Xtquant

XtQuant (QMT Python SDK by ThinkTrader)

XtQuant is the Python SDK for the QMT/miniQMT quantitative trading platform, developed by ThinkTrader (XunTou Technology). It contains two core modules:

• xtdata — Market Data Module: real-time quotes, historical K-lines, tick data, Level 2 data, financial data, sector management
• xttrade — Trading Module: order placement, position/order queries, account management, margin trading, futures/options, smart algorithms

⚠️ Requires miniQMT or QMT client running on Windows. XtQuant connects to the QMT process via local TCP. You need QMT/miniQMT access enabled by your broker.

Installation

pip install xtquant

You can also download from the official website: http://dict.thinktrader.net/nativeApi/download_xtquant.html

Architecture Overview

Your Python script (any IDE, any Python version)
    ↓ (xtquant SDK, pip install)
    ├── xtdata  → miniQMT (market data service, TCP connection)
    └── xttrade → miniQMT (trading service, TCP connection)
         ↓
    Broker trading system

Core Module Reference

Quick Start — Market Data

from xtquant import xtdata

# Connect to local miniQMT (default: localhost)
xtdata.connect()

# Download historical K-line data (must download to local cache before first access)
xtdata.download_history_data('000001.SZ', '1d', start_time='20240101', end_time='20240630')

# Get local K-line data (returns a dict of DataFrames keyed by stock code)
data = xtdata.get_market_data_ex(
    [],                    # field_list, empty list means all fields
    ['000001.SZ'],         # stock_list, list of stock codes
    period='1d',
    start_time='20240101',
    end_time='20240630',
    dividend_type='front'  # Adjustment type: none (unadjusted), front (forward-adjusted), back (backward-adjusted), front_ratio (proportional forward), back_ratio (proportional backward)
)
print(data['000001.SZ'])

Real-Time Quote Subscription

def on_data(datas):
    """Quote data callback function, receives pushed real-time data"""
    for stock_code, data in datas.items():
        print(stock_code, data)

# Subscribe to tick data for a single stock
xtdata.subscribe_quote('000001.SZ', period='tick', callback=on_data)

# Subscribe to full-market quote push
xtdata.subscribe_whole_quote(['SH', 'SZ'], callback=on_data)

xtdata.run()  # Block the current thread, continuously receiving callback data

Financial Data

# First download financial data to local cache
xtdata.download_financial_data(['000001.SZ'])
# Then retrieve financial data from local cache
data = xtdata.get_financial_data(['000001.SZ'])
# Available financial reports: Balance (balance sheet), Income (income statement), CashFlow (cash flow statement),
# PershareIndex (per-share indicators), CapitalStructure (capital structure), TOP10HOLDER (top 10 shareholders),
# TOP10FLOWHOLDER (top 10 tradable shareholders), SHAREHOLDER (shareholder count)

Instrument Info & Sectors

# Get detailed instrument info (name, price limits, tick size, etc.)
info = xtdata.get_instrument_detail('000001.SZ')
# Get security type (stock/index/fund/bond, etc.)
itype = xtdata.get_instrument_type('000001.SZ')
# Get list of stocks in a sector
stocks = xtdata.get_stock_list_in_sector('沪深A股')
# Get list of trading dates
days = xtdata.get_trading_dates('SH', start_time='20240101', end_time='20240630')

Quick Start — Trading

from xtquant import xtconstant
from xtquant.xttrader import XtQuantTrader, XtQuantTraderCallback
from xtquant.xttype import StockAccount

# Create a trader instance (path points to miniQMT's userdata_mini directory)
path = r'D:\国金证券QMT交易端\userdata_mini'
session_id = 123456  # Each strategy must use a unique session_id
xt_trader = XtQuantTrader(path, session_id)

# Register a callback class to receive real-time push notifications
class MyCallback(XtQuantTraderCallback):
    def on_disconnected(self):
        print('Disconnected')
    def on_stock_order(self, order):
        print(f'Order update: {order.stock_code} status={order.order_status}')
    def on_stock_trade(self, trade):
        print(f'Trade update: {trade.stock_code} {trade.traded_volume}@{trade.traded_price}')
    def on_order_error(self, order_error):
        print(f'Order error: {order_error.error_msg}')
    def on_order_stock_async_response(self, response):
        print(f'Async order response: order_id={response.order_id}')

xt_trader.register_callback(MyCallback())
xt_trader.start()
connect_result = xt_trader.connect()  # Returns 0 on successful connection

# Create an account object and subscribe to push notifications
account = StockAccount('your_account_id')
xt_trader.subscribe(account)  # Enable push notifications for this account

# Place order: limit buy 600000.SH, 1000 shares at price 10.5
order_id = xt_trader.order_stock(
    account, '600000.SH', xtconstant.STOCK_BUY, 1000,
    xtconstant.FIX_PRICE, 10.5, 'strategy1', 'test_order'
)
# Returns order_id > 0 on success, -1 on failure

# Query positions
positions = xt_trader.query_stock_positions(account)
for pos in positions:
    print(pos.stock_code, pos.volume, pos.can_use_volume, pos.market_value)

# Query orders
orders = xt_trader.query_stock_orders(account)

# Query assets
asset = xt_trader.query_stock_asset(account)
print(f'Available cash: {asset.cash}, Total assets: {asset.total_asset}')

# Cancel order
xt_trader.cancel_order_stock(account, order_id)

# Block the main thread, waiting for callbacks
xt_trader.run_forever()

Stock Code Format

Data Periods

tick, 1m, 5m, 15m, 30m, 1h, 1d, 1w, 1mon

Supported Asset Types

Order Type Constants (xtconstant)

Account Types

StockAccount('id')            # Regular stock account
StockAccount('id', 'CREDIT')  # Credit account (margin trading)
StockAccount('id', 'FUTURE')  # Futures account

xtdata Interface Pattern

The market data module follows a unified download → retrieve pattern:

1. Subscribe (subscribe): subscribe_quote, subscribe_whole_quote — real-time push
2. Download (download): download_history_data, download_financial_data — download from server to local cache (synchronous/blocking)
3. Retrieve (get): get_market_data_ex, get_financial_data — read from local cache (fast)

xttrade Callback System

Register an XtQuantTraderCallback subclass to receive real-time push notifications:

Advanced Features

• Smart Algorithm Trading: Execute algorithmic orders such as VWAP via smart_algo_order_async
• Securities Lending: Query available securities, apply for lending, manage contracts
• Bank-Securities Transfer: Transfer funds between bank and securities accounts
• CTP Internal Transfer: Transfer funds between futures and options accounts
• Custom Sectors: Create, manage, and query custom stock groups
• Level 2 Data: l2quote, l2order, l2transaction, l2quoteaux, l2orderqueue, l2thousand (1000-level order book), limitupperformance (consecutive limit-up tracking), snapshotindex, hfiopv, fullspeedorderbook

Usage Tips

• miniQMT must be running on Windows — xtquant connects via local TCP.
• session_id must be unique per strategy — different strategies need different IDs.
• connect() is a one-time connection — it does not auto-reconnect after disconnection; you must call it again manually.
• Always call subscribe(account) to receive trading push callbacks.
• Data is cached locally after download — subsequent reads are extremely fast.
• Use dividend_type='front' to get forward-adjusted K-line data.
• In push callbacks, use async query methods to avoid deadlocks.
• Documentation: http://dict.thinktrader.net/nativeApi/start_now.html

Advanced Examples

Batch Download Full-Market Daily K-Line Data

from xtquant import xtdata

xtdata.connect()

# Get the full list of Shanghai & Shenzhen A-shares
stock_list = xtdata.get_stock_list_in_sector('沪深A股')
print(f"Total {len(stock_list)} A-shares")

# Batch download daily K-line data (recommended to download in batches to avoid timeout)
batch_size = 50
for i in range(0, len(stock_list), batch_size):
    batch = stock_list[i:i+batch_size]
    for stock in batch:
        try:
            xtdata.download_history_data(stock, '1d', start_time='20240101', end_time='20240630')
        except Exception as e:
            print(f"Failed to download {stock}: {e}")
    print(f"Downloaded {min(i+batch_size, len(stock_list))}/{len(stock_list)}")

# Batch retrieve data
data = xtdata.get_market_data_ex(
    [], stock_list[:10], period='1d',
    start_time='20240101', end_time='20240630',
    dividend_type='front'
)
for code, df in data.items():
    print(f"{code}: {len(df)} records, latest close={df['close'].iloc[-1]}")

Real-Time Quote Monitoring + Conditional Order Trigger

from xtquant import xtdata, xtconstant
from xtquant.xttrader import XtQuantTrader, XtQuantTraderCallback
from xtquant.xttype import StockAccount
import threading

# === Trading Callbacks ===
class MyCallback(XtQuantTraderCallback):
    def on_stock_order(self, order):
        print(f'Order: {order.stock_code} status={order.order_status} {order.status_msg}')
    def on_stock_trade(self, trade):
        print(f'Trade: {trade.stock_code} {trade.traded_volume}@{trade.traded_price}')
    def on_order_error(self, error):
        print(f'Error: {error.error_msg}')

# === Initialize Trading ===
path = r'D:\券商QMT\userdata_mini'
xt_trader = XtQuantTrader(path, 888888)
xt_trader.register_callback(MyCallback())
xt_trader.start()
xt_trader.connect()
account = StockAccount('your_account')
xt_trader.subscribe(account)

# === Quote Monitoring Parameters ===
target_stock = '000001.SZ'
buy_price = 10.50    # Target buy price
sell_price = 11.50   # Target sell price
bought = False

def on_tick(datas):
    """Real-time tick callback: automatically places orders when price hits target"""
    global bought
    for code, tick in datas.items():
        price = tick['lastPrice']
        print(f'{code}: latest price={price}')

        # Price drops to or below target buy price, buy
        if price <= buy_price and not bought:
            order_id = xt_trader.order_stock(
                account, code, xtconstant.STOCK_BUY, 100,
                xtconstant.FIX_PRICE, buy_price, 'auto_buy', '条件触发买入'
            )
            print(f'Buy triggered: order_id={order_id}')
            bought = True

        # Price rises to or above target sell price, sell
        elif price >= sell_price and bought:
            order_id = xt_trader.order_stock(
                account, code, xtconstant.STOCK_SELL, 100,
                xtconstant.FIX_PRICE, sell_price, 'auto_sell', '条件触发卖出'
            )
            print(f'Sell triggered: order_id={order_id}')
            bought = False

# === Start quote subscription (separate thread) ===
xtdata.connect()
def run_data():
    xtdata.subscribe_quote(target_stock, period='tick', callback=on_tick)
    xtdata.run()

t = threading.Thread(target=run_data, daemon=True)
t.start()

# Keep the main thread running
xt_trader.run_forever()

Multi-Stock Moving Average Strategy

from xtquant import xtdata, xtconstant
from xtquant.xttrader import XtQuantTrader, XtQuantTraderCallback
from xtquant.xttype import StockAccount
import pandas as pd

xtdata.connect()

# Define stock pool
stock_pool = ['000001.SZ', '600036.SH', '601318.SH', '000858.SZ', '300750.SZ']

# Download historical data
for stock in stock_pool:
    xtdata.download_history_data(stock, '1d', start_time='20240101', end_time='20241231')

# Retrieve data and compute signals
signals = {}
for stock in stock_pool:
    data = xtdata.get_market_data_ex([], [stock], period='1d',
        start_time='20240101', end_time='20241231', dividend_type='front')
    df = data[stock]

    # Calculate 5-day and 20-day moving averages
    df['ma5'] = df['close'].rolling(5).mean()
    df['ma20'] = df['close'].rolling(20).mean()

    # Determine the latest signal
    if len(df) >= 21:
        latest = df.iloc[-1]
        prev = df.iloc[-2]
        if prev['ma5'] <= prev['ma20'] and latest['ma5'] > latest['ma20']:
            signals[stock] = 'BUY'   # Golden cross
        elif prev['ma5'] >= prev['ma20'] and latest['ma5'] < latest['ma20']:
            signals[stock] = 'SELL'  # Death cross
        else:
            signals[stock] = 'HOLD'  # Hold

print("Trading signals:")
for stock, signal in signals.items():
    print(f"  {stock}: {signal}")

Retrieve Financial Data and Screen Stocks

from xtquant import xtdata

xtdata.connect()

# Get the list of Shanghai & Shenzhen A-shares
stock_list = xtdata.get_stock_list_in_sector('沪深A股')

# Download financial data
xtdata.download_financial_data(stock_list[:100])  # Download the first 100

# Retrieve financial data
for stock in stock_list[:10]:
    data = xtdata.get_financial_data([stock])
    if stock in data and 'PershareIndex' in data[stock]:
        psi = data[stock]['PershareIndex']
        if len(psi) > 0:
            latest = psi[-1]
            roe = latest.get('du_return_on_equity', 0)
            eps = latest.get('s_fa_eps_basic', 0)
            print(f"{stock}: ROE={roe}, EPS={eps}")

社区与支持

由 大佬量化 (Boss Quant) 维护 — 量化交易教学与策略研发团队。

微信客服: bossquant1 · Bilibili · 搜索 大佬量化 on 微信公众号 / Bilibili / 抖音