dumbmoney

dumbmoney is a technical analysis and quantitative trading toolkit designed for retail investors. The current version provides a unified, transparent interface to fetch daily stock prices across A-shares, H-shares, and US markets by abstracting popular data packages like massive, tushare, akshare, and tigeropen (optional), hiding their implementation complexity.

To use the massive or tushare providers, you must set the required environment variables MASSIVE_KEY or TUSHARE_TOKEN.

tigeropen is an optional provider. To use tigeropen, you must set the required environment varialbes TIGER_ID, TIGER_ACCOUNT, TIGER_LICENSE and TIGER_PRIVATE_KEY. Please refer to Tiger's GitHub repo for more details.

📦 Installation

Install only the core:

pip install dumbmoney

Install extra tigeropen package:

pip install "dumbmoney[tiger]"

🚀 Quick Start

Stock Details

from dumbmoney import get_stock_details

os.environ["TIGER_ID"] = "xxxxxx"
os.environ["TIGER_ACCOUNT"] = "yyyyyy"
os.environ["TIGER_LICENSE"] = "xxxxxx"
os.environ["TIGER_PRIVATE_KEY"] = "yyyyyy"
os.environ["TUSHARE_TOKEN"] = "xxxxxx"
os.environ["MASSIVE_KEY"] = "yyyyyy"

details = get_stock_details('AAPL.US')
print(details)

OHLCV Data & Charts

from dumbmoney import get_ohlcv, plot

os.environ["TUSHARE_TOKEN"] = "xxxxxx"
os.environ["MASSIVE_KEY"] = "yyyyyy"

ohlcv = get_ohlcv("AAPL.US")
print(ohlcv.tail())

from dumbmoney.indicators import MovingAverage, MACD, RSI

ma5 = MovingAverage(name="MA5", window=5, ma_type="SMA")
ma5.compute(ohlcv)

ma20 = MovingAverage(name="MA20", window=20, ma_type="SMA")
ma20.compute(ohlcv)

ma60 = MovingAverage(name="MA60", window=60, ma_type="SMA")
ma60.compute(ohlcv)

vol_ma20 = MovingAverage(name="Vol_MA20", window=20, ma_type="SMA", input_col="volume")
vol_ma20.compute(ohlcv)

macd = MACD()
macd.compute(ohlcv)

rsi = RSI()
rsi.compute(ohlcv)

plot(
  ohlcv,
  indicators=[ma5, ma20, ma60, vol_ma20, macd, rsi],
  panels=[0, 0, 0, 1, 2, 3],
  title="AAPL Stock Price with Indicators (mplfinance)",
  backend="mpl", # available backends: "mpl", "plotly"
)

Strategy, Policy & Backtest

from dumbmoney import get_ohlcv
from dumbmoney.strategies import MACrossParams, MACrossStrategy
from dumbmoney.policies import LongFlatAllInConfig, LongFlatAllInPolicy
from dumbmoney.backtests.single_asset import SingleAssetBacktester

ohlcv = get_ohlcv("AAPL.US")
strategy = MACrossStrategy(MACrossParams(fast_window=20, slow_window=60))
policy = LongFlatAllInPolicy(LongFlatAllInConfig(max_long_pct=1.0, min_strength=0.5))
backtester = SingleAssetBacktester(initial_cash=100_000.0)
result = backtester.run(symbol="AAPL", ohlcv=ohlcv, strategy=strategy, policy=policy)
print(result.metrics)

✨ Features

• 🔌 One function to fetch ohlcv data: get_ohlcv(symbol, start, end)
• 🌏 Multiple markets supported
  • A-shares (.SH, .SZ)
  • H-shares (.HK)
  • US stocks (.US)
• ⚙️ Automatic provider routing
  • A-shares: TigerOpen → TuShare → AkShare
  • H-shares: TigerOpen → TuShare → AkShare
  • US stocks: TigerOpen → Massive → AkShare
• 📐 Unified normalized output
  • open, high, low, close, volume
• 🔁 Fallback logic
  • If one provider fails, the next takes over
• 🧩 Extensible architecture (plug in new providers)

Important Notice

• massive's free api key only supports retrieving data of US stocks from the most recent two years.
• tushare's free token only supports retrieving data of A-shares.
• akshare is free but depends on third-party data sources that may have variable reliability.

🏷️ Symbol Format

dumbmoney uses suffix-based symbol conventions:

Suffixes for H-shares and US stocks are required. A-share symbols may omit suffixes; however, if they are present, they must be valid and correct.

📘 API Reference

get_ohlcv(symbol, start, end, adjust="none")

Fetch normalized daily OHLCV prices.

• Parameters

• Returns

A pandas.DataFrame with:

Index is a DatetimeIndex named date.

plot(ohlcv, indicators=None, panels=None, title=None, backend="mpl", **kwargs)

Plot chart using the provided ohlcv data.

• Parameters

• Returns

The plotter object which can be used to get the figure object by plotter.fig.