News Catalyst
Sterling opens today into a session where the dollar leg of the pair is carrying all of the narrative risk and the pound is carrying almost none — which is exactly the asymmetry a low-lag trend filter is designed to trade. The dominant story is Fed communication: a CNBC Fed Survey showing 80% of respondents want Chair Warsh to say more about his read on the economy at Jackson Hole tells you positioning is unusually thin ahead of any remark he makes, and thin positioning into a speaking event is what produces the fast, one-directional 15-minute legs this system exists to capture rather than the slow drift it filters out. On the European side the economic calendar is close to empty for GBP: the only scheduled print is German GfK Consumer Confidence (forecast −29.6 against −29.6 prior), a low-impact release that reaches GBPUSD only indirectly through EUR/GBP cross flow and is unlikely to do more than add a few pips of noise in the European open. That combination — no domestic UK catalyst, an unscheduled dollar catalyst that can arrive at any moment — is the double-edge of today's tape. It means trends, when they start, will start suddenly and run without the staircase pullbacks a lagging moving average needs to stay onside; it also means the pair will spend long stretches doing nothing at all while the market waits. The strategy's answer to both is the same: a fast trend line to catch the move early, and a hard no-trade filter to sit out the waiting.
Trade Summary
This is a trend-following system built around the question most trend systems never ask — not which way, but whether to trade at all. A Hull Moving Average (21) supplies the direction with far less lag than an EMA of the same length, and Velocity(10) supplies the gate: when the rate of change in price is below a volatility-scaled threshold, the chart is in a no-trade zone and every HMA turn inside it is treated as noise, not signal. The HMA decides direction; Velocity decides whether direction means anything yet. A Candle Pattern confirmation on the signal bar and ATR(14) risk sizing complete the structure.
The strategy is directionally neutral — it takes long and short trends with identical logic — and it is explicitly built for high-volatility, catalyst-driven sessions. It performs best when a real driver produces sustained directional pressure, which is the profile a Fed-speech day tends to hand GBPUSD once the London and New York sessions overlap. It performs worst in exactly the conditions the Velocity gate is there to exclude: quiet, range-bound tape where the HMA's low lag becomes a liability, whipping direction on every minor swing and generating a stream of small losses. On GBPUSD 15-minute charts the pair trends in reasonably clean legs during the overlap and chops badly outside it, which makes the no-trade filter the highest-value component of the entire system rather than an optional extra.
The Anatomy of the Trade
The Logic: What Inefficiency Are We Exploiting?
Every moving-average trend system faces the same trade-off. A slow average is reliable but late — by the time it turns, a meaningful slice of the move is gone. A fast average is early but unreliable — it turns on noise as readily as on trend. The Hull Moving Average attacks the lag side of that trade-off directly by combining weighted averages of different lengths so the result tracks price with materially less delay than an EMA of equal period. What it does not do, and what is routinely misunderstood, is fix the second half of the problem: an HMA in a flat market turns more often than an EMA does, not less. The lag is gone, and so is the accidental filtering that the lag was providing.
That is why the edge here lives in the separation of direction from permission. Velocity measures how fast price is moving, independent of where the HMA points, and a market whose velocity sits below a multiple of its own recent average is a market where no directional read is worth acting on regardless of what the trend line says. Gating HMA signals on a Velocity threshold removes the single largest loss cluster in low-lag trend trading — the flat-market whipsaw — without touching the early-entry advantage that made the HMA worth using. The candlestick confirmation adds the final layer: a Hammer at a swing low or a Shooting Star at a swing high shows the bar that triggered the signal was actually rejected by participants, not merely a bar that happened to close on the right side of a line. Each filter removes a distinct failure mode, and dropping any one of them reintroduces its own.
Setup Requirements
- Primary indicator: Hull Moving Average (21) applied to closing price, used for both slope direction and the price-versus-line relationship.
- No-trade filter: Velocity(10) — the setup is disarmed entirely whenever absolute Velocity reads below 1.0× its own 50-bar average. This is the defining component of the strategy, not a secondary filter.
- Confirmation: Candle Pattern — Hammer or Bullish Engulfing on long signals, Shooting Star or Bearish Engulfing on short signals, closing on the signal bar.
- Risk management: ATR(14) sets stop distance, target distance and position size, so risk stays constant as GBPUSD ranges between a 12-pip and a 40-pip 15-minute ATR.
- Primary symbol: GBPUSD — the pair carries enough intraday range for a 21-period HMA to separate signal from noise, and its liquidity keeps spreads tight enough that a low-lag system's higher trade frequency does not get eaten by costs.
- Timeframe: 15-minute charts — fast enough that the HMA's reduced lag is a genuine advantage on the entry, slow enough that the Velocity reading reflects real directional pressure rather than single-tick spikes.
- Adaptability: the structure transfers well to EURUSD, USDJPY and the major index CFDs, all of which alternate clearly between trending and dead phases. It transfers poorly to thin crosses and to instruments with frequent gaps, where Velocity readings jump for structural reasons rather than momentum ones.
Entry Rules
Every condition below must align on the same bar. An HMA turn in a low-velocity tape is not a weak signal to be sized down — it is not a signal at all, and the setup should be treated as unarmed until the Velocity gate opens.
- Long entry: HMA(21) slope has turned upward on the current bar and price closes above HMA(21) and Velocity(10) is positive and reads at least 1.0× its 50-bar average and the bar closes as a Hammer or Bullish Engulfing.
- Short entry: HMA(21) slope has turned downward on the current bar and price closes below HMA(21) and Velocity(10) is negative with absolute value at least 1.0× its 50-bar average and the bar closes as a Shooting Star or Bearish Engulfing.
Enter at the close of the confirmation candle. Do not anticipate the HMA turn intrabar — the Hull calculation is weighted toward recent price and its intrabar slope can reverse completely before the bar settles, which means an early entry is acting on a value that will not exist a few minutes later.
Exit Rules
- Stop loss: 1.5× ATR(14) from entry, placed beyond the swing extreme of the signal bar.
- Take profit: 3.0× ATR(14) from entry — a 2:1 reward-to-risk minimum, which the low-lag entry makes reachable because the position is established near the start of the leg rather than a third of the way into it.
- Signal exit: close the position when HMA(21) slope reverses against the trade, regardless of open profit. This is the primary exit in practice — the same responsiveness that gets the position on early gets it off early.
- Velocity exit: close the position if absolute Velocity(10) falls below 0.5× its 50-bar average for three consecutive bars. The move has stopped being a trend; holding it is holding a range position with a trend stop.
The stop loss is non-negotiable. Low-lag entries produce a distinct psychological trap: because the position is opened early in the move, a normal retracement often takes it briefly underwater before the trend resumes, and the temptation to widen the stop “because the HMA still points the right way” is constant. The 1.5× ATR stop is what makes the strategy's arithmetic work at a sub-50% win rate; moving it converts a bounded loss into an unbounded one.
Risk Management
- Risk per trade: 1–2% of account equity. With the Velocity gate active, qualifying setups appear roughly two to four times per session on GBPUSD 15m, so there is no frequency argument for sizing beyond 2%.
- Risk-to-reward ratio: minimum 2:1. At a 1.5× ATR stop and a 3.0× ATR target, the system stays profitable above roughly a 36% win rate before costs — which matters, because a trend system with a signal-based exit will book many small losses between the runners.
- Position sizing: on a $25,000 account risking 1.5% ($375), with ATR(14) at 18 pips on the 15-minute GBPUSD chart, the stop distance is 27 pips. At roughly $10 per pip per standard lot, that gives $375 ÷ (27 × $10) = approximately 1.39 standard lots. Recalculate every trade; GBPUSD ATR can double between the Asian session and the London–New York overlap.
- Maximum concurrent positions: one on GBPUSD, and no more than two dollar-denominated pairs open at once. Running GBPUSD alongside EURUSD on the same signal is one dollar bet in two tickets, not two trades.
SYMBOL: GBPUSD
TIMEFRAME: 15m
TREND LINE:
hma = HMA(21, close)
hma_up = hma > hma[1]
hma_down = hma < hma[1]
NO-TRADE ZONE:
vel = Velocity(10)
vel_avg = average(abs(vel), 50)
armed = abs(vel) ≥ 1.0 × vel_avg
// While armed = false, no entry is valid
LONG ENTRY:
armed = true
hma_up = true AND close > hma
vel > 0
candle = Hammer OR Bullish Engulfing
SHORT ENTRY:
armed = true
hma_down = true AND close < hma
vel < 0
candle = Shooting Star OR Bearish Engulfing
STOP LOSS: 1.5 × ATR(14) from entry
// Beyond the signal bar swing extreme
TAKE PROFIT: 3.0 × ATR(14) from entry
// 2:1 minimum reward-to-risk
SIGNAL EXIT: hma slope reverses against position
STALL EXIT: abs(vel) < 0.5 × vel_avg for 3 bars
RISK: 1–2% of equity per trade
MAX POSITIONS: 1 on GBPUSDCommon Pitfalls
The failure modes below are specific to low-lag trend systems. Most of them are variations on a single mistake — trusting the HMA's speed in conditions where speed is precisely the wrong property.
Trading Through the No-Trade Zone
The most common way to break this strategy is to take the HMA signal without checking Velocity, usually because the chart looks like it is trending when viewed at a glance. In a compressed range the HMA hugs price closely and turns every few bars, and each turn looks like the start of something. The Velocity gate is not a filter that improves the strategy — it is the strategy, and the HMA is just the direction input. If you find yourself overriding the gate because a setup looks clean, you are trading a bare fast moving average and should expect the results of one.
Fed Communication and Unscheduled Dollar Headlines
GBPUSD's dollar leg is exposed to central bank commentary that arrives without a calendar entry, and Jackson Hole weeks are dense with it. A remark from the Fed Chair produces a velocity spike that arms the setup and an HMA turn that points into it — a technically perfect signal generated by a one-off repricing that may fully retrace within a few bars. Treat unscheduled central bank headlines as a stand-down condition, not a confirmation. Wait for the initial impulse to complete and re-arm on the structure that forms after it.
Relaxing the Candle Confirmation
Because the HMA turns early, there is always a bar where the slope and Velocity conditions are met but no qualifying candle has printed. Taking that bar feels like getting in even earlier for free. It is not: the candle requirement is what separates a bar that closed on the right side of the line from a bar that was actively rejected. Signals without candle confirmation degrade the win rate by more than the earlier entry recovers, and because they are the most frequent near-miss, relaxing this rule roughly doubles trade count while cutting expectancy.
Over-Optimising the HMA Period and Velocity Threshold
The 21-period HMA and the 1.0× Velocity threshold are round, defensible starting points, and a backtest will always find a pair of values that performed better on the sample you tested. Tuning both simultaneously against a single instrument on a single year is the fastest route to a curve-fitted system. If performance collapses when the HMA period moves from 21 to 18 or 24, the edge was in the parameter, not in the logic. Test a grid, keep the plateau, discard the peak.
Revenge Trading After a Whipsaw Cluster
The Velocity gate reduces whipsaws but cannot eliminate them, and when the market transitions from trending to ranging the strategy will typically give back several trades in a row before the gate closes. That cluster arrives immediately after the system's best performance, which makes it feel like something broke rather than like normal regime change. Define the daily loss limit before the session and stop at it — three consecutive stopped-out trades on GBPUSD 15m is a signal that the regime has turned, not an invitation to size up.
Build Strategy using Arconomy
The GBPUSD HMA Velocity No-Trade Zone Filter is assembled in the Arconomy Strategy Designer as a chain of rules, with the Velocity gate wired as a mandatory condition into the same Logic block as the trend and candle conditions — so the setup cannot fire while the market is inside the no-trade zone.
| Step | Rule(s) Required | Description | Key Configuration |
|---|---|---|---|
| Data | Price Data | Feed 15-minute GBPUSD candles into the strategy as the base series for every downstream rule. |
|
| Entry | Moving Average | The Hull Moving Average supplies both the directional read and the price-versus-line condition — the low-lag core of the system. |
|
| Filter | Velocity | The no-trade gate. Signals are blocked entirely while momentum speed sits below its own recent average. |
|
| Filter | Candle Pattern | Confirms the signal bar was rejected by participants, not merely closed on the right side of the trend line. |
|
| Filter | Logic | AND gate combining trend, Velocity gate and candle confirmation so all three must be true on the same bar. |
|
| Risk | Place Trade + ATR | Sizes the position from realised volatility so a fixed percentage of equity is risked in every volatility regime. |
|
| Exit | Stop Loss + Take Profit | Bounds every trade, with the HMA slope reversal and the Velocity stall condition wired as additional close signals. |
|
| Backtest | Validate across trending and ranging GBPUSD regimes, checking specifically how much the Velocity gate removes from the trade count. |
|
Backtest Considerations
Test a minimum of twelve months of GBPUSD 15-minute data, and make sure the sample straddles both regimes the strategy cares about. A window drawn only from a strong trending stretch will flatter the HMA badly, because the component that carries most of the strategy's value — the Velocity gate — contributes nothing when there is nothing to filter. The single most informative test is the paired one: run the strategy with the gate enabled and again with it disabled. If the gate is doing its job, trade count should fall substantially while profit factor rises; if disabling it barely changes the result, the threshold is set too loose to be filtering anything.
Watch profit factor above 1.3, maximum drawdown, and the distribution of results across the sample rather than the headline return. A low-lag trend system with a signal-based exit produces a long tail of small losses and a short tail of large winners, so a healthy equity curve here looks lumpier than a mean-reversion system's — check that no single month carries the entire result, and that the largest winning trade is not a double-digit percentage of net profit. Consecutive-loss runs of five or six are normal for this structure and should be verified as survivable at your chosen risk percentage before going live. The Arconomy backtesting engine reports each of these directly.
Cost assumptions matter more here than in a slower system, because the low-lag entry raises trade frequency. Model GBPUSD spreads at 1.0–1.5 pips through the London and New York sessions and widen that materially for the Asian session and the hour around the daily rollover — a strategy that looks profitable on a flat 0.8-pip assumption can lose its entire edge under realistic variable spreads. Add at least half a pip of slippage on entries, since signals cluster at the start of velocity expansions when the book is thinnest, and confirm the backtest fills at the confirmation bar's close rather than at an idealised signal price.
Key Takeaways
- The Hull Moving Average removes lag from trend detection, but it does not remove false signals — a fast average in a flat market turns more often, not less.
- The Velocity no-trade gate is the core of this strategy: it blocks entries entirely while momentum speed sits below its own recent average, removing the whipsaw cluster that defines low-lag trend trading.
- Candle confirmation on the signal bar separates a bar that closed on the right side of the line from a bar that was actively rejected, and relaxing it roughly doubles trade count while cutting expectancy.
- ATR-based stops at 1.5× and targets at 3.0× keep the arithmetic viable above roughly a 36% win rate, and widening the stop because the trend line still points the right way is the fastest way to break it.
- Stand down around unscheduled Fed commentary and validate the strategy with the Velocity gate both on and off — if disabling the gate barely changes the backtest, the threshold is too loose to be doing any work.
Credits
The strategy idea originated from the following YouTube channel. Concepts have been adapted and structured for systematic implementation by Arconomy.
ALGOBOX PRO's video walks through coding an explicit “no-trade zone” — a programmatic condition that tells an automated system when to stand aside rather than which direction to take — and that inversion is what this post builds on, pairing the gate with a low-lag Hull Moving Average on GBPUSD so the fast trend read is only ever acted on once momentum confirms the market is actually moving.
