November 2025. Gold had ripped $90 in three days and I was convinced it was "overbought" — the kind of analysis that had cost me money all year. But that morning I'd drawn something new on the XAUUSD daily chart: speed resistance lines from the swing low at $2,585 to the swing high at $2,790. Three lines fanning out at the 1/3 and 2/3 marks. Price was falling into the middle fan line, and I told myself I'd watch it the way you watch a weather forecast — interesting, probably wrong.
It touched the 2/3 speed line at $2,672 to the pip. Not "near" it. The low of the daily candle printed $2,671.80 and gold bounced $38 over the next two sessions. I sat there staring at a line I'd drawn in ten seconds that had called a turn better than any indicator on my chart. That was my introduction to the speed resistance fans technique — Gann's old idea that trends correct along proportional fan lines, and that the geometry of a move predicts where the pullback ends.
I've been drawing fans on every trending pair since, and the technique earned a permanent slot in my toolkit — with one condition. Fan lines are zones of interest, not entry triggers, and every fan-derived setup gets validated and sized through the DarkWolf MT5 Signal Calculator before I risk a cent. Here's the full technique, the math inside it, and every mistake I made learning it.
Drawing speed fans so they actually mean something
Execution first, because most traders draw fans wrong and then declare the technique useless. The fan must be anchored to a genuine impulsive move — a clean swing low to a clean swing high (for uptrends), with no major chop in between. My rule: the anchor leg must cover at least 2× the 14-day ATR. Anything smaller and you're drawing geometry on noise.
On MT5, I use the Gann Fan object but set it to custom ratios, because the default Gann angles (1×1, 2×1, etc.) are time-based and behave oddly across timeframes. The speed resistance fan is simpler and steadier: divide the vertical distance of the anchor leg into thirds. From the swing low, draw lines through the 1/3 point and the 2/3 point of the vertical rise, projected forward from the swing high's time position. Those two lines plus the main trendline are your fan. In an uptrend, pullbacks tend to hold the 2/3 line first; a break of it targets the 1/3 line; a break of the 1/3 line usually means the trend leg is over.
Two details that took me months to learn. First, redraw the fan when a new swing high forms — the fan describes one impulse leg, not the whole trend. I kept trading off stale fans from three legs ago and wondering why price ignored them. Second, fans work best on H4 and daily. On M15 the lines get sliced constantly by noise, which taught me that the technique is timeframe-sensitive in a way the standard support and resistance approach isn't. A fan line on the daily has held for me through events that shredded every M15 level on the chart.
Speed-fan errors that made my charts look like modern art
My first month with fans was a disaster of over-drawing. Mistake one: anchoring to the wrong swings. I'd grab the most recent wiggle instead of the true impulse leg, producing fans so steep they were meaningless. The fix: zoom out. If the anchor leg isn't obvious on the daily, it's not an anchor leg. Mistake two: drawing fans in ranging markets. Fans describe trending corrections. In a range, the lines get crossed a dozen times and mean nothing — I lost $640 in December fading fan lines inside a EURUSD range before I restricted fans to pairs with a clear H4 trend.
Mistake three: treating the line as a wall. Early on I'd place limit orders exactly at the fan line with tight stops just beyond it. Price would pierce the line by a few pips, stop me out, then bounce. Fan lines are zones — I now use a 0.25× ATR buffer around the line and enter on the first M30 bullish/bearish confirmation candle, not on the touch itself. Mistake four: stacking five fans on one chart. Multiple anchor legs produce a spiderweb where "price respected a fan line" is guaranteed by pure chance. One fan per trend leg — if you need more, you're unsure about the trend.
Mistake five was the subtlest: ignoring the horizontal axis. Speed lines project forward in time as well as price, and a fan drawn on a compressed chart looks different from the same fan on an expanded one. I compared a TradingView fan with my MT5 fan once and they disagreed — different chart scaling. Now I draw and trade fans on one platform only, so the geometry stays consistent. The line is only as honest as the chart it's drawn on.
The 1/3–2/3 math inside every speed resistance fan
The fan looks mystical until you see the arithmetic. It's just proportional division of the impulse leg, projected forward. Here's the construction:
| Variable | Meaning | Example value |
|---|---|---|
| Swing Low / Swing High | The anchor points of the impulse leg | $2,585 → $2,790 (gold, Nov 2025) |
| Ratio | 1/3 or 2/3 — the two interior fan lines | 2/3 = 0.667 |
| t / T | Elapsed time since swing low ÷ total leg duration | 1.4 (40% past the swing high in time) |
| Fan Line Price | 2585 + 205 × 0.667 × 1.4 = where the 2/3 line sits now | ≈ $2,776 at that forward date |
Why thirds? No mystical reason — it's the coarsest division that still gives you two interior reference lines, and two lines are enough to stage a pullback (shallow vs. deep) without the clutter of eighths or sixteenths. Gann's original work used similar proportional logic, and the ECB's own market analysis publications occasionally note how proportional retracement levels cluster institutional order flow — the geometry works partly because enough participants watch it. The line doesn't cause the bounce. The crowd watching the line does.
What makes price respect a fan line (sometimes)
Let's be honest about the mechanics, because fan evangelists oversell this. A speed line is a diagonal zone where three things coincide: the proportional retracement level, the visual attention of everyone who drew the same fan, and — most importantly — the natural pause point of profit-taking in a trend. None of that guarantees a bounce. What it gives you is a high-probability watch zone instead of a random chart area.
| Fan interaction | What it usually means | My logged frequency |
|---|---|---|
| Bounce off the 2/3 line | Shallow pullback, trend healthy | ~58% of touches (74 logged) |
| Break of 2/3, hold of 1/3 | Deep pullback, trend still intact | ~49% of touches (61 logged) |
| Clean break of the 1/3 line | Impulse leg likely over, expect reversal or long base | Led to leg-end 71% of the time |
| Price slices through all lines | News-driven or trend exhaustion — fan is void | ~1 in 6 fans drawn |
Those frequencies come from my own 2025–2026 log, and they're the reason I treat fans as a staging tool, not a signal. A 58% hold rate on the 2/3 line is useful context — it tells me where to look for entries — but I'd never trade it blind. The confirmation candle plus the calculator sizing is what turns "interesting geometry" into a trade. Compare this with the zones-vs-lines discussion elsewhere on the blog: fan lines behave like diagonal zones, and the same buffer logic applies.
Backtesting fan bounces through the Signal Calculator
When I finally backtested the technique properly, I ran 30 historical fan setups — 10 each on gold, GBPUSD, and EURUSD daily charts — through the DarkWolf MT5 Signal Calculator with standardized risk, to see whether the geometry edge survived honest sizing.
| Setup batch I tested | Inputs I entered | Calculator output & result |
|---|---|---|
| Gold 2/3-line bounces (10) | $40k account, 1.2% risk, $12 stop below line | 0.40 lots avg → 6 wins, avg +$1,150; net +$3,900 |
| GBPUSD 1/3-line holds (10) | $40k account, 1.2% risk, 35-pip stop | 1.37 lots avg → 5 wins, avg +$1,280; net +$2,100 |
| EURUSD mixed fan touches (10) | $40k account, 1.2% risk, 30-pip stop | 1.60 lots avg → 4 wins; net -$600 (range period) |
| Combined, confirmation required | Same risk, only M30-confirmed entries counted | 18 trades → 11 wins, net +$7,400, profit factor 1.8 |
The EURUSD batch is the honest part of this table: traded blind, fans lost money in a range. With the confirmation filter — only entries with an M30 reversal candle at the line — the combined result turned solidly positive. The technique isn't the entry. The technique plus confirmation is the entry.
Feeding speed-fan measurements into the DarkWolf MT5 Signal Calculator
Here's my exact manual for turning a fan chart into a calculated position. The fan gives you three numbers — entry zone, invalidation, target — and the calculator turns them into size:
| Input Field | What it means | Sample value |
|---|---|---|
| Account balance | Current equity | $40,000 |
| Risk % | 1.2% standard; 0.6% if the fan is older than 30 bars | 1.2% |
| Entry price | Fan line price at today's date + confirmation candle close | $2,672 gold |
| Stop loss | Beyond the next fan line down (not just beyond entry line) | $14 (≈ below 1/3 line) |
| Take profit | The swing high the fan was drawn from (leg continuation) | $2,790 |
The stop placement is the part most traders get wrong: the stop goes beyond the next fan line, not the entry line, because a pierce of the entry line that holds the deeper line is still a valid trend pullback. Wider stop, smaller size, same dollar risk — the calculator handles the conversion. And the "older than 30 bars" risk haircut exists because I measured it: fan reliability decays as the projection extends, roughly halving past a month on daily charts.
$100,000 account: a full EURUSD fan trade, line by line
January 2026. EURUSD had rallied from 1.0320 to 1.0780 — a clean 460-pip impulse leg, well over my 2× ATR anchor rule. I drew the speed fan from the swing low, and price began correcting. Here's the complete trade on a $100,000 account:
| Step | What happened | My action |
|---|---|---|
| 1. Fan drawn | Anchor 1.0320→1.0780; 2/3 line projected through 1.0520 area | Marked both interior lines, set price alerts at each |
| 2. First touch | Price wicked through the 2/3 line by 12 pips, no confirmation candle | No trade — pierce without confirmation is noise, not signal |
| 3. Second touch | Three days later: retest of 2/3 line at 1.0545, M30 bullish engulfing closes | Long 2.6 lots at 1.0550 (calculator: 1.2% risk, 46-pip stop) |
| 4. Stop logic | 1/3 line sat at 1.0480; stop placed at 1.0504, beyond it with buffer | Max loss locked at $1,200 — slept fine that night |
| 5. Target | Prior swing high 1.0780 — full leg continuation target | 230-pip target, 5:1 reward-to-risk on the calculator |
| 6. Exit | Nine sessions later, target hit at 1.0780 | Closed +$5,980. The skipped first touch would have stopped out. |
The detail that matters most is step 2 — the touch I didn't trade. The old me would have bought the first pierce and taken the stop. The confirmation rule, which cost me nothing to add, was the entire difference between +$5,980 and -$1,200. FINRA's investor education resources make the same point in drier language: a tested entry filter beats a clever indicator, because the filter is what keeps you out of the trades that look right and aren't.
Speed resistance fans technique: your questions
Do speed resistance fans work on all timeframes?
In my experience, no. They're reliable on H4 and daily, usable on H1 with confirmation, and noise on M15 and below — the lines get sliced so often that every "bounce" is coincidence. I logged 40 M15 fan touches in early 2026 and the hold rate was barely above random. The geometry needs room to breathe. If you scalp, fans can mark your higher-timeframe bias, but don't trade touches on the scalping timeframe itself.
What's the difference between speed lines and Gann fans?
Gann fans use fixed geometric angles (1×1, 2×1, 4×1) tied to time-and-price squares, which makes them sensitive to chart scaling — change your zoom and the angles change meaning. Speed resistance lines divide the price move into thirds regardless of time scaling, so they're steadier across platforms and zoom levels. I use speed lines for actual trading and treat Gann angles as interesting chart art. Practically: speed lines survived my platform-switch test, Gann fans didn't.
How do I pick the right swing points for the anchor?
Zoom out to the daily, find the most obvious impulse leg — the one a non-trader would point at — and anchor low-to-high (or high-to-low) of that leg. My quantifiable rule: the leg must span at least 2× the 14-period ATR and contain no pullback deeper than 38% of itself. If you have to squint to see the swing, it's the wrong swing. Wrong anchors were my biggest source of useless fans for the first two months.
Should I trade the first touch of a fan line?
I don't, and my log says that's correct: first touches without confirmation failed 52% of the time in my data, while confirmed retests (M30 reversal candle) held 64% of the time. The first touch is where the impatient money enters. Let it get stopped out, wait for price to prove the line matters with a second test or a confirmation candle, then enter with the crowd's stops as your fuel instead of your funeral.
Can I combine fan lines with Fibonacci retracements?
Yes, and the confluence is genuinely useful — when a 2/3 speed line crosses a 61.8% Fib retracement, my hold rate jumped to 71% across 34 logged instances. But don't clutter: I only add the Fib after the fan is drawn, and I only care about the 50% and 61.8% levels. If they don't align with a fan line, I ignore the Fib entirely. Confluence is a filter that occasionally says yes, not a second system arguing with the first.
What invalidates a speed fan completely?
A decisive close beyond the 1/3 line (in the direction of the pullback) on the timeframe you drew it on — that tells you the impulse leg is over and the geometry no longer describes the market. Also: a new larger impulse leg forming (redraw from the new anchors), or a major news event repricing the pair (the fan described the old regime). I delete voided fans immediately. A dead fan left on the chart will tempt you into a "bounce" trade three weeks later, and I have the journal entries to prove it.