What Data for This Category Looks Like
Data sources for multi-finance financing daily reports are daily business submissions from licensed non-bank financial institutions and publicly disclosed documents from industry self-regulatory organizations. Full data updates are completed within 1 hour after the close of each trading day. Each daily report document uses structured JSON format. It includes six core fields: target securities code, full name of financing subject, daily financing purchase amount, financing balance, securities lending sale volume, and securities lending remaining volume. Amount fields use ten thousand RMB as their unit. Volume fields use shares as their unit. No nested sub-documents are included.
Constraints on HTTP Interfaces and External Systems
The data sources include two channels: institutional submissions and public disclosures. HTTP interfaces must support permission verification and data merging for both data sources. The daily-only update schedule requires external system scheduled pull tasks to be tied to trading day time windows. This prevents invalid requests. Core fields have fixed units. Interfaces must return standardized unit identifiers, eliminating the need for external systems to perform manual unit conversions. Each document contains multiple target entries. Interfaces must support pagination query parameters. This prevents timeouts caused by excessively large single-response data volumes.
Configuration Settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
API_KEY | Independent key bound to a dedicated workflow | Prevents permission conflicts caused by sharing keys with other applications |
WORKFLOW_ID | Workflow ID corresponding to financing daily report data pulling | Accurately invokes the specified business process and avoids calling unrelated applications |
REQUEST_TIMEOUT | 600 seconds | Accommodates the maximum response time for multi-source data merging and pagination queries |
PAGE_SIZE | 50 items per page | Balances single request data volume and interface call frequency to avoid timeouts |
DATA_SOURCE_TYPE | all | Covers both institutional submission and public disclosure data sources to ensure data completeness |
UNIT_AUTO_CONVERT | Enabled | Automatically standardizes returned field units to reduce external system adaptation costs |
The parameter values provided on this page are common recommended starting points for configuration. Actual values are affected by material form, data volume, and business rules. Specific issues require individual analysis. It is recommended to test on your own samples before finalizing settings.
Three Common Mistakes
- A 403 Forbidden status code is returned, indicating no access permission. The cause is use of an API key not bound to a dedicated workflow, or the application bound to the key not having external call permissions enabled.
- The unit of amount fields returned by the interface does not match expectations, with values using non-ten-thousand RMB units. The cause is failure to enable the
UNIT_AUTO_CONVERTconfiguration, leading to use of units from original submission data. - Scheduled pull tasks execute multiple times per day, triggering interface call frequency limits. The cause is failure to tie tasks to trading day time windows, resulting in invalid requests initiated on non-trading days.
How to Confirm Proper Configuration
- The configured
API_KEYandWORKFLOW_IDare input on the FastGPT interface debugging page, a test request is sent, and the returned status code is confirmed to be 200 OK. - Fields returned by the interface are compared with the preset core field list. All required fields are confirmed to be present with no missing entries.
- Unit identifiers returned by the interface are checked. These match standard units such as ten thousand RMB and shares.
- A scheduled task call is simulated. The number of returned data items is confirmed to fall within the range specified by the
PAGE_SIZEconfiguration.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-14.