Under the hood

Our Technology

BunnyQuant is not a screener with indicators bolted on. It is a probabilistic decision framework, built to answer a different question: not what price did, but what it is likely to do next - and how sure we are.

We describe the principles, not the models. Enough to judge whether the output deserves your money.

Confidence levels

50% → 99%

Timeframes

1m → 1W

1000s

Scenarios per signal

Every decision, simulated

20–40%

Fewer false positives

Multi-frame vs single-frame

What separates a probability from an indicator

Traditional indicators - RSI, MACD, Bollinger Bands - are descriptive. They summarise what price has already done and apply a fixed rule to it. Two traders reading the same RSI get the same answer, and neither knows how much to trust it.

The BQ engine is built the other way round. Instead of "RSI is 72, therefore overbought", it asks: given this market's current structure, its volatility regime and where it sits in its larger cycle, what does the range of plausible outcomes look like - and how tightly do those outcomes agree? The answer is a probability with a stated confidence, not a verdict.

Seven Pillars

Each decision is produced from thousands of simulated forward paths for that specific instrument - not from one indicator crossing a line.

How it works

The engine learns the instrument's own behaviour - how violently it moves, how its moves cluster, how today relates to yesterday - and projects that forward many thousands of times. The output is the full range of where price could plausibly go, with how often each outcome occurred.

What you see

A direction, a confidence level, and a horizon. The confidence comes from how tightly the simulated outcomes agree, so a signal in a calm, well-behaved market is not presented with the same certainty as one in chaos.

Why it matters

An indicator cannot tell you it is unsure. A distribution can. Most losing trades are not wrong calls - they are right calls sized as if they were certain.

Traditional indicator

RSI = 68 → overbought

One number. No sense of how reliable it is.

BQ engine

70% confidence, bullish, 12 bars

Uncertainty measured and stated.

How the tiers map to technology

Tiers are not marketing bundles. Each one is bounded by real infrastructure cost, real compute cost, and by what is safe to hand someone at that level.

Class	Data pipeline	Engine	Confidence ceiling	Shorts	API
Free	Batch (15-30 min)	50% only	50%	—	—
C-Class	Delayed (10-15 min)	Full BQ engine	90% (limited)	— Hard wall	— Hard wall
E-Class	1-2 min polled feed	Full + multi-frame	95% (E350 max)	✓ D1/4H only	— Hard wall
S-Class	Real-time tick feed	Full + structure + path	99% (S650+)	✓ Full intraday	✓ R/W
Maybach	Custom feeds	Unlimited + customised	All levels	✓ Unlimited	✓ Full + bot

"Hard wall" means the capability does not exist at that tier, no matter how many daily packages are bought.

What this compares to

None of these capabilities are new to finance - institutions have had them for decades. What is new is the price at which an individual can reach them.

Capability

BunnyQuant

Market equivalent

Probabilistic signal engine

From C200 ($29.99/mo)

Quant fund infrastructure - not sold to retail

Tail-risk position sizing

From C200 ($29.99/mo)

Bloomberg MARS: ~$6,000/year

Multi-timeframe cycle detection

From C300 ($49.99/mo)

Charting platform + custom scripting: ~$600/year, with no confidence measure

Structure and path probability

S450 ($299.99/mo)

Built in-house on a quant desk, or not at all

Short signals at 90%+ confidence

E350 ($149.99/mo)

Institutional desk tooling - not sold to retail

Real-time API and signal pipeline

S550–S650 ($549.99–799.99/mo)

Broker API plus a custom quant stack: $500+/mo and a developer

The tools have always existed. Access has not.

Probabilistic engines and institutional risk frameworks spent decades behind six-figure subscriptions and trading-floor infrastructure. BunnyQuant exists to put them within reach of the people actually taking the risk.

View Pricing Try the platform

BunnyQuant is built and run by one person. Everything you are reading here is free. ♥ Support the project