coin-test 0.1.0

Robust and rigorous backtesting framework for cryptocurrencies.

coin-test

Coin-test is a backtesting library designed for cryptocurrency trading. It supports trading strategies across multiple currencies and advanced configurations of tests, including cron-based scheduled execution of strategies, synthetic data generation, slippage modeling, and trading fees.

Quick Start

Coin-test runs on Python 3.10 or higher. Install the package via pip:

pip3 install coin-test

To run a backtest, import the coin-test library. Then define your data source, strategy, and test settings to run the analysis.

import datetime as dt
import os
import pandas as pd

from coin_test.backtest import Portfolio, Strategy, MarketTradeRequest
from coin_test.data import CustomDataset
from coin_test.util import AssetPair, Ticker, Money, Side

Then, import data from a CSV or another source to load into the backtest.

dataset_file = "data/ETHUSDT-1h-monthly/BTCUSDT-1h-2017-08.csv"
header = ["Open Time", "Open", "High", "Low", "Close", "Volume", "Close Time",
    "Quote asset volume", "Number of trades", "Taker buy base asset volume",
    "Taker buy quote asset volume", "Ignore"
]
df = pd.read_csv(dataset_file, names=header)
df = df.drop(columns=["Close Time", "Quote asset volume", "Number of trades",
                      "Taker buy base asset volume",
                      "Taker buy quote asset volume", "Ignore"])
df["Open Time"] //= 1000  # To seconds
df = df.sort_values(by=["Open Time"])

# define dataset metadata
usdt = Ticker("USDT")
btc = Ticker("BTC")
asset_pair = AssetPair(btc, usdt)
freq = "H"

dataset = CustomDataset(df, freq, asset_pair)

Strategies are stored in classes as shown below. Each strategy should have a schedule, which is a cron string representing when this strategy is run, a lookback, which is how much data is accessed in the strategy, and a __call__ method which returns a list of TradeRequest objects, which represent trades the strategy wants to make.

class MACD(Strategy):
    def __init__(self, asset_pair) -> None:
        """Initialize a MACD object."""
        super().__init__(
            name="MACD",
            asset_pairs=[asset_pair],
            schedule="0 9 * * *",
            lookback=dt.timedelta(days=26),
        )
        self.perc = 0.2

    def __call__(self, time, portfolio, lookback_data):
        """Execute test strategy."""
        asset_pair = self.asset_pairs[0]
        exp1 = lookback_data[asset_pair]["Close"].ewm(span=12 * 24, adjust=False).mean()
        exp2 = lookback_data[asset_pair]["Close"].ewm(span=26 * 24, adjust=False).mean()
        macd = exp1 - exp2
        exp3 = macd.ewm(span=9 * 24, adjust=False).mean()

        if macd.iloc[-1] > exp3.iloc[-1]:
            return [MarketTradeRequest(
                asset_pair,
                Side.BUY,
                notional=portfolio.available_assets(usdt).qty * self.perc,
            )]
        elif macd.iloc[-1] < exp3.iloc[-1]:
            return [MarketTradeRequest(
                asset_pair,
                Side.SELL,
                qty=portfolio.available_assets(btc).qty * self.perc,
            )]
        return []

This package supports multiple strategies, train-test splits for historical data, synthetic data, and further customization. To run the backtest, define the datasets

from coin_test.backtest import ConstantSlippage, ConstantTransactionFeeCalculator
sc = ConstantSlippage(50)
tc = ConstantTransactionFeeCalculator(50)

datasets = [dataset]
strategies = [MACD]

results = coin_test.run(datasets, strategies, sc, tc,
                        portfolio, pd.Timedelta(days=90),
                        n_parallel=8)