Sun-BD
Stock Market AI Strategy Analysis System — Taiwan Stocks & Index Futures Dual-Core Edition
AI Strategy Ranking, One-Click Import | Full prompt → three candidate strategies → one-click import
One application, two cores: Taiwan single stocks and index futures each get their own strategy workspace and backtest data, and never affect each other.
Your settings and backtest evidence become a complete prompt; ChatGPT, Gemini or Claude return three candidate strategies you can fill straight back in. After backtesting they enter the strategy ranking, and the one you like imports with a single click. No coding, and every strategy has evidence behind it.
Available now on Windows; the mobile version (iOS / Android) is coming soon, with store listings rolling out on the Microsoft Store, Google Play and the App Store.
The public browser reads a same-origin HTTPS health gateway. The gateway then verifies both HA web nodes without exposing search or database services.
Live Ranking
Dual-Core Strategy Backtest Ranking
Live comprehensive ranking of anonymous strategies across every server — top 50. The TAIEX futures ranking is above, filterable by the seven strategy categories; the Taiwan single-stock ranking is below, merging the futures-linked and stock-data rankings into one table with a Source column. Both share the data source of Subject Strategies → Strategy Backtest Ranking in the desktop app.
This table mirrors the TAIEX futures comprehensive ranking under the app's Subject Strategies → Strategy Backtest Ranking, top 50 (switch across the seven strategy categories; points are TAIEX futures index points; estimated monthly return = monthly net points × NT$200 per point ÷ NT$700,000 capital). Taiwan single-stock rankings are a separate data set and never mix with this table. The full top 500, per-trade evidence, hourly statistics and one-click import are available in the desktop app. Past backtest results do not guarantee future profits.
This table merges the two Taiwan-stock rankings inside the app (futures-linked = correlation branch, stock's own data = stock-data branch) into one comprehensive top 50, with a Source column telling them apart. Taiwan stocks have no index points: every figure is a percentage return — (exit − entry) ÷ entry per stock, net of the 0.435% round trip (0.1425% commission × 2 + 0.15% transaction tax), then averaged across stocks; the win rate is computed per stock and then averaged, and max drawdown is a percentage too. The rightmost column is a monthly return: the cumulative return divided by the strategy's total holding trading days, times 20 (a month holds about twenty trading days). Taiwan-stock cycles run from hours to days, so holding time is converted into trading days to give every trade one common unit; only hours actually held inside the 09:00-13:30 session count (divided by 4.5), flat time is excluded, and the figure is printed beneath each return. An em dash means that strategy's backtest is being regenerated and will appear when it finishes. Only net-profitable strategies with at least 2 trades are listed. Full rankings, per-trade evidence and one-click import are in the desktop app. Past backtest results do not guarantee future profits.
Dual Core
Taiwan Stocks × Index Futures — two cores in one application
Separate workspaces, revisions and backtests. Trade only futures, only single stocks, or both
◫
Separate stock workspace
Taiwan single stocks and index futures are two separate workspaces with separate data; strategies, revisions and backtests never affect each other.
◇
Stock-to-futures correlation
Correlation strength between each stock and the index future is computed on a 5-minute grid, so high-correlation names are identified from data, not intuition.
⇄
Trigger lock, linked entries
Once a futures strategy locks a trigger, the selected stocks enter and exit according to same/inverse direction, evenly split capital and highest-correlation-first rules.
◷
Taiwan session backtests
Replayed over the 09:00-13:30 Taiwan trading session; win rate, net profit, PF and max drawdown use exactly the same definitions as the futures side.
▤
Taiwan stock ranking
The futures-linked and stock-data rankings merged into one table of percentage returns net of fees and transaction tax, with the same one-click import into a custom subject slot.
⚿
Separate securities layer
Securities and futures use separate broker connection layers; stock orders travel the securities channel and never mix with futures orders.
LIVE DESKTOP
Live: the software running on a real machine
Live: the software running on a real machine
LIVE
Demo paused
The demo machine is not broadcasting right now; this panel lights up again on its own.
A live view of the demo machine (simulation mode, demo account), broadcast one way: watch only, with no way to control that machine. The view refreshes once a second, about 1-2 seconds behind real time.
Desktop Platform
Windows Trading Platform
Ten custom strategies, five core setting areas plus three first-entry gates, revision history, individual backtests, AI prompts, and simulated trading
Sun-BD Windows — Named Strategies · Revisions · Individual Backtests · Diffs · Reports · AI Prompts
Official Tutorials
Official Full Tutorials
Watch the complete SUN-BD Stock Market AI Strategy Analysis System TAIEX-futures core workflow (a Taiwan single-stock tutorial will follow). Follow login and the six operating areas, then learn AI multidimensional strategy analysis, four-dimensional backtests, ranked-strategy imports, individual backtests, and iterative refinement.
English · EN-US
English Full Tutorial
English narration with burned-in captions, covering the complete AI strategy analysis and backtest refinement workflow.
FULL TUTORIAL09:51
Traditional Chinese · 繁體中文
繁體中文完整教學
繁體中文旁白與畫面內字幕，完整示範 AI 策略分析及回測優化工作流程。
完整教學09:51
The videos are for product training and research. Historical backtests do not guarantee future performance. Before live trading, verify the broker API and recheck the market, order size, risk controls, and every parameter.
Core Features
Key Highlights
From market phase detection to risk management — comprehensive quantitative decision support
◈
Market Phase Detection
Automatically identify six structural phases: Accumulation → Early Rally → Main Rally → Overheated → Distribution → Decline.
⟁
Multi-Period Quant Engine
Integrates 12+ technical indicators with cross-period validation to present structured entry and exit references.
◎
Custom Subject Strategies
Compose five core setting areas across ten named slots, then gate a flat-position first entry by trend state, market phase, and the Nth independent trigger before saving and testing each revision.
⊘
AI Strategy Recommendation + AI Backtest Analysis
Generate complete TXT prompts for three UI-ready candidate sets, then analyze the current revision’s latest trade-level backtest evidence.
⬡
Simulated Trading
After configuration, observe entry, scaling, stops, targets, closing, and transaction records without sending a live broker order.
Download the Windows desktop version now; the mobile version (iOS / Android) is coming soon — see the download page for store listings and overseas direct downloads
