Skip to content

Kicking - bull/bear determined by longer marubozu

Patterns pattern candlestick classic

A variation of the kicking pattern.

Visual Example

Kicking - bull/bear determined by longer marubozu — annotated preview mapping to core implementation

Synthetic ideal per library logic. Generated 2026-06-25 IST via docs/generate_all_previews.py (reproducible; maps to core Next<T> implementation).

Description

The Kicking - bull/bear determined by longer marubozu indicator is a technical analysis tool that a variation of the kicking pattern.

This indicator is primarily used for identifying key market conditions. It provides a robust signal that can be easily integrated into both simple strategies and more complex machine learning feature pipelines. Compared to its alternatives, it offers a distinct balance of responsiveness and stability.

Traders often combine this with other metrics to confirm signals and avoid false positives during sideways market regimes. It remains a standard tool for systematic trading models.

Focuses on the relative strength of the marubozus.

QuantWave implements this indicator via the universal Next<T> trait, guaranteeing bit-identical results between Rust streaming, Python streaming, and Polars batch (.ta() / map_batches) surfaces.

Formula / Specification

Recognition Rules (TA-Lib-compatible, CDLKICKINGBYLENGTH in quantwave-core/src/indicators/pattern.rs):

  1. Stateless candlestick pattern evaluated on OHLC windows.
  2. Returns a signed signal (+100 bullish, −100 bearish, 0 none) on the completion bar.
  3. Exact threshold geometry (body ratios, gap requirements, shadow lengths) matches the TA-Lib reference implementation wrapped via talib_cdl! in quantwave-core/src/indicators/pattern.rs.
  4. Validate against quantwave-core/tests/gold_standard/ vectors where present.

Parameters

Parameter Default Description
(none) — No tunable parameters for this detector.

Usage Examples

Streaming (Rust)

use quantwave_core::indicators::CDLKICKINGBYLENGTH;
use quantwave_core::traits::Next;

let mut det = CDLKICKINGBYLENGTH::new();
for (o, h, l, c) in &ohlcv {
    let sig = det.next((o, h, l, c));
}

Streaming (Python)

from quantwave import CDLKICKINGBYLENGTH

det = CDLKICKINGBYLENGTH()
for o, h, l, c in ohlcv:
    sig = det.next((o, h, l, c))

Polars Batch (Python)

import polars as pl
import quantwave as qw

def apply_kicking_bull_bear_determined_by_longer_marubozu(series: pl.Series) -> pl.Series:
    ind = qw.CDLKICKINGBYLENGTH(14)
    return pl.Series([ind.next(float(v)) for v in series.to_list()])

df = (
    pl.read_csv('ohlcv.csv')
    .lazy()
    .with_columns(
        pl.col("close").map_batches(apply_kicking_bull_bear_determined_by_longer_marubozu, return_dtype=pl.Float64).alias("kicking_bull_bear_determined_by_longer_marubozu")
    )
    .collect()
)

All surfaces are bit-identical via the single Next<T> implementation and proptests.

Edge Cases & Limitations

  • Requires sufficient complete OHLC bars; early bars yield no signal.
  • False positives are common in sideways markets — gate with trend or structure filters.
  • Pattern semantics follow TA-Lib body/shadow rules; literature variants may differ.
  • Signed output (+/−/0) should be consumed as events, not continuous features without encoding.
  • Combine with volume expansion or higher-timeframe confirmation for production use.
  • No look-ahead bias; signal is known only after the pattern window closes.

Boundary Behavior

Condition Behavior
Warm-up Pattern functions emit 0 (no pattern) until enough bars exist.
period > len Short series returns all zeros (no pattern detected).
NaN inputs Bars with NaN OHLC are treated as no pattern (0).
Invalid params N/A for most candlestick patterns.
Empty data Empty input returns an empty integer series.

Sources & References

Primary Source: https://www.investopedia.com/articles/active-trading/062315/using-bullish-candlestick-patterns-buy-stocks.asp

Implementation: quantwave-core/src/indicators/pattern.rs (CDLKICKINGBYLENGTH / CDLKICKINGBYLENGTH_METADATA). Pattern reference: TA-Lib CDL family via talib_cdl! in pattern.rs. Nison (1991) cited for psychology only — no duplicated boilerplate. Parity: quantwave-core/tests/gold_standard/cdlkickingbylength.json

Provenance: Standards bulk upgrade 2026-06-25 IST — see docs/DOCUMENTATION_STANDARDS.md.