A-Share Financial Red Flags: Finding Risk Signals with Data

Why check for landmines first

Before studying growth or valuation, confirm the company has no obvious red flags — do it in the wrong order and the rest of the analysis is wasted.

All four steps below use public filings and disclosure data and run through the API. No buy/sell advice is given; the goal is to surface risk signals.

Step 1: cross-check the statements (profit vs cash)

import requests

KEY = "ct-your-key"
H = {"Authorization": "Bearer " + KEY}

r = requests.get("https://api.ashareapi.com/v1/finance",
                 headers=H, params={"code": "sh600667"}, timeout=20)
tabs = r.json()["data"]            # three tables: income / balance sheet / cash flow
names = ["income", "balance sheet", "cash flow"]
for name, tab in zip(names, tabs):
    latest = tab[0]                # periods are descending: first row is the latest
    print(name, latest["EndDate"], latest.get("OperatingRevenue"))

• What to look at: net profit vs net operating cash flow — a persistent gap (high profit, chronically negative cash flow) is a classic signal
• Look at multiple periods: seasonality distorts single quarters; use at least 4–8 periods (`num`)
• Receivables growing faster than revenue: possibly buying revenue with looser credit

Step 2: shareholder counts (concentrating or dispersing)

r = requests.get("https://api.ashareapi.com/v1/shareholder",
                 headers=H, params={"code": "sh600667"}, timeout=20)
print(r.json().get("structured") or r.json()["data"])
# Returns: top holders / shareholder count (with change %) / institutional holdings

• Rising counts = dispersing (more retail, often high-level churn or institutional distribution)
• Falling counts = concentrating (possible accumulation, or just deteriorating liquidity)
• Read it with price position: falling counts at a low vs surging counts at a high mean opposite things

Step 3: events (lock-ups / reductions / dividends)

# All event tags (42 types: lock-up / buyback / placement / dividend / results …)
r = requests.get("https://api.ashareapi.com/v1/events",
                 headers=H, params={"code": "sh600667"}, timeout=20)
print(r.json().get("structured") or r.json()["data"])

# Dividend history (continuity)
r = requests.get("https://api.ashareapi.com/v1/dividend",
                 headers=H, params={"code": "sh600667"}, timeout=20)
print(r.json().get("structured") or r.json()["data"])

• Large lock-up expiries: the bigger the share of free float, the larger the potential supply
• Dividend continuity: steady payers are not automatically good, but an abrupt stop deserves a question
• Events are facts: a lock-up expiry is not a reduction — check announcements and actual filings

The four-step checklist

Four red flags is a different magnitude from one. Screening for landmines rules out extremes; it does not score stocks.

Common mistakes and boundaries

• Calling negative cash flow fraud → growth-stage companies often have negative cash flow; check whether it matches revenue and capex
• Concluding from one quarter → compare multiple periods (seasonality, one-offs)
• Ignoring disclosure cadence → financials update on reporting dates; if a period has not been disclosed, there is no data (not an API problem)
• Units → `amount` fields are generally in CNY; follow each field’s documentation
• This is not investment advice → risk-signal identification only