🌊 Undercurrent Cadence

E001 — NVI Regime Filter + DPO Cadence Cross

Timeframe: H1 Best fit: Crypto / Indices / US-HK Stocks Sim: ~1.5 signals/day

Concept

Most breakout-chasing systems react to the volume spike everyone already sees. Undercurrent Cadence flips that logic: it tracks a "quiet money" regime line that only updates on low-volume bars — the Negative Volume Index (NVI) — under the theory that large participants tend to move size without spiking volume. Once that undercurrent clearly leans one direction, the strategy waits for a de-trended cycle line (DPO) to tick back through zero as the actual timing trigger.

1.5x ATR

Stop Loss

2.5x ATR

Take Profit

Risk / Trade

10 / day

Max Entries

Signal Rules

Long Side

Undercurrent Line (NVI) sits above its own 14-period EMA — bullish undercurrent bias
Cadence Line (DPO, 10-period) crosses above zero, confirming the same direction as the undercurrent bias
ATR(14) is at or above 50% of its own 50-bar average (volatility floor)

Short Side

Undercurrent Line (NVI) sits below its own 14-period EMA — bearish undercurrent bias
Cadence Line (DPO, 10-period) crosses below zero, confirming the same direction as the undercurrent bias
ATR(14) is at or above 50% of its own 50-bar average (volatility floor)

Parameters

Parameter	Default	Description
NVI_EMA_Period	14	Smoothing period for the Undercurrent (NVI) regime line
DPO_Period	10	Cadence Line lookback period (internal offset = 6 bars)
ATR_Period	14	ATR period used for both risk sizing and the volatility gate
ATR_AvgPeriod	50	Rolling average period for the ATR baseline
ATR_RatioMin	0.5	Minimum ATR-to-average ratio required to trade
SL_ATR_Mult	1.5	Stop-loss distance in ATR multiples
TP_ATR_Mult	2.5	Take-profit distance in ATR multiples (~1.67:1 reward-to-risk)
MaxTradesPerDay	10	Daily entry cap, resets at midnight server time
NoSignalAlertDays	3	Watchdog alert threshold for consecutive no-trade days

Frequency Test

Entry logic was replayed in Python on synthetic H1 data (~250 trading days, regime-switching random walk). With default parameters the strategy produced an average of ~1.5 signals per day, comfortably within the 1–10/day target range, with a roughly balanced long/short split.

Risk Disclosure: This page describes a rules-based framework for research and educational purposes only. It is not financial advice and does not guarantee any outcome. Performance on historical or simulated data does not guarantee future results. Always backtest and forward-test on your own platform, use proper position sizing, and only risk capital you can afford to lose. Markets can and do move against any strategy.