What Actually Happened With Kjan Obit
The trading analytics space has been buzzing for months about Kjan Obit, and honestly, most of the noise comes from people who have never actually looked at the data. I've spent the better part of three years tracking methodology shifts across retail and institutional analysis circles, and the Kjan Obit thing is different from the usual influencer hype cycle. It's not because it's perfect. It's because the core framework actually holds up under scrutiny, which is rarer than you'd think. Kjan Obit's approach centers on a hybrid model that blends order flow imbalance metrics with regime detection, rather than relying on any single indicator. The basic mechanic is straightforward: you're looking at cumulative volume delta across time slices, then cross-referencing that against volatility regimes identified through GARCH-based clustering. Most traders skip the second half and just chase delta signals blindly. That's why they lose money. I first encountered this framework around mid-2023 when a former prop desk colleague started referencing the methodology in our Discord channels. The public materials were sparse at first — mostly chart annotations and occasional threads on what he called "structural break confirmation." People assumed it was just another repackaged ICT concept. It wasn't. The difference is in the execution layer, not the theory.
Here's how you actually run it. You need a data feed that gives you tick-level or at minimum one-second aggregated volume and trade direction. You can't do this on daily candles. Once you have the feed, you're calculating rolling cumulative delta over a 15-minute window, then tagging each segment as buyer-initiated or seller-initiated. The next step is identifying the current volatility regime — low, normal, or elevated — using a rolling 30-day realized volatility percentile. You only take delta divergence signals during low or normal regimes. Elevated regimes filter out most false signals, but they also suppress legitimate entries, which is by design. I ran into a specific edge case that almost made me write the whole thing off. During the October 2024 volatility spike in ES futures, the model generated consistent seller-side divergence across multiple sessions, but price wasn't reacting the way the backtests suggested it would. The workaround wasn't to adjust the parameters — it was to add a liquidity void filter. When the book shows thin liquidity on the passive side, delta signals get absorbed without price displacement. I started cross-referencing the delta divergence with footprint heatmaps showing absorption patterns, and the signal quality improved dramatically. That's something the published materials don't cover well because it requires real-time order book access, which most retail traders don't have. The counter-intuitive part that most beginners miss is that the model actually performs worse during major news events, even though that's when most people want to use it. High-impact news creates order flow that doesn't reflect genuine market sentiment — it reflects positioning adjustments and stop runs. The regime detection catches some of this, but not enough. I've found that manually suspending the model 30 minutes before and after major releases, like FOMC decisions or CPI prints, saves you from the worst whipsaws. The model isn't broken during those windows. It's doing exactly what it's supposed to. You just shouldn't be trading during them.
Another nuance that doesn't get discussed enough: the lookback period matters more than anyone admits. The standard 15-minute window works for intraday, but if you're applying this to daily or weekly charts, you need to extend the delta calculation to at least 60 minutes per bar, otherwise you're just measuring noise. I've seen people apply the same parameters across timeframes and then complain the signals don't translate. They don't, because the noise profile changes completely. As for getting access to the actual framework, Kjan Obit hasn't released a full public model or downloadable tool yet. What exists are analytical threads, partial visual breakdowns, and a small community around the concept. There are third-party implementations floating around on GitHub — mostly Python scripts that approximate the delta calculation and regime detection — but none of them are official. If you find a "Kjan Obit indicator download" link, it's not from the source. Be careful with those. The honest downsides are worth stating plainly. The model requires real-time data, which means you're paying for a proper exchange feed or a third-party provider like Rithmic or CQG. Free platforms won't cut it. The learning curve is steeper than a typical indicator because you're not plugging numbers into a prebuilt script — you're understanding the interaction between delta and regime, which takes actual screen time. And during low-volatility consolidation periods, you'll get very few signals. Some months might produce fewer than ten legitimate setups across all tracked instruments. If you're expecting constant action, this isn't for you.
Get the Full Details
For people who want to experiment without building from scratch, the closest thing to a working implementation right now is a set of custom Pine Script indicators for TradingView that replicate the cumulative delta and regime detection logic. They're not complete, and they're limited by TradingView's data constraints, but they're a starting point. Search for "cumulative delta regime" along with the analyst's name, and you'll find a few community-contributed versions. None of them match the full proprietary setup, but they'll get you close enough to understand whether the methodology fits your workflow. The framework itself is sound. The problem is that everyone wants the shortcut, and shortcuts don't exist here. You either put in the time to understand the mechanics or you don't, and the market doesn't care which.
