Natural-Language Stock Screening with Supermind and Rule-Based Daily Trading

Summary

This Chinese-language tutorial describes using Supermind’s iWencai query function to turn natural-language stock criteria into candidate lists for a daily trading workflow. It gives examples combining small-cap or value preferences, exclusions such as special-treatment or newly listed stocks, technical conditions, and fundamental or news filters. The accompanying strategy logic tracks holding duration, sells positions that meet exit conditions or appear in a sell list, excludes blacklisted names, and allocates positions up to a configured holding limit. Optional price-based profit and loss thresholds are also shown.

The article contrasts this screening approach with a simple single-stock MACD crossover example, arguing that natural-language filters can combine multiple data dimensions without manually coding each calculation. It outlines a route from research code to live trading through a broker interface, but provides no backtest results, execution analysis, or evidence that the screening service is faster or more reliable than direct data methods. Query interpretation, list ordering, data availability, and live order behavior all remain important practical constraints.

Key ideas

• Natural-language queries can produce stock candidate lists from technical, fundamental, news, and capital-flow criteria.
• A daily strategy can filter candidates, apply exclusions, limit the number of holdings, and size positions by portfolio share.
• The example exits positions after a holding period or when a sell-list or optional price threshold is reached.
• The tutorial positions query-based screening as an alternative to manually coding every selection rule.
• It provides implementation examples but no evidence of profitability, execution quality, or screening accuracy.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.