Trend Continuation · The Four-Engine Stack

Independent confirmations, combined.

Must-pass engines

Optional confluence

A-grade score

≥ 5 / 6

Session

Gate, not a point

Four must-pass, two optional

The core stack confirms; the confluence stack sharpens.

The four core engines are non-negotiable: price outside the Ichimoku cloud with Tenkan and Kijun ordered in trend direction, MACD histogram on the right side of zero with slope agreeing, momentum bias at strict thresholds, and ADX at 20 or better and rising. The two optional filters — the 9/21/55/233 EMA stack in perfect order, and ATR(14) slope positive — are toggles: on for fewer, cleaner signals; off for more entries with some counter-trend leakage. The split exists because must-pass conditions define the regime while confluence conditions grade it.

RSI or MFI

The momentum engine has two interchangeable hearts.

The stack's momentum read can run on RSI(14) or MFI(14) — the volume-weighted variant — and the choice is a liquidity decision, not a preference:

RSI is the baseline: purely price-based, reliable on every pair, immune to tick-volume quality.
MFI adds participation: on majors, and on crosses during London/NY hours, tick volume tracks real liquidity well enough that MFI confirms whether pushes have flow behind them.
On thin crosses and off-hours, tick volume misleads — RSI is the dependable read there.
Thresholds are strict either way: ≥ 60 for bullish bias, ≤ 40 for bearish — midline drift does not qualify as strength.

The score

Six filters, one point each. Session is a gate, not a point.

Every filter that passes contributes one point to a 0–6 confluence score plotted on every bar: 5–6 is a green, A-grade continuation regime; 3–4 is yellow, mixed; 0–2 is red, avoid. The session filter deliberately scores nothing — it gates. Being inside the London or New York window is a precondition for any signal existing at all, not evidence of continuation quality, and folding it into the score would let five real confirmations plus a clock masquerade as six.

Failure independence

Four engines were chosen for how differently they fail.

The stack's engines were selected less for individual accuracy than for uncorrelated blind spots: the conditions that fool one leave the others unimpressed. Stacking four engines that fail the same way would produce confident nonsense; stacking four that fail differently produces a veto system — which is the entire point.

Reference

Each closed trade logs its worst adverse excursion and best favorable excursion. The lab plots them into quadrants and pairs the picture with capture efficiency, giveback, and fee drag - then the EV Driver Diagnostics rank which repair (entry location, coefficient, trail logic, cost avoidance) would move net expectancy most.

Every trade plotted by pain taken (MAE) versus opportunity generated (MFE). The four quadrants diagnose entry precision and exit management at a glance.

Reference

The stack, engine by engine

Engine	Settings	Pass condition	What it reads
Ichimoku	9 / 26 / 52	Price outside cloud · Tenkan/Kijun ordered · thick cloud = strength	Structural position
MACD	12 / 26 / 9	Histogram right side of zero, slope in trend direction	Momentum acceleration
RSI or MFI	14	≥ 60 bullish · ≤ 40 bearish	Participation and flow
ADX	14	≥ 20 and rising	Trend strength building
EMA stack (optional)	9 / 21 / 55 / 233	Perfect alignment in trend direction	Multi-gear trend agreement
ATR slope (optional)	14	Expanding in trade direction	Volatility expansion

Doctrine

Connected inside MARS

This module doesn't work alone.

Connected — 01

Trend Continuation (full page)

Open system page →

Connected — 02

Agreement / Veto Reading

Open system page →

Connected — 03

ATR BE Assistant

Open system page →

Go deeper

Operator briefs on this territory.

Deep dive — 01

The stack is four engines and the score runs to six. Both numbers are correct.

Four engines answer whether. Two more answer how strongly. The score carries all six equally.

Read the full brief →

Deep dive — 02

Half the stack tests a direction of change, not a level.

Strong and fading fails. Moderate and building passes. The stack is reading the derivative.

Read the full brief →

Deep dive — 03

The continuation stack: Ichimoku, MACD, and the momentum filter.

Structure, acceleration, participation — three questions about one trend, with settings chosen for robustness over fit.

Read the full brief →

Every module ships in the complete MARS package.

One price. Eleven workbooks, three TradingView indicators, and the full manual library — $497.

Get MARS Pro — $497 Explore the Platform