The ADA Oracle is a real-time ADA/USDT trading signal engine that analyzes price data across 4 timeframes using fractal mathematics, regime detection, and multi-factor confluence to tell you when, where, and how to trade ADA.
Every 60 seconds, the engine pulls live data from Binance and runs it through a series of calculations:
The trade card uses a color gradient that tracks edge. Bright colors mean high-edge tiers; muted colors mean low-edge tiers. You can tell at a glance whether to pay attention without reading a single number.
Stop Loss always appears red. Take Profit always appears green. Those are outcome markers — they show you where you lose money and where you win — and never change based on tier.
Fractal Regime Engine v2 uses the Hurst Exponent to classify market behavior, then adjusts every calculation to match. In trending markets, it boosts momentum signals and tightens stops. In choppy markets, it boosts reversion signals and widens stops. In random markets, it dampens everything and tells you to wait.
The entry system uses multi-factor confluence across 4 timeframes (1H, 4H, Daily, Weekly) and 5 factor categories (fractals, moving averages, Bollinger Bands, Fibonacci, VWAP). The more categories that agree on a price level, the stronger the entry.
Data source: Binance ADAUSDT public API. 1000 daily + 300 4H + 600 1H + 200 weekly candles. Refreshes every 60 seconds when connected.
The Hurst exponent (H) measures the memory of the market — whether recent price moves are likely to continue or reverse.
It comes from mathematician Benoit Mandelbrot's fractal research in the 1960s. It's not another indicator — it's a layer above all indicators that tells you which ones to trust right now.
The value is calculated using R/S (Rescaled Range) analysis across four timeframes, each with a long and short lookback window — eight calculations blended into one:
The weekly layer is like knowing the season — is ADA in a multi-month trending bull run or a choppy bear market? The daily is the weather forecast. The 4H is the radar. And the 1H is the early-warning system that catches regime shifts hours before the higher timeframes show them.
Together, they tell you the nature of the game being played — not the score, but whether you're playing poker or roulette.
Volume typically peaks between 8 AM – 2 PM UTC (3 AM – 9 AM Central) when US and European sessions overlap. Weekends and late nights are consistently low.
The system runs two scoring engines in parallel — one for LONG (price going up) and one for SHORT (price going down). Each scores from 0–100% by weighting 7 factors:
The Hurst regime dynamically adjusts these weights. In a trending market, momentum factors get multiplied up and reversion factors get pushed down. In a mean-reverting market, it's the opposite.
Whichever direction scores higher wins. The percentage shown is your confidence level:
Below the score bar you’ll see both raw scores and the point gap between them. For example:
The winning side determines direction. The gap tells you how close the flip is. A 2-point gap means one small shift (RSI moving a few points, volume picking up) could reverse the direction entirely. A 20-point gap means the direction is firmly established.
Both scores update every 60 seconds. Watch the gap narrow or widen to anticipate direction changes before they happen.
The Risk:Reward ratio (R:R) compares how much you could lose (distance from entry to stop loss) versus how much you could gain (distance from entry to take profit).
A 2.0x R:R means your potential profit is twice your potential loss. You only need to be right 33% of the time to break even.
The R:R is not a fixed number — it's dynamically calculated by two independent inputs:
1. Signal Strength (score %) controls the confidence-based sizing of stops and targets:
Higher confidence = tighter stop + further target = better R:R. This is the "how sure am I?" dial.
2. Hurst Regime (H value) then modifies those stops and targets based on what the market is actually doing:
So the final R:R reflects both your confidence in the trade and the mathematical structure of the market. A 2.5x R:R in a trending regime is much more trustworthy than a 2.5x R:R in a random regime, because the trend gives the target a mathematical reason to be reached.
The colored bar shows this visually — red is your risk portion, green is your reward portion. More green = better trade.
The entry price is calculated by finding where multiple independent factors agree on a price level across multiple timeframes. More agreement = stronger entry.
Below the entry price you’ll see up to 5 small squares. Each one represents a different factor category that confirms the entry level:
The entry, stop loss, take profit, and R:R bar all share a color system tied to the score. This is designed so you can tell at a glance whether the card is showing a real signal or just informational data.
Color heuristics here predate the April 2026 tier rename. Under the new naming, bright colors no longer mean “high conviction” — they mean “price has already extended.” The highest-edge tier is PRIME (gray).
The Hurst regime also adjusts how tightly factors must agree:
The stop loss is the price where the system says "you were wrong — get out." It's your maximum allowed loss on the trade, set before you enter so emotions don't make the decision for you.
1. Signal Strength Score sets the base distance using ATR (Average True Range — a measure of how much ADA typically moves per candle):
Why wider at PRIME? Because PRIME has the strongest edge — wider stops give the setup room to breathe through normal noise, which is why PRIME has the highest win rate (73.6% Daily). A tight stop would get hit by random fluctuation, not because the trade was wrong.
Why tighter at OVERHEATING? The market has already extended — many indicators have aligned, but the price has already moved. Any reversal confirms the trade was late. Tight stops limit damage on a thesis that's already chasing.
2. Hurst Regime then modifies that base distance:
After calculating, the system checks three limits and uses the tightest one:
For longs, the stop also checks against the nearest bullish fractal — if there's proven support just below the calculated stop, it moves the stop to just below that level (×0.99) to avoid getting stopped right at the support where a bounce is likely.
For shorts, it checks the nearest bearish fractal above and places the stop just above it (×1.01).
The take profit is the price where the system says "you were right — lock in the win." It's set to capture a realistic gain based on how confident the signal is and what the market structure allows.
Below 50% (PRIME / STRONG) — uses risk-based targeting:
The TP is calculated as a multiple of your risk distance (entry to stop loss). The multiplier scales with confidence:
Why is PRIME R:R below 1.0? Because on a low-score signal the engine is placing wide stops and close targets — the geometry that historically wins ~73%. A small realistic win you actually capture beats a big theoretical win that never materializes. PRIME is the highest-edge tier despite the low R:R on paper.
50%+ (CAUTION / OVERHEATING) — uses structural targets:
Instead of a formula, the system looks for actual price levels where the market is likely to stall:
The score determines how far to reach:
After selecting the target, the Hurst regime stretches or compresses the distance:
So the final take profit reflects three layers: where the market structure says price is likely to stall, how confident the signal is, and whether the market's memory supports an extended move or a quick one.
Leverage amplifies your trade. 2x means you control twice the ADA with the same capital. 5x means five times. But it cuts both ways — gains AND losses are multiplied.
The system maps your signal strength score directly to a leverage tier:
The core principle: amplify strong edges, protect against weak ones. A 70% OVERHEATING score at the old 15x leverage would risk catastrophic loss — on a tier that historically loses ~58% of the time. That's why OVERHEATING is now capped at 1x, while PRIME (73.6% WR) earns the highest leverage. Leverage is earned by proven edge, not by signal intensity.
Warning: At 5x+, a 20% move against you can liquidate your position. The system manages liquidation risk by scaling leverage inversely with score — PRIME gets the highest leverage but also wider stops and closer targets, keeping actual risk-per-trade controlled. Exchange liquidation can still override your stop in a flash crash.
When you use leverage, the exchange lends you money. If the trade moves far enough against you, the exchange force-closes your position to recover its loan. This is called liquidation.
The liquidation price is roughly calculated as:
The 0.95 factor means the exchange liquidates you at ~95% loss of your collateral (they keep a small buffer for fees and slippage).
Critical: If your stop loss is farther from entry than your liquidation price, the stop will never trigger. You'll be liquidated first, losing your entire collateral. The system detects this and shows a ⚠ LIQ warning when it happens.
At 1x leverage there is no liquidation — you own the ADA outright. Liquidation risk starts at 2x and gets dramatically worse at higher multiples.
The position shows two numbers: your ADA in → your ADA exposure.
The formula: Position % = score ÷ 100 × 25%, applied to your total ADA capital.
The system only commits large capital when the signal quality earns it:
Base position scales with score, but under the April 2026 inversion, leverage scales inversely with score. Result: the real dollar commitment (position × leverage) is highest at PRIME where edge is strongest, and smallest at OVERHEATING where edge is weakest. The signal earned its capital deployment by winning 73.6% of the time historically.
The 25% cap is your ruin-prevention limit. Even with a perfect signal, you never commit more than a quarter of your ADA into one trade. This means you can survive 4+ consecutive max-size losses before running out of capital.
The base position is then multiplied by leverage to get your total exposure — the actual amount of ADA you control.
If the risk:reward ratio is below 1.0, the position size is further reduced proportionally (e.g., 0.8 R:R = 80% of normal size). This prevents oversizing on trades where the risk math doesn't justify a full position.
This is the exact amount of ADA you lose if the trade hits your stop loss. No surprises — you know this number before you enter.
Under the edge-weighted inversion, PRIME (highest edge, 73.6% WR Daily) gets the largest leverage — your real dollar commitment is biggest where the 5-year backtest says you win most often. You can be wrong many times in a row on PRIME trades and still survive. OVERHEATING trades run at 1x, so even a full stop-out is a small capital event.
Important: Max loss assumes your stop gets filled at the exact price. In extreme volatility (flash crashes, exchange outages), slippage can cause actual losses to exceed this number. This is especially true with high leverage.
The engine looks at three things:
Most trades end before the hold period expires. There are three ways a trade exits, whichever comes first:
You choose how many ADA to commit to this trade. The engine calculates your max loss based on the stop loss distance and leverage.
The blue line at the bottom shows what the engine suggests based on your available balance and the current signal score.
The engine checks your Strike Finance balance first (⚡) since that's where your trading margin actually lives. If Strike isn't connected, it falls back to your CIP-30 wallet (🔗).
If the risk:reward ratio is below 1.0, the engine reduces the recommended position size proportionally. This prevents oversizing on trades with bad risk math.
The engine sets a default leverage based on the score, inverted so the highest-edge tier gets the most leverage. You can override it anytime. Higher leverage = bigger max loss for the same trade size.
The score at the top is the verdict — it tells you what the engine thinks right now based on all the math.
Things to Watch are the early warnings — they tell you what’s shifting underneath the score and what might cause it to change next.
Because it’s telling you what’s coming, not what’s here.
For example, if the score says PRIME SHORT but an insight says “buying pressure is building” — that’s not a contradiction. It means the short signal is still winning overall, but momentum is starting to push the other way. If that buying pressure keeps building, the score could weaken or even flip to long on the next refresh.
You don’t act on the insights alone — you act on the score. But the insights help you anticipate when the score is about to change, so you’re not caught off guard.
Insights are sorted by importance — the most urgent one is always first. A maximum of 4 show at a time to keep things focused.
There are 8 categories. Each one has a matching icon so you can connect what you see on the card to what it means here.
Every 60 seconds, the engine checks all 8 categories, generates every insight that’s currently true, sorts by urgency, and shows you the top 4. If fewer are active, the market is straightforward and the score alone tells the story.
When enabled, Nova-Kali evaluates every engine cycle (every 5 minutes) and can execute trades on Strike Finance without you pressing a button. Every trade has SL/TP set on Strike’s infrastructure — positions are protected even if the server crashes.
Auto-trade runs on the Daily engine (4H primary timeframe, 24-48 hour holds). With require_qualifying_combo enabled, only two setups qualify:
These are the two highest-probability setups in the entire 5-year backtest. The auto-trader only fires on these.
Leverage is inverted by edge — the tier with the highest win rate gets the highest leverage, not the other way around.
Your leverage cap setting overrides these — if you set leverage cap to 3, a PRIME trade uses 3x instead of 8x.
Auto-trade executes real trades with real money on Strike Finance. All trading involves risk. The 500 ADA daily loss cap and SL/TP on every trade provide safety rails, but losses can and will occur. Past backtest performance does not guarantee future results.
Shadow Trader is a parallel simulation that runs every 5 minutes alongside the live algorithm. It logs what the system would have done without any human intervention — no pauses, no skipped trades, no overrides. Then once a month, it compares those simulated trades to your actual trades so you can see whether your judgment is adding edge or subtracting it.
Shadow Trader has its own on/off switch. You can run shadow even when auto-trade is paused (useful: "what would the algo have done while I was away?"). Or run auto-trade without shadow. Or both. Or neither.
Shadow trades don’t touch your wallet or Strike account. They’re pure simulation — zero risk, zero fees, just information.
Shadow Trader needs at least 5 closed trades per config before performance summary numbers are shown — below that, sample size is too small to mean anything. First shadow trade only fires when a signal passes the full gate (score ≥ 35, confluence ≥ 3, qualifying combo). That may take hours or days depending on market conditions.