Skip to content
Research before risk · live evidence

Strategies canlook good once.We test what happens next.

Markets change. So we retest every strategy on unseen data, across different regimes and in live forward periods — then publish only the few that still hold up.

  1. 01Idea
  2. 02Stress test
  3. 03Live retest
  4. 04Evidence
strategies discovered
proven robust
certified
First 2 months free — nothing is charged until then.
500 founding seats · wave 1 open · applications reviewed by people.
Certified · passed every test
HaloEngine
BTC/USDT · 4H
net · backtested
+108.0%
win rate
66.5%
157W / 79L of 236 trades
2017-08-17 → 2026-06-26 (8.9y) · 236 trades
TV method, our replay: 66.1% · 706 exits
profit factor
2.32
gross win / gross loss · avg 0.41R / 0.34R per exit
expectancy +0.458R per closed position · 236 positions, 706 exits
2017-08-17 → 2026-06-26 (8.9y) · 236 trades
max drawdown
4.5%
intrabar peak→trough · longest 253 days
2017-08-17 → 2026-06-26 (8.9y) · 236 trades
vs hold · return / DD
1.7×
raw return +108% vs hold +1155%
(net% / maxDD) ÷ (hold% / holdDD)
net − hold = −1047 pp
Backtest #26091 · 2017-08-17 → 2026-06-26 (8.9y) · 236 trades · computed 2026-06-27 02:25
Out-of-sample split: not exposed per strategy by the public API — see the method.
Recent edge · public API snapshot · 2026-08-28
12-month evidence
30 trades · 53.3% win
+5.5R
6M
+6.6R
WR 67%
12 trades
5M
+7.6R
WR 73%
11 trades
4M
+7.6R
WR 73%
11 trades
3M
+5.2R
WR 67%
9 trades
2M
+4.1R
WR 67%
6 trades
1M
+5.2R
WR 80%
5 trades
Past performance does not guarantee future results. Validation method

Every strategy in the indicator. Real numbers.

This is the exact list members pick from inside the TradingView indicator — every certified strategy plus the top robust ones per coin and timeframe. Numbers come straight from the engine and refresh live. We delimit by confidence: certified (full validation), robust · confirmed (edge alive on the base gate too) and to-watch (in only through the indicator's recent-edge filter — not the base gate, so size it small).

strategies
certified
coins
Markets BTC ETH SOL XRP

live from the engine · every % shows its sample size · click any row for the full report · to-watch = in via the indicator filter, not the base gate — trade it small

Eight tools, one path — raw idea to proven edge.

Eight stages, eight different instruments. Each one shows its evidence in the form that stage actually produces — and most of them read off live numbers, not illustrations.

01StageStrategy Discovery
Live numbers

Find edges you'd never code by hand.

Discovery happens in parameter space, not on one chart. The funnel counts live: how many combinations were tried, and how few survive each filter.

  1. combinations tested
  2. kept in catalog
  3. survived robustness
  4. certified
What this gives youYou skip months of blind trial and error.

Thousands of strategies enter, almost all fail — and you see the failures too.

Validation funnelSyncing feed
5,712Strategy ideasgenerated per round
≈18,500,000Parameter setsstress-tested per round
99.97%Screened outstress + unseen data
Screening result0.03% advance
99.97% rejected0.03% advance
Verified survivorsSyncing feed
Live catalogsyncing
Robustsyncing
Forward testsyncing

Backtest assumptions: Binance spot data, 2017–2026, every timeframe the venue exposes; fees 0.08% per side; broker-emulator fills with partial exits and trailing stops; no leverage. Backtested results are hypothetical; past performance does not guarantee future results.

The robustness proof
Syncing feed

Only strategies that survive stress, unseen data and a live trial reach this card. Every figure below comes from the engine — never from a placeholder.

Checking the verified feed

This preview will not fabricate a win rate while the engine cannot be reached.

win ratesyncing
profit factorsyncing
max drawdownsyncing
How a strategy earns its place

Why we claim better

QuantRhinoHow it’s usually sold

  • Tested on unseen data

    QuantRhino: held out from tuning, then judged on years it never saw

    How it’s usually sold: scored on the very period it was tuned on

  • Corrected for luck

    QuantRhino: Sharpe deflated by the number of attempts; overfitting probability scored

    How it’s usually sold: no correction — enough attempts make luck look like skill

  • How many were tried

    QuantRhino: the number is published

    How it’s usually sold: only the survivor is shown

“How it’s usually sold” describes common practice, not any particular company — and the point is the method, not the marketing.

Regime evidence

Markets change. The strategy has to keep up.

Bull, range and bear are not separate demos. Move through one certified backtest and see what the same rules produced in each market condition.

Expansion

Trend-led periods. Every result earned while momentum was rising stays visible.

Scroll through the section or choose a regime

Visual previewRegime simulationBull · Range · Bear
syncing certified data
net
max DD
trades
Results measured separately across bull, range and bearOne curve · one rule set · no hidden regime

What membership includes

Free for the first 2 monthsCreate an account and use everything free for 60 days — nothing is charged until then.

These are not two plans to choose between. The subscription is the access — everything already validated. Credits are optional fuel on top, only if you want the engine to go looking for something new, on coins you pick.

Subscription · access
31.25€first paid month50€
free 2 months, then 31.25€ (−37.5%), then 50€/month · cancel anytime
  • ·Full validated catalog + every metric, live
  • ·Complete recipes + Pine v6 export for TradingView
  • ·Discovery signals + the forward ledger
Credits · your own searches
1€per coin + timeframe search
exactly 1 credit per request · one timeframe per request · separate from the subscription
  • ·Launch YOUR own search on the coins you choose
  • ·Pick coin + timeframe + criteria (profit window, filters)
  • ·A dedicated box just for you searches & validates
  • ·Your searches also grow the shared catalog
Buy credits

Take one of the 500 founding seats.

Applications are reviewed by people. Not everyone is accepted.

Begin application
Founding members
/ 500
seats assigned · wave —
reading live count…
Wave 1 · 500 seatsopened 12 Jun 2026

Waves close when their seats are assigned — not on a timer. Every seat counted above is a real database row.

Example sealQUANTRHINOFOUNDING MEMBER#137

The #137 drawn here is an illustration — it is not one of the seats counted above. Every member receives a seal generated from their number — no two are alike. Yours is drawn the moment you are accepted.

Beta testers get free access while they help us test. Decisions within 72 hours.

Questions

Start here, even if trading is new to you.

All 17 questions, on the Help page

Trading involves substantial risk of loss. Past performance does not guarantee future results.