Core coverage
Analyze stock-level flow by order size
Research use
Track Stock Connect capital direction
Extended capability
Study northbound and southbound holdings changes
REST API · PYTHON SDK · MCP
One request into your workflow
Every endpoint uses consistent authentication, security symbols and JSON responses, with REST API, Python SDK and MCP access available.
# pip install quantcoda import quantcoda as qc qc.set_token("qc_live_your_key") df = qc.moneyflow(symbol="600519.SH", start_date="20260101")
What this China market dataset includes
Money Flow is exposed through 4 documented endpoints. Each response preserves exchange-qualified security symbols, explicit trading or reporting dates and stable field meanings so that the same query can be reviewed in code, a backtest or an agent workflow.
Stock money flow
Typical filters: symbol, start_date, end_date, trade_date
Stock Connect money flow (market-wide)
Typical filters: start_date, end_date, trade_date
Northbound holdings detail
披露节奏说明：港交所自 2024 年 8 月起不再每日披露北向持股明细，本接口数据为港交所定期披露口径（低频），历史每日数据完整保留。查询建议不带日期或用较宽日期范围。
Typical filters: symbol, start_date, end_date, trade_date
Southbound holdings detail
南向持股明细为 T+1 披露：最新数据滞后一个交易日。
Typical filters: symbol, start_date, end_date, trade_date
CONSISTENT INPUTS
Symbols and dates that remain explicit
China stock symbols include their exchange suffix, such as 600519.SH, 000001.SZ or 430047.BJ. Calendar dates use YYYYMMDD; financial endpoints distinguish announcement dates from reporting periods.
RESEARCH OUTPUT
JSON for APIs, DataFrames for Python
REST responses use typed JSON values. The official Python SDK converts the same records into pandas DataFrames, while MCP exposes the same authenticated data contract to compatible AI clients.
FIELD-LEVEL DOCS
Inspect meaning before modelling
The documentation lists request parameters, response columns, units and plan requirements for every endpoint. Missing values remain explicit rather than being silently imputed by the API.