CVD Approximation
Cumulative Volume Delta (CVD) estimates whether recent trading activity leaned toward buyers or sellers. BlitzPulse computes an approximation from L1 market data, then treats the result as a directional clue rather than proof.
Why It Is An Approximation
L1 data gives the best bid, best ask, last price, and size. It does not provide the full queue of resting liquidity at every price. Because of that, BlitzPulse must classify trades with transparent heuristics, such as whether a print moved up, moved down, or repeated the prior price.
This is enough to detect some useful divergences, but it is not a complete view of market microstructure.
What It Can Suggest
CVD approximation is most useful when price and estimated participation disagree:
- Price rises while CVD weakens: buying pressure may be fading.
- Price falls while CVD improves: selling pressure may be fading.
- Price breaks a range while CVD confirms: the move may have broader support.
These readings are stronger when volume profile, trend, and risk context agree.
What It Does Not Prove
CVD approximation does not identify specific institutions, hidden liquidity, or future order intent. It can misclassify prints during fast markets, wide spreads, auction periods, or delayed data windows.
How BlitzPulse Uses It
BlitzPulse combines approximate CVD with:
- price structure
- volume profile levels
- unusually large prints
- symbol-specific thresholds
- data freshness checks
The output is a confluence input. It is not a trading instruction.
Practical Guardrails
- Treat divergence as a reason to investigate, not as a command to trade.
- Be more cautious around market open, market close, and major news.
- Prefer repeated confirmation over one isolated reading.