Dual-System Forex Trading Strategy: Mean Reversion + Trend Following for a Smoother Equity Curve

Philosophy and objectives

The core idea: one edge is not enough

Markets don’t behave the same way all the time. Sometimes they mean-revert (stretch → snapback). Other times they trend (stretch → keep stretching). Most traders blow up because they try to force one behaviour onto all regimes.

Disclaimer:
This guide is for educational and informational purposes only. It is not financial advice. Trading leveraged products involves significant risk and you can lose more than your initial deposit. Past performance is not indicative of future results. I accept full responsibility for my own decisions, risk, and outcomes.

My philosophy

I run two behaviours, not two “ideas”: mean reversion (core) + trend capture (hedge).
I protect the equity curve with filters + exposure limits + risk caps.
I don’t chase constant returns; I chase survival + consistency.

So we run two systems that thrive in opposite conditions:

KnoxMagnet = the core engine
- counter-trend / pullback / “magnet” mean reversion
- aims for regularity, clean exits, and a smoother curve
- enhanced by a controlled “win-streak reinvestment” ladder
TrendCapture = the safety net
- trend-following breakout / runner capture
- expects false breakouts (small losses) to catch the big move
- kept on fixed risk, no “add to winners” (because trends are lumpy)

This is a portfolio, not a single trick.

The style: “process beats prediction”

We do not need to guess direction. We need:

a defined setup,
defined risk,
defined exit logic,
and consistency.

Your edge comes from how you behave:

you only trade when conditions align,
you avoid the worst environments (strong trends for counter-trend),
you let probability play out across many trades.

The equity curve goal: smooth doesn’t mean flat

You want a “nice smooth curve” because it’s psychologically and professionally attractive.

A smooth curve comes from:

tight risk boundaries (risk caps, max trades, spread filters),
not stacking correlated exposure,
taking fewer but higher-quality trades,
accepting that some months will be quieter.

Aiming for 4–6% per month can be a motivating target, but the “smooth” version of that goal looks like:

strong months,
average months,
and a small number of weak months,
with drawdowns that are controlled and recoverable.

The job is not “never draw down”. The job is “never blow up”.

Why your staking ladder makes sense (and where it can bite)

Your revised staking approach is a controlled anti-martingale:

Base risk starts at 1%
If the pair is “hot” (recent wins), you reinvest a portion of profits
But you deliberately reduce the add-on on the second win (50% → 25%)
And after 3 wins you reset to base, stopping runaway risk

That’s a very sensible structure because it:

captures momentum in your edge without turning into uncontrolled compounding,
naturally avoids “chasing” after a loss (you reset),
prevents long streaks from creating stupid position sizes (you reset after 3 wins).

Where it can still bite:

if you allow too many simultaneous positions across correlated pairs,
if spreads blow out (entries become worse than tested),
if you ignore hard caps.

So: the staking ladder is a performance enhancer, not the foundation.
The foundation is still:

filters,
stop discipline,
and portfolio limits.

The “set & forget” philosophy (remote-friendly)

Set & forget works only if you define what “forget” means.

You’re not abandoning the system — you’re moving your effort to:

design
risk boundaries
review and adjustments
not watching every candle.

Remote management rules:

you intervene only for errors, risk limits, or exception events.
you do not intervene based on emotion or “gut feel”.

The hierarchy: what matters most

If you remember nothing else, remember this order:

Survival (risk cap, max trades, no margin stress)
Consistency (same rules, same conditions, same execution)
Quality of setups (ADX/volatility filters do the heavy lifting)
Edge extraction (targets/trailing, partials)
Performance enhancement (staking ladder)

Most people reverse that order and pay the price.

The rules you will not break

These are “non-negotiables” if you want smoothness:

Hard stop loss always (no “I’ll just hold it”)
Spread filter always (don’t enter when costs are distorted)
Max open trades always (overexposure kills smooth curves)
Risk cap always (prevents a single event from dominating the month)
One trade per symbol (keeps behaviour clean and trackable)

What success looks like (measurable)

You’re “on track” when:

you are executing the plan without exceptions,
drawdowns are within your defined bounds,
you can leave the system unattended without fear,
weekly review shows no creeping rule-breaking.

You’re off track when:

you change inputs constantly,
you widen stops emotionally,
you increase exposure because of boredom,
you override the system repeatedly.

How the two systems work together

The barbell portfolio: steady mean reversion + trend insurance

The simplest way to think about it

I run two systems because I’m solving two different problems:

KnoxMagnet makes money when price stretches and then reverts back toward “fair value” and key levels (pivots/magnets).
TrendCapture makes money when price breaks out and then keeps going (the very environment that can punish counter-trend logic).

So my portfolio is designed to do this:

In normal markets: KnoxMagnet does most of the work.
In strong directional markets: TrendCapture helps offset the periods when KnoxMagnet is naturally less comfortable.

It’s not about having “two EAs.” It’s about having two behaviours that respond differently to the same market.

Why this combination can smooth my curve

A mean-reversion system’s worst periods usually look like:

price trends hard,
pullbacks are shallow,
RSI stays extreme,
divergence prints multiple times,
magnets get ignored until a late, violent snapback (if it comes).

A breakout / trend-following system’s best periods usually look like:

price trends hard,
breaks levels and keeps breaking,
volatility expands,
pullbacks are shallow and trend resumes,
runners become the outsized winners.

Those are the same environments.

So if I keep TrendCapture small and disciplined, it can act like an “equity curve stabiliser” across regime changes.

My intent (and why it’s right)

I run KnoxMagnet as my core and TrendCapture as my safety net.

That tells me exactly how I size and manage them:

KnoxMagnet gets the attention and the upgrades (my staking ladder).
TrendCapture stays simple, fixed risk, and patient.

This keeps me out of overengineering.

A practical framework: my “traffic light” regime lens

This is the quick mental model I use so I stay consistent without overthinking:

✅ Green market (best for KnoxMagnet)

ADX is low or falling
price is oscillating / ranging
pivots/magnets get respected
snapbacks happen regularly

How I behave:
KnoxMagnet does most of the work. TrendCapture may be quiet.

⚠️ Amber market (mixed)

ADX rising but not extreme
volatility expanding
some reversions work, some don’t
false breakouts occur

How I behave:
KnoxMagnet still trades but my filters matter. TrendCapture may start to appear.

🔴 Red market (hard trend / freight train)

ADX high and rising
repeated one-way moves
“oversold stays oversold / overbought stays overbought”
magnets are skipped for longer than usual

How I behave:
KnoxMagnet must be protected by strict ADX/spread/exposure rules.
TrendCapture has the best chance of catching runners.

This is why I built the hedge in the first place.

Pair split: what I run on what

KnoxMagnet (my core 6)

These are the first six pairs I run for the mean-reversion engine:

AUDNZD
EURCAD
NZDCAD
USDCHF
USDCAD
AUDCHF

Why: they generally “breathe” more cleanly and mean-revert more reliably than the wilder GBP crosses, and the costs/spreads are usually more manageable.

TrendCapture (my hedge set)

I keep this smaller, more selective, and more liquid:

EURUSD
GBPUSD
AUDUSD
USDCAD (shared pair is fine if I cap total risk)

I can expand later, but I don’t need to at the start.

Portfolio rules that keep my curve smooth

This is where I win or lose the “nice curve” goal.

Rule 1: I limit simultaneous exposure

Even good systems look ugly if I stack trades.

My default:

KnoxMagnet: MaxOpenTradesAllSymbols = 1–2
TrendCapture: MaxOpenTradesAll = 1–2

If both are running, I’m effectively running a portfolio — so I keep total combined open trades tight.

Rule 2: I keep one trade per symbol

I keep:

OneTradePerSymbol = true

This prevents “spiral exposure” where a pair catches me twice in the same regime.

Rule 3: I use different Magic Numbers

This keeps trade management clean and avoids cross-system confusion.

KnoxMagnet: MagicNumber A
TrendCapture: MagicNumber B

Non-negotiable.

Rule 4: I cap risk where it matters

My win-streak ladder builds momentum. Good.
But smoothness comes from ceilings, not accelerators.

For investor-grade behaviour:

KnoxMagnet RiskCapPercent: 2.0–2.5%
TrendCapture risk per trade: 0.25–0.5%
TrendCapture portfolio heat cap: 1–1.5%

This way TrendCapture can help in trend regimes without making my equity curve noisy.

My “priority” rule when both want to trade

Sometimes both systems want a trade around the same time. I use a simple tie-breaker:

I respect portfolio caps first (max trades / heat cap / spread limits).
If I’m at the limit, TrendCapture gets priority only in a clear Red market (strong trend regime).
Otherwise, KnoxMagnet gets priority because it’s my primary engine.

Keeping it simple stops me overtrading.

The behaviour I’m aiming for over time

This portfolio is built to behave like this:
KnoxMagnet provides frequency + regularity
TrendCapture provides convexity (occasional big winners)
Risk caps provide survivability + smoothness
Filters provide discipline under stress
If those four things are true, my “steady curve” goal becomes realistic.

KnoxMagnet system rules

My core mean-reversion engine: stretch → reversal → magnet target

What KnoxMagnet is trying to do

KnoxMagnet is my “regularity” system. I’m not trying to catch the start of a new trend. I’m trying to catch the snapback after price has stretched too far, too fast, and the move begins to lose momentum.

My edge comes from combining:

stretch (RSI(5) across multiple timeframes),
momentum loss / reversal evidence (Knoxville divergence),
environment control (ADX filter so I don’t fight freight trains),
and a realistic target (missed pivots/magnets).

The most important part is this:
I don’t need a lot of trades. I need the right trades, in the right conditions.

The setup in one sentence

On H1, I trade reversals only when RSI is stretched across timeframes, Knoxville divergence confirms, ADX says the market isn’t in full trend mode, and there’s a clear magnet target with enough room to justify the trade.

My “must-have” entry conditions

KnoxMagnet only enters when all of these line up:

1) Timeframe and timing

I run it on H1.
Signals are evaluated on the closed candle (no guessing mid-candle).
That keeps execution consistent and stops me from reacting to noise.

2) Stretch filter: RSI(5) across multiple timeframes

I want the market to be genuinely stretched, not just “a bit extended”.

RSI(5) is checked on: M1, M5, M15, M30, H1
I require at least 4 out of 5 to agree.

Typical thresholds:

Oversold: RSI ≤ 30
Overbought: RSI ≥ 70

Buy bias when the stretch says oversold.
Sell bias when the stretch says overbought.

This part ensures I’m trading a real stretch, not a random wiggle.

3) Regime control: ADX filter (H1)

This is one of my biggest drawdown reducers.

I only take counter-trend entries when ADX(14) is below my threshold (typically around 22–25).
If I’m in a period where trends have been brutal, I tighten it (closer to 20–22).

The logic is simple:

When ADX is high and rising, trend continuation is more likely than snapback.
My job is to avoid that environment.

4) Trigger / confirmation: Knoxville Divergence

Stretch alone isn’t enough. I want evidence that momentum is fading and reversal pressure is building.

A valid Knoxville signal is:

any of the divergence buffers printing on the closed candle (depending on how strict I set it)

I can optionally require:

the stronger “alt” divergence lines only (more selective),
and/or the reversal arrow confirmation.

This gives me a “stacked” entry:
stretch + reversal evidence rather than “RSI said so”.

My “no trade” conditions (these protect my curve)

Even if I see stretch + divergence, I don’t want to take trades in bad operating conditions.

I skip trades when:

1) Spread is too high

If the spread is inflated, the edge gets eaten before I even start.

I keep a MaxSpread filter on.
I’m especially strict around rollover.

2) There’s no room to a sensible magnet target

I don’t take trades where the nearest target is too close.

I require my first target (TP1) to offer at least 2R relative to my stop distance.
If it doesn’t, I skip.

This avoids “good signals” that have no runway.

3) I’m already exposed

I keep:

one trade per symbol,
and a low max open trades limit across symbols.

This stops my equity curve getting jagged.

Stop loss (how I define “I’m wrong”)

I always use a stop. No exceptions.

My preferred default is ATR-based because it adapts to volatility:

SL = 2 × ATR(14) from entry (H1 ATR)

If I want something more structural, I can use a swing-based stop:

below/above the last swing over a lookback window (plus a buffer)

The point isn’t “tight”. The point is invalidated.

Targets and trade management (the magnet logic)

KnoxMagnet is built around the idea that old pivots behave like magnets.

TP1: the nearest missed pivot magnet (D/W/M)

The EA scans:

daily, weekly, monthly pivots,
identifies levels that weren’t touched when “they should have been”,
then checks whether price has still not returned to them,
and uses the nearest valid one in the direction of my trade as TP1.

If no missed pivot is found, it falls back to the nearest current pivot level.

Runner: optional

I like using a runner because sometimes the snapback turns into a full reversal.

So I split the position:

TP1 leg closes at the magnet
Runner leg trails (ATR trailing) and tries to catch the bigger move

This gives me two benefits:

regular, banked profits,
and occasional outsized wins without needing prediction.

The “set & forget” operating rules I follow

When it’s running live, I don’t micro-manage.

I intervene only if:

there’s a technical issue (platform, file error, trade execution errors),
a spread anomaly is causing repeated skips or poor fills,
or I decide to tighten/loosen parameters after a planned review.

I don’t intervene because a candle looks scary.

The one thing I watch like a hawk

Counter-trend systems live or die by environment control.

So the single most important dial for KnoxMagnet is:

ADX strictness (and secondarily: exposure limits).

If I’m seeing too many trades that “should revert but don’t,” I don’t blame RSI or divergence first.
I tighten the trend filter and reduce simultaneous exposure.

KnoxMagnet staking ladder

How I build momentum on winning pairs without letting risk run away

Why I use a staking ladder at all

My KnoxMagnet edge tends to show up in clusters. When the market is behaving in a mean-reverting way, I often get:

clean stretches,
clean snapbacks,
pivots/magnets getting hit regularly.

That’s the environment where it makes sense to press slightly harder — not by gambling, but by reinvesting a portion of profits while conditions are favourable.

The goal of my staking ladder is:

build momentum on a “hot” pair,
while still resetting quickly when the environment shifts.

The principle

I never scale after a loss.
I only scale after a win, and only for a short streak.

That means I’m never “chasing” recovery. I’m compounding success.

My exact staking rules (per pair)

Everything below is pair-specific. AUDNZD has its own streak, EURCAD has its own streak, etc.

Base rule (always on)

Every new trade starts from a base risk of 1% of equity.

Win 1 → add 50% of the previous profit

If my previous trade on that same pair was a win, then the next trade risk becomes:

1% of equity + 50% of the previous trade’s profit

This is the “momentum press.” I’m using house money, not increasing my baseline.

Win 2 → add 25% of the previous profit

If I then get a second win in a row on the same pair, then the next trade risk becomes:

1% of equity + 25% of the previous trade’s profit

This is deliberate: I press less on the second win than the first.
It keeps the curve smoother and stops the ladder from getting too aggressive.

After 3 wins in a row → reset to base risk

If I’ve had three wins in a row on the same pair, the next trade goes back to:

1% of equity

This is my “take the money and reset” rule.

It prevents the classic compounding trap where a hot streak makes position sizes grow right before the market flips regime.

Any loss → reset immediately

If a trade is a loss, the next trade on that pair goes straight back to:

1% of equity

No exceptions.

This keeps losses contained and stops me from trying to force the next trade to “make it back.”

What counts as “one trade” (important)

Because KnoxMagnet can place two orders per entry (TP1 + Runner), I treat that as one combined trade event.

So:

the win/loss outcome is based on the combined result of both legs,
not each order separately.

That keeps the streak logic honest.

My safety rails (what stops this becoming reckless)

The ladder gives me a growth boost, but the smoothness comes from my guardrails.

1) Risk cap (non-negotiable)

I cap total risk per new trade at a fixed % of equity.

My default:

RiskCapPercent = 2.0–2.5% for smoothness
3.0% only if I’m deliberately running it more aggressively

This cap is the difference between:

“enhanced compounding”
and
“one loss wipes a month.”

2) Exposure limits (protect the curve)

Even a good ladder turns ugly if I run too many trades at once.

My default:

MaxOpenTradesAllSymbols = 1–2
OneTradePerSymbol = true

3) I don’t let the ladder override quality

The ladder never forces trades. It only changes stake size after a valid signal.

If signals dry up, it does nothing.
That’s a feature, not a bug.

What this ladder is designed to achieve

This is my intended behaviour:

In a favourable mean-reversion regime, I get a handful of good trades and the ladder gives me an extra push.
When the market shifts, a loss resets risk immediately.
After a strong run (3 wins), I reset anyway to avoid “giving it back.”

It’s momentum funding with built-in brakes.

The reality check (how I stay grounded)

I like the ambition of 4–6% per month, but I don’t measure success by “constant monthly targets.”

I measure success by:

keeping drawdowns controlled,
keeping position size sane,
executing rules consistently,
and letting the positive months stack up.

Smooth equity curves come from discipline, not from forcing returns.

My operating checklist for the ladder

Before I run this live, I confirm:

RiskCap is set (2–3%)
Max trades are set (1–2)
Spread filter is strict
ADX filter is strict enough
I’m running on the right pairs (my core 6)

If those are correct, I can let it run.

TrendCapture system rules

My trend safety net: breakouts + runners, kept simple and fixed risk

What TrendCapture is trying to do

TrendCapture exists for one reason:

When markets stop mean-reverting and start trending hard, my counter-trend system can have a rough patch. TrendCapture is my insurance policy for those environments.

It’s designed to:

accept that breakouts often fail,
keep losses controlled and repeatable,
and stay in the game long enough to catch the occasional big trend runner.

This system is not meant to feel “busy” or “clever.”
It’s meant to be patient, mechanical, and occasionally spectacular.

The mindset I need for trend systems

Trend systems only work if I accept three truths:

False breakouts are part of the edge
I will take small losses. That’s not failure — that’s the entry fee.
The big winners pay for everything
One runner can cover multiple losses and still leave profit.
If I add to winning streaks, I usually increase chaos
Trend outcomes are lumpy. I want stable sizing, not compounding spikes.

So: TrendCapture stays on fixed risk. Always.

My setup in one sentence

On H4, I trade breakouts only in the direction of the dominant trend, only when volatility is expanding, and I manage the trade with an ATR stop, a partial at +2R, and a trailing runner.

My “must-have” entry conditions

TrendCapture only enters when all of these line up:

1) Timeframe and timing

I run it on H4.
It evaluates signals on the closed candle.
It executes on the new bar so the decision is always consistent.

2) Trend filter: EMA(200) + slope

I only want breakouts that align with the bigger picture.

Longs only if:
- price is above EMA(200)
- EMA(200) is sloping upward
Shorts only if:
- price is below EMA(200)
- EMA(200) is sloping downward

This stops me taking “breakouts” that are really just noise inside a larger downtrend/uptrend.

3) Breakout trigger: Donchian channel breakout

I use a Donchian channel breakout because it’s clean and objective.

Long trigger: close breaks above the recent Donchian high
Short trigger: close breaks below the recent Donchian low

To reduce weak pokes, I add an ATR-based breakout buffer:

the breakout must clear the level by a small fraction of ATR

This keeps entries slightly more selective without turning the system into a curve-fit monster.

4) Volatility expansion filter

Trend moves tend to happen when volatility expands.

So I require:

ATR now > ATR average (SMA of ATR) by a small margin

If volatility isn’t alive, I’d rather stay out than get chopped up.

5) Optional: ADX rising filter

If I want even cleaner trades, I require:

ADX is above a minimum level
and rising

This pushes TrendCapture toward the environments where trends are actually developing.

My “no trade” conditions (these matter unattended)

1) Spread too high

Breakout systems get destroyed by poor execution.

So I skip entries when:

spread is above my threshold
especially around rollover

2) Rollover window

I block entries around rollover because:

spreads spike
fills worsen
and “breakouts” can appear purely due to pricing distortions

Even if the setup is valid, the execution quality is often not.

3) I’m already at my portfolio limits

TrendCapture is not allowed to dominate my risk.

If I’m already at:

max open trades, or
portfolio heat cap,

I skip.

That’s the whole point of using it as a safety net — it helps without taking over.

Stop loss and trade structure (how I keep losses repeatable)

Initial stop loss: ATR-based

I set the SL using ATR so it adapts to each pair’s volatility:

SL = entry ± (ATR × multiplier)

This makes the system consistent across EURUSD and something like AUDUSD without me manually adjusting stops.

Two-leg management: why I do it

Trend systems have a psychological problem:

you can be right, but still give back a lot before the trend ends.

So I use two legs:

TP leg: closes at +2R
- this pays for the trade quickly and calms the equity curve
Runner leg: no TP, just trails
- this is the part that catches the “10x+” style move

This combination keeps the system survivable while preserving the upside.

Trailing stop: Chandelier / ATR style

The runner trails behind price using:

a volatility-aware stop that tightens as the move progresses

This lets me stay in trends longer without babysitting.

Position sizing (fixed, boring, deliberate)

TrendCapture stays on fixed risk because that’s what keeps it stable and useful as a hedge.

My typical sizing approach:

0.25%–0.5% risk per trade
strict portfolio heat cap (so multiple false breakouts don’t stack up)

This way TrendCapture can do its job without turning my equity curve into a heart monitor.

What I expect from TrendCapture

I don’t expect it to win often.
I expect it to:

lose small, and lose predictably
win big when it’s right
be quiet for long periods
occasionally deliver a month that makes the system worth having

This is how it balances KnoxMagnet.

Portfolio rules

The “accounting” that keeps the equity curve smooth

This section is where most traders quietly lose the plot. The systems can be good, but the portfolio behaviour can still be messy if I let exposure stack up.

So I treat portfolio rules like seatbelts:

they don’t make the car faster,
they stop me dying when something unexpected happens.

1) I run each system with its own identity (Magic Numbers)

I always run different Magic Numbers so the systems don’t interfere with each other’s trade management and reporting.

My default:

KnoxMagnet MagicNumber = 20260120
TrendCapture MagicNumber = 20260121

This keeps everything clean:

trade counts stay accurate,
management logic doesn’t cross wires,
performance tracking is simpler.

2) I keep total open trades low (smoothness lever #1)

A smooth curve comes from controlling how many things can go wrong at once.

My defaults:

KnoxMagnet: max 1–2 open trades total
TrendCapture: max 1–2 open trades total
Combined portfolio: I aim to stay around 2–3 open trades at any time

If the account is “busy,” the equity curve gets jagged.

So I’d rather skip decent setups than let exposure creep up.

3) One trade per symbol (non-negotiable)

I keep:

OneTradePerSymbol = true

This stops the most common spiral:

one pair starts trending,
the system keeps firing signals,
and I accidentally stack multiple entries in the worst possible regime.

One trade per pair keeps behaviour predictable and reviewable.

4) I control correlation (the silent killer)

Correlation is the easiest way to “think I’m diversified” while actually betting the same idea six times.

Examples:

AUDNZD + AUDCHF + AUDUSD can all lean on AUD themes
EURCAD + EURAUD + EURUSD can cluster on EUR moves
USDCHF + USDCAD + EURUSD can stack USD exposure indirectly

So I apply a simple rule:

My correlation rule (simple and effective)

I don’t allow more than:

2 trades that share the same base currency, and
2 trades that share the same quote currency,
at the same time.

If I’m already in two AUD-related positions, I skip the third AUD setup even if it looks good.

This one habit does more for smoothness than most indicator tweaks.

5) I separate “core engine” vs “hedge” behaviour

I keep the intent clear:

KnoxMagnet is the core engine

higher priority in normal conditions
win-streak staking ladder ON
risk cap ON (2–2.5% for smoothness)
strict regime control

TrendCapture is the hedge

smaller sizing
fixed risk only (no ladder)
allowed to be quiet
allowed to be “lumpy”
exists to help in trend regimes

If I let TrendCapture grow to the same size as KnoxMagnet, I lose the point of the portfolio.

6) I use a clear “priority rule” when both systems want a trade

Sometimes both systems want exposure at the same time. I handle it the same way every time:

Risk limits come first (max trades, heat cap, spread filter)
If I’m at capacity:
- In a clear Red market, TrendCapture can take priority
- Otherwise, KnoxMagnet takes priority
If I’m unsure, I skip. Consistency beats FOMO.

Skipping is a feature. It keeps my curve smooth.

7) My default “what runs where” template

This is my starting portfolio setup:

KnoxMagnet (core 6)

AUDNZD
EURCAD
NZDCAD
USDCHF
USDCAD
AUDCHF

TrendCapture (hedge 3–4)

EURUSD
GBPUSD
AUDUSD
USDCAD (optional overlap)

If I keep TrendCapture smaller and more selective, it behaves like insurance rather than noise.

8) I avoid operational mistakes (remote trading rules)

Because I manage remotely, I keep a few extra rules:

I prefer fewer charts, fewer trades, fewer surprises
I run on a VPS
I don’t change settings mid-week unless there’s a real reason
I review changes only during planned review windows

Remote trading success is mostly operational discipline.

9) My “portfolio sanity checklist” (quick)

Before I walk away and let the account run, I check:

Magic Numbers are different per system
Max open trades is low (2–3 total)
One trade per symbol is on
Risk caps are on (KnoxMagnet 2–2.5% cap; TrendCapture small fixed risk)
Spread filters and rollover blocks are active
I’m not overloaded on one currency theme

If those are correct, I can genuinely set and forget.

Risk framework

The rules that stop me blowing up and keep the curve investable

This is the part I take most seriously. A good strategy with bad risk control is just a delayed failure. My job is to stay in the game long enough for the edge to show up.

I don’t manage risk to avoid losses.
I manage risk to avoid unrecoverable losses.

1) My definition of risk (what actually matters)

Broker leverage (500:1 etc.) doesn’t protect me. It just changes margin requirements.

Real risk is:

how much I lose when I’m wrong,
how many times I can be wrong in a row,
and how quickly that compounds when multiple trades overlap.

So I focus on:

% risk per trade
portfolio heat (total open risk)
drawdown limits
execution conditions (spread/rollover)

That’s what controls survival.

2) My baseline risk targets

I keep two different “risk profiles” because the systems behave differently.

KnoxMagnet (core engine)

Base risk: 1%
Win-streak add: +50% profit, then +25% profit
Reset after 3 wins or any loss
Hard cap: 2.0–2.5% (3% only if I’m deliberately being aggressive)
Low trade concurrency (1–2 open trades total)

KnoxMagnet is about regularity. That means tighter ceilings.

TrendCapture (hedge)

Fixed risk only: 0.25–0.5%
Portfolio heat cap: 1–1.5%
Max trades: 1–2

TrendCapture is about insurance. It should help, not dominate.

3) My “portfolio heat” rule (my biggest curve smoother)

I never want a situation where one move hits multiple open trades and dents the account hard.

So I cap the combined risk across open positions.

My typical maximum combined open risk:

3% total across everything (core + hedge)

If I’m already at that exposure, I don’t add another trade. Even if it looks perfect.

This one rule does more for smoothness than almost anything else.

4) My hard stop rule (no exceptions)

I always trade with a stop loss.

I don’t:

widen it because I’m “sure”
remove it to avoid being stopped
hold and hope

Hope is not a strategy, and it’s not remotely manageable.

Stops are the thing that makes set-and-forget possible.

5) My kill-switch rules (the emergency brakes)

These are the rules that stop a bad day becoming a bad month.

I use R-multiples (risk units) because they work across all pairs.

Daily kill-switch (simple)

If I hit -3R in a day, I stop trading for that day.

That might mean:

disable AutoTrading
or set MaxOpenTrades to 0 temporarily

Weekly kill-switch (protects the curve)

If I hit -8R to -10R in a week, I stop trading until the next week.

This prevents “death by a thousand cuts” in ugly regimes.

I’d rather lose one week of opportunity than lose confidence in the system.

(If I want an even smoother investor-style curve, I tighten these.)

6) My drawdown tolerance (and what I consider “normal”)

I separate drawdowns into three categories:

Normal drawdown (acceptable)

A drawdown that fits within my expected operating range.

Typical:

small pullbacks while the system cycles

Regime drawdown (warning)

A drawdown that suggests the market regime has shifted.

This is when I tighten:

ADX thresholds
max open trades
risk cap

Structural drawdown (stop and review)

A drawdown that is too deep to ignore.

At that point I stop and review:

spread conditions
execution quality
whether the system is trading when it shouldn’t
whether my settings drifted away from what was tested

The most important part: I don’t “double down” to recover.
I reduce risk and restore process.

7) How I think about my return goal (4–6% per month)

I treat 4–6% as an aspiration, not a requirement.

Trying to force a constant monthly return is how people:

overtrade
loosen filters
increase exposure
and blow up

So I focus on the controllables:

quality filters
exposure limits
and consistent execution

If the edge is there, the returns follow.

8) What I tune first when results get rough

When performance degrades, I don’t immediately change the “signal logic.” I change risk posture first.

My tuning order:

Reduce max open trades
Tighten ADX filter (for KnoxMagnet)
Reduce risk cap (2.5% → 2.0%)
Tighten spread filter / rollover window
Only then consider signal tweaks

This keeps me from curve-fitting.

9) The uncomfortable truth I accept

Even great systems can have:

losing streaks
flat periods
sudden drawdowns

My edge isn’t “never losing.”
My edge is surviving those periods while staying consistent until the favourable regimes return.

That’s what investors buy: discipline and robustness.

Remote operation SOP

How I run this set-and-forget portfolio like a professional

Remote trading only works when I stop treating it like entertainment and start treating it like an operating system. The goal is not to “watch trades.” The goal is to keep the machine running cleanly, and only touch it when the rules say to.

1) My setup philosophy (minimum moving parts)

I keep the operational stack simple:

MT4 runs on a VPS
Both systems run with:
- strict spread filters
- rollover block
- hard stops
- low max open trades
I monitor from my phone (MT4 mobile)

I don’t want a fragile setup. I want a boring setup.

2) My daily routine (2–5 minutes)

This is what I check once per day:

A) Platform health

Is MT4 connected?
Is AutoTrading enabled?
Is my VPS up?

B) Errors and warnings

I check the Experts and Journal tabs for:

“cannot open file”
order send errors (invalid stops, off quotes, requotes)
repeated spread filter skips
any indicator load issues

If I see errors, I fix the cause.
If I don’t see errors, I do nothing.

C) Exposure snapshot

How many open trades do I have?
Are they clustered in one currency theme (AUD-heavy, USD-heavy, etc.)?
Am I within my risk caps?

If exposure is too concentrated, I don’t panic-close.
I simply stop adding new exposure until it normalises.

3) My weekly routine (15–30 minutes)

Once a week I do a proper review:

A) System behaviour review

I look for:

Did KnoxMagnet take trades mainly in low-ADX conditions?
Did TrendCapture only trigger in valid volatility/trend conditions?
Any unusual trade clustering?

B) Execution quality review

I check:

spreads at entry vs normal spreads
slippage patterns
whether the rollover block is doing its job

Bad fills can destroy a good system quietly.

C) Risk posture adjustment (if needed)

If the week felt rough, I adjust only the “risk levers” first:

tighten ADX threshold
reduce max open trades
reduce risk cap
tighten spread filter

I don’t rebuild the strategy mid-week.
I tighten the seatbelts.

4) When I intervene (clear rules)

I only intervene for four reasons:

1) A technical failure

Examples:

platform errors
indicator not loading
EA not trading due to missing files
repeated order send errors

2) Risk limits are being threatened

Examples:

too many open trades
exposure concentrated in correlated pairs
abnormal volatility causing too much equity swing

3) Execution conditions are abnormal

Examples:

spreads are consistently higher than expected
rollover behaviour is causing ugly entries

4) Planned parameter updates

I only change settings during planned review windows, not emotionally mid-run.

5) When I do not intervene (staying disciplined)

I do NOT intervene because:

a trade is temporarily in drawdown
a candle looks scary
I’m bored and want action
I “feel” like the market will reverse

If I interfere like that, I turn a system into a discretionary mess and I lose the whole point of automation.

6) My incident response plan (what I do when something goes wrong)

A) “Cannot open file” error

This usually means:

the EA or indicator isn’t compiled into a .ex4
or the file name doesn’t match what the EA is calling

My fix steps:

MT4 → File → Open Data Folder
confirm files are in the correct folders:
- MQL4/Experts/ for EAs
- MQL4/Indicators/ for indicators
open MetaEditor and compile:
- the EA → creates .ex4
- the indicator → creates .ex4
restart MT4 or refresh Navigator
confirm the indicator name matches exactly (no “(1)” etc.)

B) Order send errors (common MT4 ones)

If I see:

Invalid stops → my stop distance is too tight for broker rules (increase buffers or check digit precision)
Off quotes / Requote → poor liquidity moment; reduce slippage setting or avoid that time
Not enough money → my lot sizing is wrong or too many trades are open

I treat error codes like diagnostics. The platform is telling me what to fix.

C) Spread spike problem

If I see the EA skipping trades repeatedly due to spread:

I check whether it’s rollover time
I check whether the broker widened spreads unusually
I tighten the rollover block
or I reduce the pair list to cleaner pairs

A spread spike isn’t a “missed trade.” It’s often a trade I’m lucky to avoid.

7) My “remote emergency” actions

If I need to intervene fast, I keep it simple:

Option 1: stop new entries

toggle AutoTrading OFF
or set max open trades to 0 (if I’m editing inputs)

Option 2: reduce risk immediately

If I’m seeing unusually rough behaviour:

reduce risk cap
reduce max open trades
tighten ADX threshold

Option 3: flatten exposure (rare)

I only manually close trades if:

there’s a platform error and stops can’t be trusted
there’s a major broker issue
or the account is in an abnormal situation

Normally, the stops and logic should do their job.

8) My “do not touch” rule during live runs

Once I set parameters for a run, I don’t tweak settings daily.

Constant tweaking creates:

inconsistent data
inconsistent behaviour
and self-sabotage

I review, adjust, and then leave it alone long enough to get meaningful feedback.

9) What I log (so I can improve without guessing)

I keep a simple log (weekly is enough):

weekly return
max drawdown
number of trades per system
any error messages seen
any parameter changes made (and why)

That’s all I need to stay on track without drowning in noise.

Testing and optimisation

How I test properly without curve-fitting myself into a trap

Testing is where I either build confidence or build illusions. My goal isn’t to find the “best backtest.” My goal is to find settings that survive real market variety and still behave sensibly.

I don’t optimise for perfection. I optimise for robustness.

1) My testing philosophy (what I’m actually trying to learn)

When I test, I’m trying to answer three questions:

Does the system have a real edge or am I just lucky in one period?
What environments does it fail in?
Which settings improve stability without reducing the edge to nothing?

If I can answer those, I can trade with confidence.

2) My backtest rules (MT4 reality)

MT4 backtesting is useful, but it has limits, especially for:

spread variation
rollover behaviour
slippage
execution during volatility

So I treat MT4 backtests as:

a filter for bad ideas,
and a way to compare settings,
not a promise of future results.

My minimum test standard

test at least 5 years per pair if possible
include different regimes (quiet years + volatile years)
test per pair individually and as a portfolio (as close as possible)

If I can only do 1–2 years, I treat it as “early evidence,” not truth.

3) What I optimise (and what I refuse to optimise)

For KnoxMagnet, I optimise only the “environment control” dials first

These are the settings that improve the curve without changing the identity of the strategy:

ADX threshold (the big one)
RSI agreement level (4/5 vs 5/5)
ATR stop multiplier (2.0 vs 2.2 vs 2.5)
spread cap (strictness)
max open trades (portfolio smoothness)
risk cap % (behaviour control)

These are safe to tune because they’re not “chasing the past.” They’re shaping risk and trade quality.

What I avoid optimising heavily (curve-fit danger)

I avoid hyper-optimising:

exact RSI levels (like 68 vs 71)
exact divergence parameters
exact lookback windows to the point of fine-grained perfection

If I find myself tweaking tiny numbers for tiny gains, I’m probably curve-fitting.

4) How I judge if an “improvement” is real

I use a simple robustness checklist:

A) Does it improve more than one period?

If a setting helps only in one year and hurts others, it’s not an improvement.

B) Does it improve multiple pairs, not just one?

If it only works on one pair, it’s likely overfit.

C) Does it reduce drawdown without killing expectancy?

Smoothness matters, but I don’t want to neuter the system.

D) Does it make intuitive sense?

If I can’t explain why it works, I don’t trust it.

For example:

Tightening ADX to avoid trends makes sense.
Choosing Donchian length = 17 because it “looks best” is suspicious unless it generalises.

5) My “two-stage testing” method (simple and effective)

Stage 1: Baseline backtest

I run the default settings and record:

profit factor
max drawdown
average trade
win rate
consecutive losses
number of trades per month
equity curve shape (smooth vs lumpy)

I’m looking for “does this behave like a real system?”

Stage 2: One change at a time

I only change one parameter at a time and compare.

If I change 5 things at once, I learn nothing.

This is how I keep myself honest.

6) Forward testing (the step most people skip)

After backtests, I always do:

a demo forward test or small live test
because forward testing reveals:
real spreads
rollover behaviour
slippage
missed fills
VPS reliability

Backtests can’t show those properly.

My minimum forward test:

at least 4–8 weeks
enough trades to see behaviour (even if profit is small)

I’m testing behaviour more than profit at this stage.

7) How I test the staking ladder (without fooling myself)

My staking ladder is designed to enhance growth in favourable regimes, but it also changes distribution of returns.

So I test it in two ways:

A) Strategy expectancy remains positive without the ladder

If KnoxMagnet doesn’t have an edge at base risk, the ladder won’t save it — it will just amplify noise.

B) The ladder improves growth without blowing up the curve

I compare:

max drawdown
volatility of returns
worst week
recovery time after drawdown

If the ladder increases drawdown too much, I tighten:

risk cap %
max open trades
ADX strictness

The ladder is adjustable, not sacred.

8) Testing TrendCapture (what I expect)

Trend systems can look awful in some ranges and brilliant in trend years.

So I don’t judge TrendCapture by:

win rate
short-term month-to-month smoothness

I judge it by:

controlled losses
occasional large winners
how well it offsets KnoxMagnet during trend regimes

TrendCapture is a hedge. If it loses small while KnoxMagnet is winning, I’m fine with that.

9) My minimum evidence rule before I trust changes live

Before I accept any new settings live, I want:

5-year backtest evidence (or as close as possible)
across at least 3–6 pairs
and at least 4–8 weeks forward behaviour confirmation

If I can’t get that, I keep the change small and reversible.

Troubleshooting MT4

The fixes that save me hours and stop dumb errors killing good systems

Most MT4 problems aren’t “mystical.” They’re usually one of:

missing files
wrong folders
not compiled
name mismatch
broker rules (stops, digits, lot steps)
or platform connection issues

So I keep a simple diagnostic routine and I don’t waste time guessing.

1) The golden rule: read Experts + Journal first

Whenever something looks wrong, I check:

Experts tab (EA messages + errors)
Journal tab (platform events + connection issues)

If I don’t read those first, I’m blind.

2) “Cannot open file” (the most common one)

This error almost always means MT4 can’t find something the EA is trying to load.

Most common causes

the file isn’t in the right folder
the file exists but isn’t compiled to .ex4
the file name doesn’t match what the EA calls
the indicator is in a subfolder but the EA name doesn’t include the subfolder path

My fix steps (in order)

MT4 → File → Open Data Folder
Confirm file placement:
- EAs go in: MQL4/Experts/
- Indicators go in: MQL4/Indicators/
Open MetaEditor
Compile:
- the EA (.mq4) → should create .ex4
- the indicator (.mq4) → should create .ex4
Restart MT4 (or Navigator → right-click → Refresh)
Verify the name is exact:
- no “(1)”
- no extra spaces
- same spelling/case as the EA expects

Special case: indicator in a subfolder

If my indicator is in:
MQL4/Indicators/MyTools/Knoxville Divergence v3.5.ex4

Then the EA must call it like:
"MyTools\\Knoxville Divergence v3.5"

So if I use subfolders, I must include the path.

3) Compile errors (why a file “exists” but still won’t run)

Sometimes I copied the .mq4, but MT4 won’t create the .ex4 because it won’t compile.

My quick fix

Open MetaEditor
Compile
If it errors, I read the error lines and fix them first

No .ex4, no working EA/indicator. Simple.

4) EA attaches but doesn’t trade (my checklist)

If the EA loads but doesn’t take trades, it’s usually one of these:

A) AutoTrading disabled

The big AutoTrading button must be ON
Also: the EA settings must allow live trading
- “Allow live trading” checked

B) Wrong chart timeframe

KnoxMagnet is designed for H1
TrendCapture is designed for H4
If I put them on the wrong timeframe, I might get:
no signals
or strange behaviour

C) Filters doing their job (not a bug)

If spread filter, ADX filter, and target rules are strict, the EA may skip a lot.
That’s often correct behaviour.

So I check Experts tab messages:

if it says “Spread too high” or “Not enough RR,” that’s the system protecting me.

D) Market closed / no ticks

No ticks, no OnTick activity. Especially on weekends.

5) Common MT4 trade errors (and what I do)

MT4 error codes are basically the platform telling me exactly what’s wrong.

“Invalid stops”

Cause:

my SL/TP is too close for broker minimum stop distance
Fix:
increase stop buffer
check broker’s stop level requirements
ensure I’m calculating pips/points correctly (5-digit brokers)

“Off quotes” / “Requote”

Cause:

price moved, low liquidity moment, or volatility spike
Fix:
increase allowed slippage slightly
avoid trading during rollover
check connection quality

“Not enough money”

Cause:

lot size too big
too many trades open
broker margin rules
Fix:
reduce risk %
reduce max open trades
check min/max lot and lot step

“Trade context busy”

Cause:

too many operations at once
Fix:
add small delays between sends/modifies (if coding)
keep the system low-frequency (which mine is)

6) Spread and rollover issues (the silent killers)

If results look worse live than in tests, spreads and rollover are often the reason.

My response:

tighten spread filter
widen the rollover “no trade” window
avoid adding pairs that have unpredictable spreads at certain times

This is how I keep the curve smoother.

7) File + folder structure I expect to see

This is what “correct” looks like:

For KnoxMagnet

MQL4/Experts/
- KD_RSI5_ADX_KnoxMagnet_EA_SetForget_....ex4
MQL4/Indicators/
- Knoxville Divergence v3.5.ex4
- any other required indicators used by the EA

For TrendCapture

MQL4/Experts/
- TrendCapture_Donchian_ATR_EA.ex4
MQL4/Indicators/
- TrendCapture_Visuals.ex4 (optional visuals)

If something is missing from that structure, MT4 will complain.

8) My “fast diagnose” decision tree

When something breaks, I run this:

Is MT4 connected and AutoTrading on?
Any errors in Experts/Journal?
If “cannot open file”:
- check folder → compile → restart → name match
If “invalid stops”:
- check broker stop level + digits
If no trades:
- timeframe correct?
- filters too strict (or spread too high)?
- market open?

This fixes 90% of problems quickly.

Weekly/monthly review template (how I stay on track)

How I stay on track without obsessing or randomly tweaking

My review process has one job: keep me consistent and improving without turning the system into a constant experiment.

I don’t review to “find a reason to meddle.”
I review to make sure:

the systems are behaving as designed,
risk is controlled,
and any changes I make are planned and justified.

1) My weekly review (15–30 minutes)

I do this once per week at a consistent time.

A) Health check (non-negotiable)

Any MT4 errors in Experts/Journal?
Any missing files / indicator load issues?
Any broker issues (disconnects, freezes, abnormal spreads)?

If yes, I fix operations first. No point analysing performance if execution is broken.

B) Portfolio behaviour check

I ask myself:

Did I respect my max open trades limits?
Did I get trade clustering in one currency theme?
Did the systems behave like “core + hedge” or did the hedge become noise?
Any unusual exposure overlap?

If I see clustering or noisy behaviour, I tighten:

max open trades
correlation limits
or I reduce the pair list

C) Trade quality check (not just results)

I look at a handful of trades from each system:

For KnoxMagnet:

Was ADX low enough?
Was RSI stretch actually strong across TFs?
Was there realistic room to the magnet?

For TrendCapture:

Was the breakout aligned with EMA trend?
Was volatility expanding?
Was it a reasonable breakout, not a dead market poke?

If trades are being taken in the wrong environments, I tighten filters.
If trades are clean but results are poor, I accept that it may be normal variance.

D) Metrics I record weekly (simple and useful)

I log only what helps decision-making:

Weekly return %
Max drawdown % for the week
Number of trades per system
Win rate per system (lightly weighted)
Biggest win and biggest loss (in R)
Any platform issues / unusual spreads

This keeps me anchored.

E) My weekly decision rule (tighten vs hold)

If behaviour is clean and within drawdown expectations → I change nothing
If drawdown is elevated and trade quality looks worse → I tighten risk posture
If trade frequency is too low and results are stable → I can loosen slightly (carefully)

Most weeks the correct action is: do nothing.

2) My monthly review (45–90 minutes)

Monthly review is where I allow deeper changes — still controlled.

A) Month summary questions

Was the month “Green / Amber / Red” overall?
Which system carried the month?
Did the hedge actually hedge, or did it just add noise?
Was drawdown within my acceptable band?

B) System-by-system scorecard

I separate results:

KnoxMagnet

Is it still producing regularity?
Are losses clustered during high-ADX conditions?
Is the staking ladder helping or increasing volatility?

TrendCapture

Did it stay small and supportive?
Did it catch at least one meaningful move or was it pure chop?
Is it trading too often in dead conditions?

C) Parameter stability rule (anti-overfitting)

I don’t change more than one major dial per system per month.

A “major dial” is:

ADX threshold
risk cap %
max open trades
spread cap strictness
ATR stop multiplier

If I change multiple dials at once, I learn nothing.

D) My adjustment hierarchy (what I change first)

If I need improvement, I follow this order:

Reduce max open trades (smoothness lever)
Tighten ADX filter (KnoxMagnet protection)
Tighten spread/rollover filters (execution protection)
Reduce risk cap (behaviour control)
Only then consider deeper signal tweaks

This stops me blaming the wrong thing.

3) My “if this happens, I do that” playbook

This is my simple response system so I don’t react emotionally.

If I get a sharp drawdown week

I do:

reduce max open trades by 1
tighten ADX threshold slightly
tighten risk cap slightly

I do NOT:

switch pairs randomly
increase risk to recover
override trades manually

If I’m getting lots of small losses in a row (KnoxMagnet)

I assume the regime is trending.
I do:

tighten ADX threshold
reduce exposure
consider pausing the worst-behaving pair for a week

If TrendCapture is taking many false breakouts

That can be normal, but if it’s excessive:

I tighten the volatility filter
I require ADX rising
I reduce TrendCapture to the most liquid pairs only

If performance is great and I feel tempted to raise risk

I don’t.
Good periods are when traders get greedy and break their system.

If I want to scale, I do it slowly:

increase capital
not % risk

4) My “stay on track” rules (personal discipline)

This is what stops me drifting:

I don’t change settings mid-week unless it’s an operational problem
I don’t chase constant monthly targets
I record changes and reasons
I treat boredom as a warning sign, not a reason to trade more
I let the systems breathe

Consistency is the real edge.

5) My review template (copy/paste format)

Here’s the exact structure I use:

Week/Month:
Market regime: Green / Amber / Red
Return: _%
Max drawdown: _%
Trades (KnoxMagnet / TrendCapture): _ / _
Largest win (R): _
Largest loss (R): _
Any errors / execution issues?: Yes/No (details)
Observations (what I noticed):

Decision: Hold / Tighten / Loosen
If change: (one dial only)

Change: ______
Reason: ______
Date applied: ______

Appendix

That’s enough to stay professional without drowning in data.

Appendix A — Default settings templates

My starting presets (then I tweak only one dial at a time)

A1) KnoxMagnet (Core engine — H1)

Intent: high-quality mean-reversion trades, smooth curve, win-streak ladder ON.

Chart/timeframe

Run on: H1
One chart per pair
OneTradePerSymbol = true

Core risk + staking (your new ladder)

BaseRiskPercent = 1.0
UseProfitAddSizing = true
Win1AddFraction = 0.50
Win2AddFraction = 0.25
WinStreakResetAfter = 3
RiskCapPercent = 2.0 (smooth) or 2.5 (balanced) or 3.0 (aggressive)

Exposure controls

MaxOpenTradesAllSymbols = 1 (smoothest) or 2 (more opportunity)
Keep correlation rule: max 2 trades per currency theme

Filters (curve protection)

ADX filter (H1):
- start around 22–25
- tighten to 20–22 if trends are punishing
RSI MTF agreement:
- RSI_MinAgree = 4 (good balance)
- use 5 only if you want fewer but stricter trades

Stops + exits

Stop loss: ATR-based default (your EA uses this style)
- start around 2× ATR
TP1: magnet/pivot target with minimum RR filter (keep MinRR ≥ 2.0)
Runner: ON (optional), ATR trail ON (helps catch bigger reversals)

Execution hygiene

Spread filter: strict (especially on CHF crosses)
Rollover block: ON

Indicator loading

KnoxvilleIndicatorName = "Knoxville Divergence v3.5"
(must match file name exactly / include folder path if needed)

A2) TrendCapture (Safety net — H4)

Intent: hedge trend regimes, fixed risk only, accept false breakouts.

Chart/timeframe

Run on: H4
Keep pair list small and liquid (EURUSD/GBPUSD/AUDUSD/USDCAD)

Risk (fixed only)

RiskPercentPerTrade = 0.25 to 0.50
MaxTotalOpenRiskPct = 1.0 to 1.5
MaxOpenTradesAll = 1 or 2
No “add to winners” logic here

Signals

EMA_Period = 200
EMASlopeBars = 6
DonchianLength = 20
EntryBufferATR = 0.10
Volatility expansion: ATR_Expansion_Min = 1.05
ADX (optional but helpful):
- UseADXFilter = true
- ADX_Min = 18
- RequireADXRising = true

Stops + exits

SL_ATR_Mult = 2.2
Two-leg structure ON:
- TP leg: TP_R_Multiple = 2.0
- Runner: chandelier/ATR trail
  - TrailLookbackBars = 22
  - Trail_ATR_Mult = 3.0

Execution hygiene

Spread filter: strict
Rollover block: ON

Appendix B — Quick-start checklist (new VPS / fresh MT4 install)

B1) Install + file placement

Install MT4 (broker version)
MT4 → File → Open Data Folder
Copy files:
- EAs → MQL4/Experts/
- Indicators → MQL4/Indicators/

B2) Compile (this prevents 90% of issues)

Open MetaEditor
Compile:
- KnoxMagnet EA → must produce .ex4
- Knoxville indicator → must produce .ex4
- TrendCapture EA → must produce .ex4
- TrendCapture visuals (optional) → must produce .ex4

B3) Platform settings I always enable

AutoTrading ON
EA properties → Allow live trading ✅
DLL imports only if truly needed (generally avoid)

B4) Chart setup (clean + consistent)

KnoxMagnet charts: H1, one chart per core pair
TrendCapture charts: H4, one chart per hedge pair
Save templates so I can reapply quickly

B5) Final operational check

Experts/Journal tabs show no errors
Spreads look normal
VPS stable (no disconnect loops)

Appendix C — Glossary (my plain-English definitions)

ADX: Measures trend strength. Higher ADX = stronger trend. For counter-trend systems, high ADX is danger.

ATR: Volatility measure. Used for stops/trailing so the system adapts to market “breathing.”

Bar close / closed candle: I only act on confirmed candles to avoid mid-candle noise and false signals.

Donchian breakout: Break above/below the high/low of the last N bars. Simple, objective breakout method.

Equity curve: The account’s growth line over time. “Smooth” means controlled drawdowns and stable risk behaviour.

Heat / Portfolio heat: Total open risk across all trades. A key smoothness control.

Magnet / missed pivot: Old pivot levels that price tends to revisit later, especially after regime changes.

R (R-multiple): A unit of risk.

If I risk £100 and make £200 → +2R
If I lose £100 → -1R

Streak sizing (anti-martingale): Increasing stake after wins (not after losses). My ladder is a controlled version of this.

TrendCapture: My hedge system that aims for occasional big winners. It’s allowed to be lumpy.

Win-streak reset: After 3 wins in a row, I revert to base risk to prevent runaway compounding.

Appendix D — One-page rules summary (print/pin this)

My philosophy

I run two behaviours: mean reversion (core) + trend (hedge).
I protect the curve with filters + exposure limits + risk caps.
I don’t chase constant returns; I chase consistency and survival.

KnoxMagnet (Core) — H1

Pairs: AUDNZD, EURCAD, NZDCAD, USDCHF, USDCAD, AUDCHF
Goal: regular mean-reversion profits to magnets/pivots

Entry must-have:

MTF RSI stretch (≥ 4/5 timeframes agree)
Knoxville divergence confirmation
ADX filter says “not a freight train”
Enough room to target (MinRR ≥ 2.0)

Risk & staking:

Base risk: 1%
After win #1: 1% + 50% of last profit
After win #2: 1% + 25% of last profit
After 3 wins: reset to 1%
Any loss: reset to 1%
Risk cap: 2–2.5% (3% only if aggressive)

Exposure:

One trade per symbol
Max open trades total: 1–2
Avoid currency clustering (no more than 2 trades sharing the same theme)

TrendCapture (Hedge) — H4

Pairs: EURUSD, GBPUSD, AUDUSD, (USDCAD optional)
Goal: offset trend regimes and catch occasional runners

Entry must-have:

EMA200 trend alignment + slope
Donchian breakout + ATR buffer
Volatility expanding (ATR expansion)
Optional ADX rising filter

Risk:

Fixed only: 0.25–0.5% per trade
Heat cap: 1–1.5%
Max open trades: 1–2
No “add to winners”

Non-negotiables

Hard stop loss always
Spread filter always
Rollover block always
Low max open trades always
Different Magic Numbers per system
Weekly review, not daily tweaking

My review cadence

Daily: platform health + errors + exposure snapshot (2–5 min)
Weekly: trade quality + risk posture (15–30 min)
Monthly: one major dial change max, only with evidence (45–90 min)

Discover more from Stocked And Shared

Subscribe to get the latest posts sent to your email.