stood out to me was the behaviour from Back Test to Forward Test. With the baseline VBX-3, the Forward Test already performs respectably. However, once NeQsis Extreme is added, the Forward Test doesn’t degrade — it actually improves. Trade count drops materially (as expected), average trade duration increases, and most importantly, the Forward Test hit rate and per-trade expectancy strengthen rather than weaken. On some forward segments, the hit rate is pushing into the mid-70% range, which — for a fully mechanical, end-of-day system — is getting uncomfortably close to territory we don’t often see without discretionary input. What makes this particularly interesting (at least to me) is what *isn’t* happening here: - No AI - No machine learning - No optimisation of VBX-3 parameters - No changes to the original strategy logic - No visual signal stacking This is purely indicator-level logic doing one job: removing trades that look “technically valid” but occur in statistically poor conditions. I’ve only been working with NeQsis for a couple of weeks, but the takeaway so far is fairly simple: if it can materially improve a solid, well-established plug-in like VBX-3 without touching its core logic, it raises a bigger question — what else inside OmniTrader improves once you stop asking “can I trade here?” and instead ask “should I trade here?” I’ve attached a Performance Summary screenshot above showing the Back Test vs Forward Test behaviour for reference. Sharing this as an observation rather than a conclusion. Still early days, but the BT → FT behaviour is encouraging enough to keep digging. Regards Jerry [Edited by jsigalow on 12/20/2025 9:27 AM]  Attached file : VBX-3 NeQsis.png (358KB - 113 downloads) |
|
|
|
|
Administrator
Forum Moderator |
Registered User
Unregistered User |
 E-Mail this thread to a friend |
Toggle e-mail notification |
