Stock Hot Momentum Analyst

Problem It Solves

When analyzing A-share hot sectors, the bottleneck is often not the indicators but the data pipeline: local install of Tongdaxin or pytdx, fetching quotes, and recomputing sector and stock scores. That usually means heavy environment dependencies and inconsistent field definitions, leaving analysts to manually assemble a table of “which sectors are hot today” and “which stocks are in the pool.” Stock Hot Momentum Analyst reduces this by consuming daily public JSON files and generating a hot momentum report with standard-library Python.

How It Works

In default user mode, it avoids local market terminals and extra pip packages. The script pulls three JSON files from a COS public-read bucket: longlong_homework.json provides the daily stock pool, longlong_homework_boards.json provides sector gains and constituents, and longlong_homework_stocks.json provides cloud-computed four-dimension stock scores.

The pipeline is:
- Ranks sectors by daily gain and takes the top HOT_TOPN hot sectors, defaulting to the top 20.
- Intersects sector constituents with the candidate stock pool using set operations.
- Reuses scores from stocks.json instead of recalculating from external quotes.
- Outputs student_hot_momentum.json and student_hot_momentum.html for structured results or a web-style report.

The scoring weights are: hot coverage 20%, sector strength 30%, volume confirmation 30%, and stock momentum 20%. Local code aggregates and ranks the upstream scores rather than recomputing the full market feed.

Limitations

It is suited for daily cross-analysis of hot sectors and a candidate pool, not real-time trading. If upstream JSON is stale, missing fields, or has corrupted constituents, the report may keep real gains and skip constituent intersections. Results depend on upstream data availability and definitions, so it is best for quick review rather than a full research process.