Algo Trading and Futu OpenAPI Skills Practical Private Sharing Session

Main speaker: Ryan, Head of Frontend Product at Futu

The mission of Algo equality and traditional barriers (00:06:49 - 00:12:15)

The four-step process of quantitative trading—strategy research, code implementation, backtesting optimization, and live deployment

Three major pain points: high programming skill threshold, lack of user-friendly backtesting tools, and complex live deployment. Futu has launched the Algo project, committed to achieving 'Algo equality.'

Innovative product one: Lego block-style zero-code strategy (00:12:15 - 00:16:00)

Abstract any complex trading strategy into 'condition + action' building block modules, allowing users to build strategies by drag-and-drop without writing a single line of code. For example: constructing, backtesting, and visualizing buy/sell points for an MACD golden cross buy/death cross sell strategy.

Innovative product two: AI natural language strategy generation (00:16:00 - 00:17:25)

Users only need to describe strategy logic in natural language, and AI automatically generates a complete block strategy diagram, then one-click backtesting or live trading can be performed, completely eliminating manual assembly steps.

Comparison of dual product lines: Desktop Argo vs. Futu OpenAPI (00:17:25 - 00:22:45)

Futu Algo is divided into two product lines:

One is the OpenAPI for developers—highly flexible, with market data requiring an additional fee and supporting self-deployment.

The other is the desktop client Algo for all investors—shared market data access, no extra fees, one-click backtesting, and live trading.

Futu AI Skills Live Demonstration (00:31:00 - 00:44:30)

Install Futu AI Skills via Claude Code (EasyCloud) → Natural language query $Tesla (TSLA.US)$ Stock price → Place order → MACD strategy backtest (Total return 30%, Annualized 9%) → Profit-taking and stop-loss parameter optimization (3×3 Cartesian combination loop backtest) → Paper and live trading. Fully automated by AI without any manual coding, including environment setup, strategy generation, and execution.

Futu Algo June Version Preview (00:44:30 - 00:52:30)

In June, the Futubull desktop will natively integrate AI agents, achieving a complete AI + Algo closed loop: natural language strategy generation, visualized backtesting, one-click cloud deployment for live trading, without relying on third-party AI agents. The REST API will also be released simultaneously, significantly expanding API capabilities to include advanced data such as stock screeners, short-selling data, and options anomalies.

Q&A Session (00:54:51 - 01:05:18)

Q: Can Skills be directly placed into the System Prompt?

Yes, but it will result in carrying a large number of tokens in each conversation, which is costly and inefficient. Not recommended.

Q: Should I use Claude Code or EasyCloud?

Both are viable; those who prefer a command-line interface will find Claude Code most efficient, while EasyCloud is more convenient for users on domestic networks.

Q: Which tool should new users download now?

For those focused on Algo, wait for the new Futu version in June; for those wanting to fully understand AI Agent capabilities, it's recommended to install an intelligent agent experience immediately.

Q: Can orders placed manually bypass restrictions via API (e.g., risk controls for low-priced stocks)?

No, the underlying trading system for the API is the same as manual order placement, with identical risk control rules.

The above content is AI-assisted and is for reference only. It does not constitute investment advice.

74K Views