Smart Fund Picker

Problem

In public fund screening, metrics such as performance, drawdown, fund manager, size, and fees are spread across different reports. Manually checking conditions is easy to miss items with and hard to reproduce under the same screening criteria.

How It Works and Boundaries

The skill is built on the Hang Seng Juyuan MCP financial database and turns natural language conditions or preset strategy templates into fund screening tasks. It covers public fund categories such as equity, hybrid, bond, money market, and QDII funds, and can filter by fields like 3-year return, maximum drawdown, Sharpe ratio, fund size, and management fee. A typical workflow is: input screening conditions, call tools such as SmartFundSelection to query the full fund market, combine basic info, performance, holdings, and manager data to form a candidate pool, and output results with timestamps and data sources. Note that NAV is usually updated T+1 on trading days, holdings depend on quarterly disclosures, private fund coverage is partial, and newer funds may have incomplete fields. Historical performance does not guarantee future results; the output is a data-organization tool and does not constitute investment advice.