What is 61.8 Signals?
61.8 Signals is a Fibonacci 61.8% retracement alert system across 15 markets — indices, commodities, bonds, the US Dollar, and crypto. Built around a single principle: fewer alerts, higher quality, honest about what we don't trade.
This guide walks you through every feature, top to bottom — from the math behind the strategy, to the dashboard, to placing your first simulated trade.
Open Dashboard Open Telegram Bot01 The math behind 61.8
61.8% is the inverse of the golden ratio φ ≈ 1.618. The same proportion appears in nautilus shells, the Parthenon, and 90+ years of market chart analysis.
The formula
φ = (1 + √5) ÷ 2 ≈ 1.6180339887... 1/φ = (√5 − 1) ÷ 2 ≈ 0.6180339887...
The 61.8% Fibonacci retracement level is 1 ÷ φ expressed as a percentage. When a market makes an impulsive move and pulls back, the 61.8% level is a common reversal zone — both because of math and because every trader watches it.
01b The vocabulary we use
Three terms keep coming up across the dashboard, charts, and Telegram alerts. Learn them once and everything below clicks faster.
| Burst | One clean directional move — a swing from low to high (UP burst) or high to low (DOWN burst). The fib is always drawn on the most recent burst. |
| Corrected | Price has retraced exactly 61.8% of the burst. Once corrected, the burst is "done" — the bot starts watching for the next burst forming in the opposite direction. |
| Chain | Each corrected retracement becomes the start of the next burst. Auto-Fib rolls over automatically as the chain advances. You can trade the same instrument all day as it walks through one burst after another. |
Throughout this guide you'll see "burst" instead of "impulse" and "corrected" instead of "retraced 61.8%." Same concepts, simpler vocabulary that matches how an experienced trader reads the chart by eye.
02 The strategy in plain English
Three components: detect a burst, wait for it to be corrected (retraced 61.8%), and only fire if the volatility regime makes mean reversion likely.
Bursts are read on 15-minute candles by default — the same timeframe most discretionary traders use to spot mean-reversion setups by eye. Entry labels on the chart and Telegram alerts are tagged with the order side: BUY E1 when the burst was UP (you're entering long on a pullback), SELL E1 when the burst was DOWN (you're entering short on a bounce). TP and SL labels stay generic.
Why BUY on an UP burst (and SELL on a DOWN burst)?
This is the part that confuses most new users. The bot is not predicting reversals — it's betting that strong, recent momentum resumes after a deep pullback. You're trading against the short-term retracement, with the original burst direction:
| UP burst | Price ran from low to high. It now pulls back 61.8% of the way down. You BUY the dip — betting the up-move resumes back toward the burst high. BUY E1 / E2 / E3 are three progressively deeper buy levels. |
| DOWN burst | Price dropped from high to low. It now bounces 61.8% of the way back up. You SELL the rally — betting the down-move resumes back toward the burst low. SELL E1 / E2 / E3 are three progressively higher short-entry levels. |
In both cases you are trading against the current direction of price action (which is the retracement), and with the direction of the original burst (which is the dominant move). When the entry hits and price reverses back toward the burst-extreme TP, the trade pays. When price keeps retracing through all three entries and breaches the burst origin (the 100% level), the shared stop-loss fires and the position exits. See § 02b for the SL placement and per-entry R:R math.
The 3-entry ladder
When an impulse is detected and price starts retracing, the system offers three entries — each at a deeper retracement level:
Take-profit ladder
| TP1 (Conservative) | 50% retracement — most likely to hit |
| TP2 (Moderate) | 38.2% retracement — mid-tier target |
| TP3 (Extra) | 23.6% retracement — stretch target |
The volatility-regime gate
Before any signal fires, we check the instrument's volatility coefficient:
VolC = recent average daily range ÷ older average daily range
- VolC < 1.0 — volatility falling, regime favors mean reversion → signal fires.
- VolC > 1.0 — volatility rising, regime hostile to retracement plays → signal silenced.
02b Risk management — the shared-stop ladder
Three entries, one stop-loss, placed just past the burst extreme. The ladder lets price retrace deeper without forcing a single perfect entry, and the shared stop fires only when the entire setup is invalidated.
The rule
All three entries on a signal (E1, E2, E3) share the same stop-loss. The stop sits just past the 100% level — the original burst origin — with a small instrument-specific buffer:
SL = burst-extreme ± instrument-specific buffer (the "stopBuffer")
For a BUY (an UP burst), SL goes just below the burst low. For a SELL (a DOWN burst), just above the burst high. If price ever closes through the SL, the burst's direction is dead — the whole ladder exits together.
Per-instrument buffers
Each instrument has a tuned stopBuffer that fits its typical wick size. Major ETFs like SPY/NDX/RUT use buffers around 0.05–0.1% of price; gold and silver around 0.15–0.3%; crypto (BTC, ETH, SOL, AVAX, XRP, LINK) around 0.2–0.5%. The buffer keeps the SL out of random noise without giving up so much room that losses get heavy.
Why this differs from a fixed R:R
Because the SL is anchored to the burst extreme (not to the entry), the risk-to-reward ratio varies by which entry fires:
| E1 (61.8% retracement) | Furthest from burst extreme → widest SL relative to TP. Lowest R:R per unit risked, but it's the highest-probability fill. |
| E2 (midpoint) | Closer to the burst extreme → tighter SL distance → better R:R if you ladder in here. |
| E3 (100% retracement) | Right at the burst origin → SL is just the buffer away → best R:R, but only fills on a full round-trip. |
The intuition: by laddering, you average down naturally if the move retraces deeper, and the single shared SL caps the worst case. When price reverses back toward TP, all three entries profit together. The 60-day backtest results (see § 03) are computed with this shared-stop formula.
03 The multi-instrument dashboard
The dashboard shows all 15 instruments at once, each as a mini-card with a live price, setup status, and a small candle chart. Click any card to drill into the focused chart view.
sixty-one-eight.pages.dev/61-8-app. Lime-bordered cards have an active signal; gray-bordered cards are scanning; red-bordered are vol-gated and won't broadcast to Telegram.What each card shows
- Instrument name + icon — quick visual identification.
- Phase — SCANNING (waiting for impulse), SETUP VALID (impulse found, watching for retracement), ENTRY_1/2/3_SIGNAL (signal fired).
- Current price + day change — live ticker.
- Mini candle chart — recent price action with HIGH/LOW markers and fib lines.
- VolC + win rate badge — instrument's vol regime status and backtest win rate.
04 The focused chart
Click any dashboard card to open the focused view. This is where you'll spend most of your time — live candles, all the fib levels, entry pills, and the expected-resolution projection.
What you can do on this view
- Watch live — chart refreshes every 5 seconds with new candle data.
- Toggle Auto-Fib — turn the automatic fib detection on/off (default ON).
- Toggle Hint — show/hide the gold status hint ("Retracing → wait for 61.8%") that overlays the chart.
- Resize the Hint — drag the bottom-LEFT corner of the hint to make it bigger or smaller.
- Adjust Min Impulse + Stop Buffer — per-instrument config to fine-tune detection sensitivity.
- Draw a manual fib — test your own analysis (see Section 6).
- Run sim trades — backtest the current setup on past candles (see Section 8).
04a Min Impulse — the noise filter
"Min Impulse" is the single most important configuration knob on the focused chart. It tells the analyzer how big a price move has to be before it counts as a real impulse — versus random noise that the strategy should ignore.
What it actually measures
An "impulse" in 61.8 is the price range between an absolute high and an absolute
low within a recent lookback window of candles. For an UP setup, the impulse is
measured from the low to the high preceding it; for a DOWN setup, from the high to the low.
Either way, the size is just |high − low|.
The Min Impulse threshold says: if that range is smaller than this number, treat it as noise. Don't compute fib levels. Don't fire any signals.
if (impulseSize < minImpulse) {
phase = 'IMPULSE_TOO_SMALL';
return; // no fib levels drawn, no signals broadcast
}
Why the threshold exists
- Filter out micro-moves. A 0.02-point wobble on UUP isn't a real impulse — it's tick noise. There's nothing meaningful to retrace.
- Improve signal-to-noise ratio. Without this filter, the analyzer would fire on every minor swing, generating dozens of low-quality alerts per hour.
- Keep retracement targets meaningful. A 61.8% retracement of a 5-cent move is 3 cents — too small to trade profitably against any reasonable spread.
- Protect the backtest stats. Including tiny noise-impulses would inflate trade counts and distort win-rate calculations.
Per-instrument defaults
Min Impulse scales with the instrument's typical daily range, so the default value varies widely. A 50-point move on Bitcoin is trivial; a 50-point move on UUP would never happen.
| US Dollar (UUP) | ~0.04 — low-volatility currency ETF |
| T-Bonds (TLT) | ~0.05 — bond ETF |
| Gold (GLD) | ~0.10 — gold ETF |
| Crude Oil (USO) | ~0.10 — oil ETF |
| Silver (SLV) | ~0.50 |
| Dow / SPY / QQQ | ~5 — index ETFs |
| Russell 2000 (IWM) | ~1 |
| Ethereum / Solana | ~5 / ~0.5 |
| Bitcoin | ~50 — extreme intraday range |
These are tuned to roughly half of each instrument's recent 30-day average daily range. That ratio gives the analyzer enough room to detect a meaningful intraday swing without firing on every minor tick.
The "Auto" button — what it actually computes
Next to the Min Impulse and Stop Buffer fields, the Auto button pulls the
instrument's average daily range over the last 30 trading days from the worker's
market-history endpoint, divides it by 2, and writes the result into the Min Impulse field.
// Pseudo-code for what Auto does: const dailyRange = avg(candles.map(c => c.h - c.l)); // 30-day mean const recommended = dailyRange / 2; inputMinImpulse.value = recommended.toFixed(2);
Why half of the daily range? Because we want to detect intraday swings, not full-day moves. Half-day-range is empirically a reasonable threshold: small enough to catch real impulses, big enough to filter out chop.
The "30-day avg" status line displayed in the cfg-row shows this number live, so you can see how the auto value compares to the instrument's actual recent volatility.
How to tune it manually
| Increase Min Impulse (e.g., 50 → 80 on BTC) | Fewer setups detected. Higher quality. Misses smaller-but-still-valid impulses. |
| Decrease Min Impulse (e.g., 50 → 30 on BTC) | More setups detected. Higher noise. May fire on chop or micro-trends. |
| Match Auto exactly | Recommended starting point. Conservative balance. |
What happens BELOW the threshold
When the current price action's impulse falls below Min Impulse, the chart enters
IMPULSE_TOO_SMALL phase. You'll see:
- HIGH and LOW markers still drawn — so you can see what range the analyzer is measuring.
- NO fib levels rendered — there's nothing to retrace from.
- NO entry pills — no E1/E2/E3/TP/SL pricing.
- NO Telegram broadcast — the bot stays silent on this instrument.
- Status text: "Waiting for a bigger impulse" or similar.
This is the analyzer being honest: I don't see anything worth trading right now. It's not broken; it's behaving correctly.
Relationship to other parameters
| Stop Buffer | Independent. Even if Min Impulse passes, Stop Buffer controls how much cushion the stop gets past the impulse extreme. |
| Vol-regime gate | Layered. First the impulse must be ≥ Min Impulse. THEN the vol-coefficient is checked. Both must pass for a Telegram broadcast. |
| Satisfaction logic | Once an impulse is detected (passes Min Impulse), it's tracked individually. Even if a later micro-impulse falls below Min Impulse, the original tracked impulse keeps going until retraced 61.8%. |
Common mistakes
- Set it to 0 or "tiny." You'll get flooded with noise alerts and the backtest stats will lose all meaning. The analyzer was designed assuming meaningful filtering.
- Crank it up to "huge" hoping to only catch monster moves. You'll wait days between setups, the strategy underperforms on small-impulse days, and you defeat the point of having a 15-market scanner.
- Use one number for all instruments. Bitcoin's noise floor IS higher than UUP's signal. The per-instrument defaults exist for a reason.
- Re-tune daily. The 30-day average changes slowly. Tuning Min Impulse every day is over-fitting; let it stabilize.
A worked example
Suppose you're watching UUP and the 30-day average daily range is reported as 0.085.
The Auto button writes 0.04 into Min Impulse (rounded down from 0.0425).
- UUP swings from 28.45 to 28.51 over 30 minutes — that's a 0.06 impulse. Above 0.04 → analyzer registers it, draws fib levels, watches for retracement.
- UUP wiggles between 28.49 and 28.51 for 20 minutes — that's a 0.02 impulse. Below 0.04 → analyzer ignores it, phase stays "IMPULSE_TOO_SMALL".
- UUP makes a clean 28.40 to 28.60 move — that's a 0.20 impulse. Well above 0.04 → strong setup; signal fires if it retraces to the 61.8% level (~28.48) AND vol-coefficient is favorable.
Why this is one of the most underrated knobs
Most retail traders fixate on entry signals — "the strategy" — and ignore the meta-question of when the strategy is even worth running. Min Impulse is the gatekeeper for that. A well-tuned Min Impulse is the difference between a system that signals 50 times a day (mostly noise) and one that signals 5 times a week (mostly actionable).
If you only adjust two knobs on the dashboard, make them Min Impulse and your position sizer's Risk %. Everything else is downstream of those two.
04b Auto-Fib Pending — anticipating the next burst
The focused chart has two toggles in the controls bar: Auto-Fib 1 and Auto-Fib Pending. Knowing what each does turns the chart from a snapshot of one burst into a live preview of the next one.
| Auto-Fib 1 | The current burst — the one the bot is actively trading. Pills are labeled BUY E1 / SELL E1 / TP1 / etc. |
| Auto-Fib Pending | The NEXT burst, projected from the current burst's last extreme. Pills are prefixed with "P" (P BUY E1, P TP1, etc.) and the anchor markers are cyan / red with a small "P" label. |
Pending fires as soon as ANY retracement starts off the current burst's end — you don't have to wait for the 61.8% line to be hit. As price moves further in the new direction, the pending fib stretches and its entries/TPs reposition. When the current burst gets corrected, the chain advances and the pending fib becomes the new Auto-Fib 1.
04c Chart reference markers
A few visual aids show up on every focused chart:
| Green ▼ (above a candle) | Auto-Fib 1's HIGH anchor — the candle whose high defines the burst's top. |
| Red ▲ (below a candle) | Auto-Fib 1's LOW anchor — the candle whose low defines the burst's bottom. |
| Cyan ▼ + "P" | Pending Fib's HIGH anchor — the candle Pending is projecting from. |
| Red ▲ + "P" | Pending Fib's LOW anchor — the candle Pending's other endpoint sits on. |
| White dotted vertical line | NY cash market open (9:30 AM ET, Mon–Fri). Helps you spot whether a setup formed during US session vs. overnight. DST-aware. |
| Lime horizontal line (bold) | The 61.8% retracement of the current burst — the line price needs to tag for E1 to fire. |
| Dashed lime/red lines | E2 (midpoint between E1 and E3) and SL (just past 100%). Drawn as additional reference levels alongside the standard fib lines. |
Chain advance — what to expect over time
As the chart updates, watch the green ▼ and red ▲ markers move. When price forms a new opposite-direction extreme, the algorithm rolls over: the OLD burst's end becomes the NEW burst's start, and the markers reposition to the new candles. This is the chain in action. No reload, no toggle, no manual reset — auto-fib follows the chart.
04d Strategy performance — 60-day backtest
Honest measurement. The strategy was back-tested over 60 days of 15-minute candles across all 15 instruments using the same algorithm that runs in production.
What "profitable" means here
Profit factor (PF) above 1.0 means the strategy made more money than it lost. PF of 2.0 means $2 won for every $1 lost. PF below 1.0 means the strategy bled.
Results — sorted by profit factor
| NASDAQ 100 | PF 2.67 · 58% win rate · 38 trades · +172 pts |
| Russell 2000 | PF 1.81 · 48% win · 48 trades · +28 pts |
| S&P 500 | PF 1.74 · 50% win · 68 trades · +47 pts |
| Dow Jones | PF 1.72 · 53% win · 70 trades · +325 pts |
| T-Bonds (TLT) | PF 1.59 · 44% win · 36 trades · +$0.18 |
| US Dollar (UUP) | PF 1.50 · 43% win · 28 trades · +$0.03 |
| Ethereum | PF 1.43 · 46% win · 50 trades · +$21.79 |
| Bitcoin | PF 1.28 · 40% win · 144 trades · +$643.71 |
| Silver (SLV) | PF 1.09 · 37% win · 75 trades · +$0.21 |
| Crude Oil (USO) | PF 1.09 · 39% win · 80 trades · +$0.54 |
| Gold (GLD) | PF 1.08 · 41% win · 75 trades · +$0.65 |
11 of the 12 mainstream instruments are profitable. The three smaller-cap cryptos (Solana, Avalanche, XRP, Chainlink) had too few qualifying setups in the window to yield meaningful statistics — they're not excluded but the sample is too thin to call.
How to read this with the chart
The pills on every focused chart (E1/E2/E3, TP1/TP2/TP3, SL) reflect the same entry and stop rules used to generate these backtest numbers. The Telegram alerts include the exact Entry / SL / TP prices to use. No mental math required.
05 Auto-detected signals
The Auto-Fib analyzer continuously scans for impulses and watches for 61.8% retracements. When all conditions align, it broadcasts a signal to Telegram.
The phase machine
| SCANNING | No impulse yet — system is watching for a qualifying move. |
| SETUP_VALID | Impulse found, price is retracing — not yet at 61.8%. |
| ENTRY_1_SIGNAL | Price reached 61.8% — primary entry fires. |
| ENTRY_2_SIGNAL | Price reached 76.4% — secondary entry fires. |
| ENTRY_3_SIGNAL | Price reached 88.6% — final entry before invalidation. |
ENTRY_1/2/3_SIGNAL phases broadcast to Telegram. SETUP_VALID is silent
(it's the "warming up" phase — not yet actionable). The worker also dedupes — same
instrument + direction + phase + impulse band gets ONE alert per 15-minute window, even if
multiple browser sessions post the same setup.
The "satisfaction" filter
Once an impulse has been retraced 61.8%, the analyzer retires it. This prevents the same setup from re-firing every 10 seconds while price wobbles around the level. The next signal will only fire when a brand-new impulse forms.
06 Manual fib drawing
Want to test your own setup ideas? Draw your own fib anywhere on the chart. Same entry/SL/TP math as the auto analyzer — but using your anchor points.
- Click DRAW MANUAL FIBCursor changes to crosshair mode.
- Click two points on the chartThe first click sets the impulse start (100%), the second sets the end (0%). Direction is determined by which is higher.
- Drag the endpoints to adjustBlue draggable nodes appear at start and end. Move them to refine your anchors.
- Read the levelsBlue dashed lines show 23.6 / 38.2 / 50 / 61.8 / 76.4 / 100%. Blue E1/E2/E3/SL/TP1/TP2/TP3 pills appear on the LEFT side (auto-fib pills stay on the right).
07 Position sizer
Fixed-fractional risk model — the same method used by professional discretionary traders. You set your account balance and risk tolerance; the sizer computes how many shares to buy per entry.
The math
position_size = (account_balance × risk_pct) ÷ (entry_price − stop_loss)
Example:
- Account: $10,000
- Risk per trade: 1% = $100
- Entry (E1): $28.50
- Stop: $28.25
- Risk per share = $0.25
- Position size = $100 ÷ $0.25 = 400 shares
Three sizing modes
| Split across 3 (default) | Risk budget ÷ 3, each entry gets one-third. Max total risk if all 3 fill = your configured %. |
| Each full risk | Each entry sized to full risk % independently. Max total risk = 3× configured % (aggressive). |
| Equal shares | Same share count per entry, equal to the minimum across the three. Simpler but uneven dollar risk. |
Industry best-practice defaults
| Conservative | 0.5% risk per trade |
| Standard (default) | 1.0% risk per trade |
| Aggressive | 2.0% risk per trade (do not exceed) |
08 Sim Trade
Test what would happen if you took the current setup. Two buttons: SIM AUTO walks the analyzer's setup forward through candles; SIM MANUAL does the same for your drawn fib.
What the sim is doing
- Computes E1, E2, E3, SL, TP1, TP2, TP3 from the active setup (auto or manual).
- Walks candle data forward from the setup's anchor point.
- For each entry, checks whether price reached the entry level — if yes, marks it filled.
- For each filled entry, walks further forward and checks SL or TP1/TP2/TP3 hit first.
- Multiplies (price P&L) × (your computed position size) to get dollar P&L per entry.
- Pushes the result into a log card with a colored AUTO or MANUAL badge.
09 Bot vs You — lifetime leaderboard
Every sim trade updates a running total stored on your device. AUTO totals (the bot's setups) sit next to MANUAL totals (your setups). Over time, you'll get hard data on whether your eye beats the analyzer.
10 Telegram alerts
Signals are delivered to your phone via Telegram. No app needed beyond Telegram itself. Free during preview. Cancel anytime.
What each signal looks like
Bot commands
| /start | Welcome message + linking instructions. |
| /link CODE | Link your Telegram account to your dashboard. Get CODE from the dashboard's "Get my link code" button. |
| /status | Show your subscription status and link state. |
| /unlink | Disconnect Telegram from your dashboard. |
| /help | List all commands. |
11 Get started — 90 seconds
- Open the dashboardVisit sixty-one-eight.pages.dev/61-8-app in your browser. Bookmark it.
- Open the Telegram botTap t.me/SixtyOneEightBot on your phone. Tap START.
- Get your link codeOn the dashboard, scroll to the "Connect Telegram" panel. Click Get my link code. You'll see a 6-character code.
- Send it to the botIn Telegram, type
/link XXXXXX(your code). Hit send. - ConfirmThe bot replies "✓ Linked." Signals will arrive automatically as they fire.
12 Reading the resolution projection zone
When the dashboard finds an active 61.8 setup, the focused chart overlays a projection zone on the right side. This zone shows where the trade is statistically expected to resolve — either by hitting your Take-Profit or your Stop-Loss — based on historical backtest data for that instrument.
What each element means
How to use it
- Set your alarms by the median. When ~15m says 15 minutes, set a phone alarm for 15 minutes after entry. If the trade hasn't resolved, check it manually — something unusual is happening.
- Use p75 as your patience ceiling. If the trade is still open past p75, the historical pattern says it's running long. Consider manually closing or tightening your stop.
- Per-instrument differences are real. BTC resolves in different time-windows than SPX. Solana resolves differently from Gold. The numbers update per instrument so you don't have to memorize them.
- Past performance is not a guarantee. These are statistical projections from historical backtest data. They're directionally useful but not precise.
FAQ Frequently asked questions
Getting started
Is 61.8 Signals really free?
Yes — during preview. We're capping the free phase at 100 subscribers. After that, $25/week. Cancel anytime, no annual lock-in. Existing free subscribers will be grandfathered or given paid upgrade options when the time comes.
What if I don't use Telegram?
Telegram is currently the only delivery channel. It's free, available on iOS / Android / desktop, and the bot setup takes 90 seconds. We may add SMS or email delivery in the future, but Telegram's free push notifications make it the best UX for now.
How long does setup take?
About 90 seconds end-to-end: open dashboard → click "Get my link code" → send /link CODE to the bot → done. Signals start arriving as they fire.
What does my account need to start using this?
Any brokerage account that supports bracket orders (entry + stop-loss + take-profit). Most major brokers do: Schwab, Interactive Brokers, TastyTrade, Tradovate. For crypto: Binance, Kraken, Coinbase Advanced. Robinhood has limited bracket support.
Strategy
Why 61.8%? Why not 50% or 76.4%?
61.8% is the inverse of the golden ratio φ (≈ 1.618). Every Fibonacci retracement charting tool draws it. Because so many traders watch it, it tends to act as a self-reinforcing support/resistance zone. The empirical edge has held in our 22-day backtest across the 11 currently-eligible instruments.
What are the risks of using 61.8 Signals?
All trading involves risk of capital loss. Past performance (whether from backtests or live signals) does not guarantee future results. 61.8 Signals is an indicator-only service; we provide alerts but you make all execution decisions. Position sizing, broker choice, and risk tolerance are your responsibility.
The vol-regime filter and 3-entry ladder are designed to control risk, but no system eliminates the possibility of losing trades — or losing weeks. Use proper position sizing (≤1% risk per trade is the industry standard) and never risk capital you can't afford to lose.
This is not investment advice. Consult a licensed financial advisor before making trading decisions.