Fundamental

In Taiwan stock fundamental data, we have 13 datasets, as follows:

• Income Statement TaiwanStockFinancialStatements
• Balance Sheet TaiwanStockBalanceSheet
• Cash Flows Statement TaiwanStockCashFlowsStatement
• Dividend Policy Table TaiwanStockDividend
• Ex-Dividend/Ex-Right Result Table TaiwanStockDividendResult
• Monthly Revenue Table TaiwanStockMonthRevenue
• Capital Reduction Resumption Reference Price TaiwanStockCapitalReductionReferencePrice
• Taiwan Stock Market Value Table TaiwanStockMarketValue
• Taiwan Stock Delisting Table TaiwanStockDelisting
• Taiwan Stock Market Value Weight Table TaiwanStockMarketValueWeight
• Taiwan Stock Post-Split Reference Price TaiwanStockSplitPrice
• Taiwan Stock Par Value Change Resumption Reference Price TaiwanStockParValueChange
• Taiwan Stock Financial Report and Annual Report PDF TaiwanStockFinancialReportFile

---

Income Statement TaiwanStockFinancialStatements

• Data range: 1990-03-01 ~ now
• Coverage: listed (TWSE), OTC (TPEx) and emerging market companies, distinguished by stock_id (use TaiwanStockInfo to look up market type)

Amount unit, currency and how single-quarter values are derived

• Currency: all companies are reported in New Taiwan Dollars (TWD), including foreign (KY) companies. No other currency is used.
• Unit: value is in TWD (dollars). The financial statements report amounts in thousand TWD; this table multiplies them by 1000. EPS is in TWD per share and is not converted.
• Single-quarter values: date is the quarter-end date, and every amount item is a single-quarter figure. Financial statements report year-to-date cumulative amounts; this table subtracts consecutive quarters within the same fiscal year: Q1 = Q1 YTD, Q2 = H1 YTD − Q1 YTD, Q3 = 9-month YTD − H1 YTD, Q4 = full year − 9-month YTD, so the four quarters sum to the full-year figure. For example, 2330's 2024 quarterly Revenue is 592.6 / 673.5 / 759.7 / 868.5 billion TWD, matching the sum of the three months of TaiwanStockMonthRevenue in each quarter.
• Companies that only file semi-annual and annual reports (some emerging-market and public companies): such a year has only two rows, 06-30 and 12-31. The 06-30 row is the first-half total and the 12-31 row is the second-half total (full year − first half). Check whether the year has 03-31 / 09-30 rows to tell.
• Financial companies: banks, financial holding companies, securities firms and insurers use different income statement items. For example, a financial holding company's Revenue maps to "net revenue" (淨收益). Use the origin_name column to see the original item name of each row.

EPS is a single-quarter figure and is retroactively restated after a stock dividend / split — summing quarters will not match the cumulative EPS

The EPS (basic earnings per share) in this table is a single-quarter figure, and each quarter stores the value as originally reported in that quarter's financial statements.

Under IAS 33, share-count changes with no consideration received — stock dividends (bonus issues) and stock splits — require the weighted average number of shares to be retroactively restated for every period presented. So once a bonus issue happens, that quarter's report restates the EPS of all earlier quarters downward, but this table does not rewrite the values already published. The result: within the same fiscal year, the share base behind each quarter's EPS is inconsistent, so summing them will not match the cumulative EPS in the financial statements.

Example — 3081 (Landmark Optoelectronics), FY2026. The company went ex-rights on 2026-07-15 with a stock dividend of NT$1 (100 shares per 1,000 shares held, i.e. a 10% bonus issue). This table reports EPS of 3.44 for 2026 Q1 (as originally reported, pre-bonus share count) and 4.62 for 2026 Q2 (already on the restated share count), which sum to 8.06 — but the reported first-half cumulative EPS is 7.75. The gap arises because Q1's 3.44 was restated in the Q2 report to 3.44 ÷ 1.1 = 3.13, and 3.13 + 4.62 = 7.75.

Recommended approach: to aggregate across quarters, do not sum EPS. Sum IncomeAfterTaxes (net income for the period) instead and divide by a single weighted average share count. For the example above: (317,515,000 + 468,487,000) TWD ÷ approx. 101,420,000 shares = 7.75.

To detect whether a bonus issue occurred within your query range, check the StockEarningsDistribution (stock dividend) field in TaiwanStockDividend; in the example above its value is 1.0.

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_financial_statement(
    stock_id="2330",
    start_date='2019-01-01',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockFinancialStatements",
    "data_id": "2330",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockFinancialStatements",
        data_id="2330",
        start_date= "2019-01-02"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_financial_statement(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	type	value	origin_name
0	2019-03-31	2330	CostOfGoodsSold	1.28352e+11	營業成本
1	2019-03-31	2330	EPS	2.37	基本每股盈餘（元）
2	2019-03-31	2330	EquityAttributableToOwnersOfParent	6.60098e+10	綜合損益總額歸屬於母公司業主
3	2019-03-31	2330	GrossProfit	9.03576e+10	營業毛利（毛損）淨額
4	2019-03-31	2330	IncomeAfterTaxes	6.13873e+10	本期淨利（淨損）

{
    date: str, # date
    stock_id: str, # stock code
    type: str, # category
    value: float64, # value
    origin_name: str # original name
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_financial_statement(
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockFinancialStatements",
    "start_date": "2019-03-31",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data)

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockFinancialStatements",
        start_date= "2019-03-31"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	type	value	origin_name
0	2019-03-31	000116	EPS	0.24	基本每股盈餘（元）
1	2019-03-31	000116	EquityAttributableToOwnersOfParent	4.65569e+08	綜合損益總額歸屬於母公司業主
2	2019-03-31	000116	Expense	8.95498e+08	支出及費用
3	2019-03-31	000116	Income	1.07791e+09	收益
4	2019-03-31	000116	IncomeAfterTaxes	2.74322e+08	本期淨利（淨損）

{
    date: str, # date
    stock_id: str, # stock code
    type: str, # category
    value: float64, # value
    origin_name: str # original name
}

---

Balance Sheet TaiwanStockBalanceSheet

• Data range: 2011-12-01 ~ now
• Coverage: listed (TWSE), OTC (TPEx) and emerging market companies, distinguished by stock_id (use TaiwanStockInfo to look up market type)

Amount unit, currency and the _per items

• Currency / unit: all companies are reported in New Taiwan Dollars (TWD), including foreign (KY) companies. value is in TWD (dollars) — the statements report thousand TWD, and this table multiplies by 1000.
• Point-in-time values: the balance sheet shows balances as of each quarter-end date, so there is no single-quarter vs. cumulative distinction.
• type values ending in _per (e.g. TotalAssets_per): the item as a percentage of total assets, in %, not an amount. TotalAssets_per is 100.

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_balance_sheet(
    stock_id="2330",
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockBalanceSheet",
    "data_id": "2330",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockBalanceSheet",
        data_id="2330",
        start_date= "2019-01-02"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_balance_sheet(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	type	value	origin_name
0	2019-03-31	2330	AccountsPayable	2.71009e+10	應付帳款
1	2019-03-31	2330	AccountsPayable_per	1.24	應付帳款
2	2019-03-31	2330	AccountsPayableToRelatedParties	5.60941e+08	應付帳款－關係人
3	2019-03-31	2330	AccountsPayableToRelatedParties_per	0.03	應付帳款－關係人
4	2019-03-31	2330	AccountsReceivableDuefromRelatedPartiesNet	3.09821e+08	應收帳款－關係人淨額

{
    date: str, # date
    stock_id: str, # stock code
    type: str, # category
    value: float64, # value
    origin_name: str # original name
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_balance_sheet(
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockBalanceSheet",
    "start_date": "2019-03-31",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockBalanceSheet",
        start_date= "2019-03-31"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	type	value	origin_name
0	2019-03-31	1101	AccountsPayable	7.84411e+09	應付帳款
1	2019-03-31	1101	AccountsPayable_per	2.15	應付帳款
2	2019-03-31	1101	AccountsReceivableDuefromRelatedPartiesNet	2.64638e+08	應收帳款－關係人淨額
3	2019-03-31	1101	AccountsReceivableDuefromRelatedPartiesNet_per	0.07	應收帳款－關係人淨額
4	2019-03-31	1101	AccountsReceivableNet	8.3396e+09	應收帳款淨額

{
    date: str, # date
    stock_id: str, # stock code
    type: str, # category
    value: float64, # value
    origin_name: str # original name
}

---

Cash Flows Statement TaiwanStockCashFlowsStatement

• Data range: 2008-06-01 ~ now
• Coverage: listed (TWSE), OTC (TPEx) and emerging market companies, distinguished by stock_id (use TaiwanStockInfo to look up market type)

Amount unit, currency and year-to-date cumulative values

• Currency / unit: all companies are reported in New Taiwan Dollars (TWD), including foreign (KY) companies. value is in TWD (dollars) — the statements report thousand TWD, and this table multiplies by 1000.
• Year-to-date cumulative: unlike the income statement, this table is not converted to single quarters. Each quarter stores the year-to-date cumulative amount as reported, so the 12-31 row is the full-year figure. For example, 2330's 2024 CashFlowsFromOperatingActivities is 436.3 / 814.0 / 1,206.0 / 1,826.2 billion TWD. To get single-quarter values, subtract the previous quarter within the same fiscal year (Q1 needs no subtraction).

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_cash_flows_statement(
    stock_id="2330",
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockCashFlowsStatement",
    "data_id": "2330",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockCashFlowsStatement",
        data_id="2330",
        start_date= "2019-01-02"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_cash_flows_statement(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	type	value	origin_name
0	2019-03-31	2330	HedgingFinancialLiabilities	-2.27383e+08	除列避險之金融負債
1	2019-03-31	2330	CashFlowsFromOperatingActivities	1.5267e+11	營業活動之淨現金流入（流出）
2	2019-03-31	2330	CashProvidedByInvestingActivities	-6.41885e+10	投資活動之淨現金流入(流出)
3	2019-03-31	2330	CashBalancesIncrease	6.78559e+10	本期現金及約當現金增加（減少）數
4	2019-03-31	2330	NetIncomeBeforeTax	6.81817e+10	本期稅前淨利（淨損）

{
    date: str, # date
    stock_id: str, # stock code
    type: str, # category
    value: float64, # value
    origin_name: str # original name
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_cash_flows_statement(
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockCashFlowsStatement",
    "start_date": "2019-03-31",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockCashFlowsStatement",
        start_date= "2019-03-31"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	type	value	origin_name
0	2019-03-31	1101	DecreaseInShortTermLoans	7.59053e+09	短期借款減少
1	2019-03-31	1101	ReceivableIncrease	-1.15069e+08	應收帳款(增加)減少
2	2019-03-31	1101	PropertyAndPlantAndEquipment	-1.48367e+09	取得不動產、廠房及設備
3	2019-03-31	1101	NetIncomeBeforeTax	5.6035e+09	本期稅前淨利（淨損）
4	2019-03-31	1101	CashProvidedByInvestingActivities	-4.31058e+09	投資活動之淨現金流入(流出)

{
    date: str, # date
    stock_id: str, # stock code
    type: str, # category
    value: float64, # value
    origin_name: str # original name
}

---

Dividend Policy Table TaiwanStockDividend

• Data range: 2005-05-01 ~ now

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_dividend(
    stock_id="2330",
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockDividend",
    "data_id": "2330",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockStockDividend",
        data_id="2330",
        start_date= "2019-01-02"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_dividend(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	year	StockEarningsDistribution	StockStatutorySurplus	StockExDividendTradingDate	TotalEmployeeStockDividend	TotalEmployeeStockDividendAmount	RatioOfEmployeeStockDividendOfTotal	RatioOfEmployeeStockDividend	CashEarningsDistribution	CashStatutorySurplus	CashExDividendTradingDate	CashDividendPaymentDate	TotalEmployeeCashDividend	TotalNumberOfCashCapitalIncrease	CashIncreaseSubscriptionRate	CashIncreaseSubscriptionpRrice	RemunerationOfDirectorsAndSupervisors	ParticipateDistributionOfTotalShares	AnnouncementDate	AnnouncementTime
0	2019-06-30	2330	107年	0	0		0	0	0	0	8	0	2019-06-24	2019-07-18	0	0	0	0	0	2.59304e+10	2019-06-06	15:47:30
1	2019-09-25	2330	108年第1季	0	0		0	0	0	0	2	0	2019-09-19	2019-10-17	0	0	0	0	0	2.59304e+10	2019-07-09	18:33:02
2	2019-12-25	2330	108年第2季	0	0		0	0	0	0	2.5	0	2019-12-19	2020-01-16	0	0	0	0	0	2.59304e+10	2019-08-14	15:27:02
3	2020-03-25	2330	108年第3季	0	0		0	0	0	0	2.5	0	2020-03-19	2020-04-16	0	0	0	0	0	2.59304e+10	2019-11-14	17:01:07
4	2020-06-24	2330	108年第4季	0	0		0	0	0	0	2.5	0	2020-06-18	2020-07-16	0	0	0	0	0	2.59304e+10	2020-02-14	15:10:50

{
    date: str, # rights distribution record date
    stock_id: str, # stock code
    year: str, # dividend fiscal year
    StockEarningsDistribution: float64, # stock dividend: capitalization of earnings
    StockStatutorySurplus: float64, # stock dividend: capitalization of statutory and capital surplus
    StockExDividendTradingDate: str, # ex-rights trading date
    TotalEmployeeStockDividend: float64, # employee stock dividend
    TotalEmployeeStockDividendAmount: float64, # employee stock dividend amount
    RatioOfEmployeeStockDividendOfTotal: float64, # ratio of employee stock dividend to total earnings dividend
    RatioOfEmployeeStockDividend: float64, # employee stock dividend ratio
    CashEarningsDistribution: float64, # cash dividend: capitalization of earnings
    CashStatutorySurplus: float64, # cash dividend: capitalization of statutory and capital surplus
    CashExDividendTradingDate: str, # ex-dividend trading date
    CashDividendPaymentDate: str, # cash dividend payment date
    TotalEmployeeCashDividend: float64, # total employee cash bonus
    TotalNumberOfCashCapitalIncrease: float64, # total shares of cash capital increase
    CashIncreaseSubscriptionRate: float64, # cash capital increase subscription ratio
    CashIncreaseSubscriptionpRrice: float64, # cash capital increase subscription price
    RemunerationOfDirectorsAndSupervisors: float64, # directors and supervisors remuneration
    ParticipateDistributionOfTotalShares: float64, # total shares participating in distribution
    AnnouncementDate: str, # announcement date
    AnnouncementTime: str # announcement time
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_dividend(
    start_date='2025-10-06',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockDividend",
    "start_date": "2025-10-06",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockStockDividend",
        start_date= "2025-10-06"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	year	StockEarningsDistribution	StockStatutorySurplus	StockExDividendTradingDate	TotalEmployeeStockDividend	TotalEmployeeStockDividendAmount	RatioOfEmployeeStockDividendOfTotal	RatioOfEmployeeStockDividend	CashEarningsDistribution	CashStatutorySurplus	CashExDividendTradingDate	CashDividendPaymentDate	TotalEmployeeCashDividend	TotalNumberOfCashCapitalIncrease	CashIncreaseSubscriptionRate	CashIncreaseSubscriptionpRrice	RemunerationOfDirectorsAndSupervisors	ParticipateDistributionOfTotalShares	AnnouncementDate	AnnouncementTime
0	2025-10-06	2540	113年	3	0.999999	2025-09-30	0	0	0	0	0.4	0.6	2025-09-30	2025-10-31	0	0	0	0	0	6.7491e+08	2025-09-12	17:35:32
1	2025-10-06	3312	不適用	0	0	2025-09-30	0	0	0	0	0	0			0	2e+07	9.84	39.8	0	1.62627e+08	2025-09-22	16:52:52

{
    date: str, # rights distribution record date
    stock_id: str, # stock code
    year: str, # dividend fiscal year
    StockEarningsDistribution: float64, # stock dividend: capitalization of earnings
    StockStatutorySurplus: float64, # stock dividend: capitalization of statutory and capital surplus
    StockExDividendTradingDate: str, # ex-rights trading date
    TotalEmployeeStockDividend: float64, # employee stock dividend
    TotalEmployeeStockDividendAmount: float64, # employee stock dividend amount
    RatioOfEmployeeStockDividendOfTotal: float64, # ratio of employee stock dividend to total earnings dividend
    RatioOfEmployeeStockDividend: float64, # employee stock dividend ratio
    CashEarningsDistribution: float64, # cash dividend: capitalization of earnings
    CashStatutorySurplus: float64, # cash dividend: capitalization of statutory and capital surplus
    CashExDividendTradingDate: str, # ex-dividend trading date
    CashDividendPaymentDate: str, # cash dividend payment date
    TotalEmployeeCashDividend: float64, # total employee cash bonus
    TotalNumberOfCashCapitalIncrease: float64, # total shares of cash capital increase
    CashIncreaseSubscriptionRate: float64, # cash capital increase subscription ratio
    CashIncreaseSubscriptionpRrice: float64, # cash capital increase subscription price
    RemunerationOfDirectorsAndSupervisors: float64, # directors and supervisors remuneration
    ParticipateDistributionOfTotalShares: float64, # total shares participating in distribution
    AnnouncementDate: str, # announcement date
    AnnouncementTime: str # announcement time
}

---

Ex-Dividend/Ex-Right Result Table TaiwanStockDividendResult

• Data range: 2003-05-01 ~ now

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_dividend_result(
    stock_id="2330",
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockDividendResult",
    "data_id": "2330",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockDividendResult",
        data_id="2330",
        start_date= "2019-01-02"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_dividend_result(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	before_price	after_price	stock_and_cache_dividend	stock_or_cache_dividend	max_price	min_price	open_price	reference_price
0	2019-06-24	2330	248.5	240.5	8	息	264.5	216.5	240.5	240.5
1	2019-09-19	2330	267	265	2	息	291.5	238.5	265	265
2	2019-12-19	2330	344.5	342	2.5	息	376	308	342	342
3	2020-03-19	2330	260	257.5	2.5	息	283	232	257.5	257.5
4	2020-06-18	2330	315	312.5	2.5	息	343.5	281.5	312.5	312.5

{
    date: str, # date
    stock_id: str, # stock code
    before_price: float32, # closing price before ex-dividend/ex-right
    after_price: float32, # closing price after ex-dividend/ex-right
    stock_and_cache_dividend: float32, # dividend value
    stock_or_cache_dividend: float32, # right/dividend
    max_price: float32, # limit-up price
    min_price: float32, # limit-down price
    open_price: float32, # open price
    reference_price: float32 # reference price after dividend deduction
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_dividend_result(
    start_date='2019-06-24',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockDividendResult",
    "start_date": "2019-06-24",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockDividendResult",
        start_date= "2019-06-24"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	before_price	after_price	stock_and_cache_dividend	stock_or_cache_dividend	max_price	min_price	open_price	reference_price
0	2019-06-24	00697B	42.48	42.21	0.27	除息	9999.95	0.01	42.21	42.21
1	2019-06-24	00751B	46.05	45.46	0.59	除息	9999.95	0.01	45.46	45.46
2	2019-06-24	1707	220	213.5	6.5	息	234.5	192.5	213.5	213.5
3	2019-06-24	1711	17	16.5	0.5	息	18.15	14.85	16.5	16.5
4	2019-06-24	1906	13.55	13.05	0.5	息	14.35	11.75	13.05	13.05

{
    date: str, # date
    stock_id: str, # stock code
    before_price: float32, # closing price before ex-dividend/ex-right
    after_price: float32, # closing price after ex-dividend/ex-right
    stock_and_cache_dividend: float32, # dividend value
    stock_or_cache_dividend: float32, # right/dividend
    max_price: float32, # limit-up price
    min_price: float32, # limit-down price
    open_price: float32, # open price
    reference_price: float32 # reference price after dividend deduction
}

---

Monthly Revenue Table TaiwanStockMonthRevenue

• Data range: 2002-02-01 ~ now
• Coverage: listed (TWSE), OTC (TPEx) and emerging market companies, distinguished by stock_id (use TaiwanStockInfo to look up market type)

create_time only exists from 2026-04-21 onwards, and reflects when the row entered the FinMind database

create_time is the date (YYYY-MM-DD) on which the monthly revenue row entered the FinMind database. It has only been recorded since 2026-04-21; rows older than that carry no create_time and the field is an empty string.

It is therefore not the official announcement time filed by the company on the Market Observation Post System - it only tells you when FinMind observed the row. That said, FinMind collects monthly revenue several times a day, so in practice the gap to the announcement time should be negligible, and the field can still be used as an approximation of "roughly when this revenue figure was published".

Note also that rows written on 2026-04-21 (the day the column went live) all carry create_time = 2026-04-21, which is the initial backfilled value of the column rather than the day those figures were announced.

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_month_revenue(
    stock_id="2330",
    start_date='2019-03-31',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockMonthRevenue",
    "data_id": "2330",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockMonthRevenue",
        data_id="2330",
        start_date= "2019-01-02"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_month_revenue(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	country	revenue	revenue_month	revenue_year	create_time
0	2019-04-01	2330	Taiwan	79721587000	3	2019
1	2019-05-01	2330	Taiwan	74693615000	4	2019
2	2019-06-01	2330	Taiwan	80436931000	5	2019
3	2019-07-01	2330	Taiwan	85867929000	6	2019
4	2019-08-01	2330	Taiwan	84757724000	7	2019

{
    date: str, # date
    stock_id: str, # stock code
    country: str, # country
    revenue: int64, # revenue
    revenue_month: int64, # revenue month
    revenue_year: int64, # revenue year
    create_time: str # creation time (YYYY-MM-DD); recorded only since 2026-04-21, empty string for older historical rows
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_month_revenue(
    start_date='2019-04-01',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockMonthRevenue",
    "start_date": "2019-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockMonthRevenue",
        start_date= "2019-01-01"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	country	revenue	revenue_month	revenue_year	create_time
0	2019-05-01	1101	Taiwan	10596314000	4	2019
1	2019-05-01	1102	Taiwan	8434811000	4	2019
2	2019-05-01	1103	Taiwan	160751000	4	2019
3	2019-05-01	1104	Taiwan	418992000	4	2019
4	2019-05-01	1108	Taiwan	323834000	4	2019

{
    date: str, # date
    stock_id: str, # stock code
    country: str, # country
    revenue: int64, # revenue
    revenue_month: int64, # revenue month
    revenue_year: int64, # revenue year
    create_time: str # creation time (YYYY-MM-DD); recorded only since 2026-04-21, empty string for older historical rows
}

---

Capital Reduction Resumption Reference Price TaiwanStockCapitalReductionReferencePrice

• Data range: 2011-01-01 ~ now

data_id (stock id) is not required

Omitting data_id and providing only the date range returns the capital reduction resumption reference prices of all stocks within that range. This whole-market query without data_id is available to backer/sponsor members; free members should query a single stock with data_id.

Example

PythonRPackage-Async

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockCapitalReductionReferencePrice",
    "data_id": "2327",
    "start_date": "2010-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

        date stock_id  ClosingPriceonTheLastTradingDay  PostReductionReferencePrice  LimitUp  LimitDown  OpeningReferencePrice  ExrightReferencePrice ReasonforCapitalReduction
0  2013-09-18     2327                            10.20                        10.28    10.95       9.57                   10.3                   -1.0               Cash refund
1  2014-10-09     2327                            22.05                        49.82    53.30      46.35                   49.8                   -1.0               Cash refund
2  2016-08-15     2327                            54.80                        65.96    72.50      59.40                   66.0                   -1.0               Cash refund
3  2017-08-18     2327                           120.50                       168.13   184.50     151.50                  168.0                   -1.0               Cash refund

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockCapitalReductionReferencePrice",
        data_id="2327",
        start_date= "2010-01-01"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

        date stock_id  ClosingPriceonTheLastTradingDay  PostReductionReferencePrice  LimitUp  LimitDown  OpeningReferencePrice  ExrightReferencePrice ReasonforCapitalReduction
1  2013-09-18     2327                            10.20                        10.28    10.95       9.57                   10.3                   -1.0               Cash refund
2  2014-10-09     2327                            22.05                        49.82    53.30      46.35                   49.8                   -1.0               Cash refund
3  2016-08-15     2327                            54.80                        65.96    72.50      59.40                   66.0                   -1.0               Cash refund
4  2017-08-18     2327                           120.50                       168.13   184.50     151.50                  168.0                   -1.0               Cash refund

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_capital_reduction_reference_price(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

---

Taiwan Stock Market Value Table TaiwanStockMarketValue (available only to backer, sponsor members)

• Data range: 2004-01-01 ~ now
• Data update time: Monday to Friday 23:30. The actual update time is based on the API data.

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_market_value(
    stock_id='2330',
    start_date='2023-01-01',
    end_date='2024-01-01'
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockMarketValue",
    "data_id": "2330",
    "start_date": "2023-01-01",
    "end_date": "2024-01-01",
}
resp = requests.get(url, headers=headers, params=parameter)
data = resp.json()
data = pd.DataFrame(data["data"])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
    response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockMarketValue",
        data_id= "2330",
        start_date= "2023-01-01",
        end_date= "2024-01-01"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	market_value
0	2023-01-03	2330	1.174646e+13
1	2023-01-04	2330	1.165571e+13
2	2023-01-05	2330	1.188908e+13
3	2023-01-06	2330	1.188908e+13
4	2023-01-09	2330	1.247251e+13

{
    date: str, # date
    stock_id: str, # stock code
    market_value: int64 # market value
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_daily(
    start_date='2023-01-03',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockMarketValue",
    "start_date": "2023-01-03",
}
resp = requests.get(url, headers=headers, params=parameter)
data = resp.json()
data = pd.DataFrame(data["data"])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockMarketValue",
        start_date= "2023-01-03"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	market_value
0	2023-01-03	0050	2.561094e+11
1	2023-01-03	0051	7.967000e+08
2	2023-01-03	0052	5.644650e+09
3	2023-01-03	0053	2.611218e+08
4	2023-01-03	0055	1.625804e+09

{
    date: str, # date
    stock_id: str, # stock code
    market_value: int64 # market value
}

---

Taiwan Stock Delisting Table TaiwanStockDelisting

• Data range: 2001-01-01 ~ now
• Data update time: Monday to Friday 23:30. The actual update time is based on the API data.

Example

PackagePython-requestRPackage-Async

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_delisting()

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockDelisting",
}
resp = requests.get(url, headers=headers, params=parameter)
data = resp.json()
data = pd.DataFrame(data["data"])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockDelisting"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

from FinMind.data import DataLoader
from loguru import logger
import datetime

api = DataLoader()
api.login_by_token(api_token='token')

start = datetime.datetime.now()
df = api.taiwan_stock_delisting(
    stock_id_list=['2330', '2317', '2454', '3008'],
    start_date='2024-01-01',
    end_date='2024-12-31',
    use_async=True,
)
cost = datetime.datetime.now() - start
logger.info(cost)

Output

DataFrameSchema

	date	stock_id	stock_name
0	2005-10-04	1204	津津
1	2001-11-01	1230	聯成食品
2	2005-10-04	1306	合發興業
3	2006-06-26	1408	中興紡織
4	2002-11-08	1431	新燕實業

{
    date: str, # date
    stock_id: str, # stock code
    stock_name: str # stock name
}

---

Taiwan Stock Market Value Weight Table TaiwanStockMarketValueWeight (available only to backer, sponsor members)

• Data range: 2024-10-30 ~ now
• Data update time: 23:45 on the 1st, 2nd, 3rd, 28th, 29th, 30th, and 31st of each month. The actual update time is based on the API data.

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_market_value_weight(
    stock_id='2330',
    start_date='2024-01-01',
    end_date='2025-01-01'
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockMarketValueWeight",
    "data_id": "2330",
    "start_date": "2024-01-01",
    "end_date": "2025-01-01",
}
resp = requests.get(url, headers=headers, params=parameter)
data = resp.json()
data = pd.DataFrame(data["data"])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockMarketValueWeight",
        data_id= "2330",
        start_date= "2024-01-01",
        end_date= "2025-01-01"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	rank	stock_id	stock_name	weight_per	date	type
0	1	2330	台積電	36.8397	2024-10-30	twse

{
    rank: int64, # rank
    stock_id: str, # stock code
    stock_name: str, # stock name
    weight_per: float32, # weight percentage
    date: str, # date
    type: str # listed (twse) / OTC (tpex)
}

Fetch all data for a specific date at once (available only to backer, sponsor members)

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_market_value_weight(
    start_date='2024-10-30',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockMarketValueWeight",
    "start_date": "2024-10-30",
}
resp = requests.get(url, headers=headers, params=parameter)
data = resp.json()
data = pd.DataFrame(data["data"])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockMarketValueWeight",
        start_date= "2024-10-30"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	rank	stock_id	stock_name	weight_per	date	type
0	43	1101	台泥	0.3327	2024-10-30	twse
0	63	1102	亞泥	0.2282	2024-10-30	twse
0	394	1103	嘉泥	0.0192	2024-10-30	twse
0	305	1104	環泥	0.0286	2024-10-30	twse
0	651	1108	幸福	0.0082	2024-10-30	twse

{
    rank: int64, # rank
    stock_id: str, # stock code
    stock_name: str, # stock name
    weight_per: float32, # weight percentage
    date: str, # date
    type: str # listed (twse) / OTC (tpex)
}

---

Taiwan Stock Post-Split Reference Price TaiwanStockSplitPrice

• Provides Taiwan stock post-split reference prices.
• Data update time: Monday to Friday 18:00. The actual update time is based on the API data.

Example

Python-requestR

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockSplitPrice",
}
resp = requests.get(url, headers=headers, params=parameter)
data = resp.json()
data = pd.DataFrame(data["data"])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockSplitPrice"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	type	before_price	after_price	max_price	min_price	open_price
0	2024-12-11	00632R	反分割	3.28	22.96	25.25	20.67	22.96
1	2025-02-19	00676R	反分割	2.04	12.23	13.45	11.01	12.23
2	2025-06-11	00663L	分割	170.15	24.3	29.16	19.44	24.3
3	2025-06-18	0050	分割	188.65	47.16	51.85	42.45	47.16

{
    date: str, # split date
    stock_id: str, # stock code
    type: str, # split type
    before_price: float, # pre-split price
    after_price: float, # post-split price
    max_price: float, # post-split max price
    min_price: float, # post-split min price
    open_price: float # post-split open price
}

---

Taiwan Stock Par Value Change Resumption Reference Price TaiwanStockParValueChange

• Data range: 2020-01-01 ~ now

Example

PackagePython-requestR

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
df = api.taiwan_stock_par_value_change(
    start_date='2020-01-01',
)

import requests
import pandas as pd
url = "https://api.finmindtrade.com/api/v4/data"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockParValueChange",
    "start_date": "2020-01-01",
}
data = requests.get(url, headers=headers, params=parameter)
data = data.json()
data = pd.DataFrame(data['data'])
print(data.head())

library(httr)
library(data.table)
library(dplyr)
url = 'https://api.finmindtrade.com/api/v4/data'
token = "" # Refer to login to obtain the token
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockParValueChange",
        start_date= "2020-01-01"
    ),
    add_headers(Authorization = paste("Bearer", token))
)
data = content(response)
df = data$data %>%
do.call('rbind',.) %>%
data.table
head(df)

Output

DataFrameSchema

	date	stock_id	stock_name	before_close	after_ref_close	after_ref_max	after_ref_min	after_ref_open
0	2020-08-17	8070	長華	190	19	20.9	17.1	19
1	2021-10-18	6531	愛普	750	375	412.5	337.5	375
2	2022-07-13	6415	矽力-K	2485	621.25	683	560	621
3	2024-11-11	8476	台境	58.8	29.4	32.3	26.5	29.4
4	2025-06-30	4763	材料-KY	885	88.5	97.3	79.7	88.5

{
    date: str, # date
    stock_id: str, # stock code
    stock_name: str, # stock name
    before_close: float64, # closing price before suspension
    after_ref_close: float64, # resumption reference price
    after_ref_max: float64, # limit-up price
    after_ref_min: float64, # limit-down price
    after_ref_open: float64 # opening auction reference price
}

---

Taiwan Stock Financial Report and Annual Report PDF TaiwanStockFinancialReportFile (available only to sponsorpro members)

• Provides the full PDF files of financial reports and annual shareholders' meeting reports of TWSE-listed, TPEx-listed and emerging-market companies.
• Data range: financial reports 2013-03-31 ~ now (from 2013 Q1, one per quarter); annual reports 2013-12-31 ~ now (from fiscal year 2013, one per year)
• Download one file per request via /api/v4/storage_objects with data_id (stock id), year (calendar year), quarter (1 to 4) and report_category; all four are required. Whole-market downloads are not supported.
• quarter=1 to 3 is the financial report for that quarter; quarter=4 has the annual financial report (financial_report) and the annual shareholders' meeting report (annual_report).

Download parameters and response

Parameter	Description
dataset	TaiwanStockFinancialReportFile
data_id	Stock id, required
year	Calendar year, required
quarter	Quarter 1 to 4, required
report_category	financial_report (financial report) or annual_report (annual shareholders' meeting report), required; annual_report is only available for quarter=4

• The response is a redirect; follow it to download the file itself.
• Downloaded file name format: TaiwanStockFinancialReportFile_{data_id}_{date}_{report_category}.pdf (date is the period end date, e.g. 2025-12-31).
• Returns 404 when there is no matching file for that period.
• Returns 400 when parameters are missing or invalid (e.g. quarter not 1 to 4, report_category not one of the two values above, or annual_report with a quarter other than 4).

Provides PDF files, not financial statement figures

This dataset provides the report PDF files. For financial statement figures such as income, balance sheet and cash flow items, please use Income Statement TaiwanStockFinancialStatements, Balance Sheet TaiwanStockBalanceSheet and Cash Flows Statement TaiwanStockCashFlowsStatement.

Coverage

• Financial reports (financial_report): one Chinese report per period, primarily the consolidated financial report; when no consolidated financial report is available, the individual, restated, or parent-company-only financial report is included, in that order. English versions are not included.
• Files are mostly PDF; for a small number of 2013 financial reports, the source only provides a Word file, so the downloaded file is a .doc file.
• Annual reports (annual_report) are the Chinese annual shareholders' meeting reports.
• The text layer of some annual report PDFs may be garbled or incomplete (as in the source files); OCR may be needed if you need to extract text.

Usage scope

Rights in the files belong to the issuing company. See Commercial Use Q&A.

Read reports with AI (MCP)

With the FinMind MCP Server (version 0.1.1 or later) you don't need to write download code: just ask "What is the inventory breakdown in the notes of TSMC's Q2 2026 financial report?" and the AI uses download_financial_report to download the PDF, check its table of contents and read the relevant pages.

Example

PackagePython-requestRMultiple periods

from FinMind.data import DataLoader

api = DataLoader()
# api.login_by_token(api_token='token')
# Downloads one file and returns its content (bytes); raises an exception when there is no file for the period
content = api.taiwan_stock_financial_report_file_download(
    stock_id="2330",
    year=2025,
    quarter=4,
    report_category="annual_report",
)
with open("TaiwanStockFinancialReportFile_2330_2025-12-31_annual_report.pdf", "wb") as f:
    f.write(content)

import requests

url = "https://api.finmindtrade.com/api/v4/storage_objects"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
parameter = {
    "dataset": "TaiwanStockFinancialReportFile",
    "data_id": "2330",
    "year": 2025,
    "quarter": 4,
    "report_category": "annual_report",
}
# requests follows redirects by default and gets the file itself
resp = requests.get(url, headers=headers, params=parameter)
resp.raise_for_status()  # 404 when there is no file for the period
# A small number of 2013 financial reports are Word files (.doc); others are PDF
ext = "doc" if resp.headers.get("Content-Type") == "application/msword" else "pdf"
with open(f"TaiwanStockFinancialReportFile_2330_2025-12-31_annual_report.{ext}", "wb") as f:
    f.write(resp.content)

library(httr)

url = 'https://api.finmindtrade.com/api/v4/storage_objects'
token = "" # Refer to login to obtain the token
# httr follows redirects by default; write_disk writes the file to disk
response = httr::GET(
    url = url,
    query = list(
        dataset="TaiwanStockFinancialReportFile",
        data_id= "2330",
        year= 2025,
        quarter= 4,
        report_category= "annual_report"
    ),
    add_headers(Authorization = paste("Bearer", token)),
    write_disk("TaiwanStockFinancialReportFile_2330_2025-12-31_annual_report.pdf", overwrite = TRUE)
)
stop_for_status(response) # 404 when there is no file for the period

import requests

url = "https://api.finmindtrade.com/api/v4/storage_objects"
token = "" # Refer to login to obtain the token
headers = {"Authorization": f"Bearer {token}"}
quarter_end = {1: "03-31", 2: "06-30", 3: "09-30", 4: "12-31"}

# Download 2330's quarterly financial reports and annual reports for 2024-2025
for year in [2024, 2025]:
    for quarter in [1, 2, 3, 4]:
        categories = ["financial_report"]
        if quarter == 4:
            categories.append("annual_report")  # annual reports are Q4 only
        for report_category in categories:
            parameter = {
                "dataset": "TaiwanStockFinancialReportFile",
                "data_id": "2330",
                "year": year,
                "quarter": quarter,
                "report_category": report_category,
            }
            resp = requests.get(url, headers=headers, params=parameter)
            if resp.status_code == 404:
                print(year, quarter, report_category, "no file for this period")
                continue
            resp.raise_for_status()
            ext = "doc" if resp.headers.get("Content-Type") == "application/msword" else "pdf"
            file_name = (
                f"TaiwanStockFinancialReportFile_2330_"
                f"{year}-{quarter_end[quarter]}_{report_category}.{ext}"
            )
            with open(file_name, "wb") as f:
                f.write(resp.content)
            print(file_name, len(resp.content))