"Kronos-Backtester" is a Python-based toolkit for financial data analysis and backtesting of trading strategies. It provides an effective platform for simulating trading strategies using historical data to evaluate their performance and potential profitability.
$ pip install kronos-backtester
fromkronos_backtesterimportBacktester# Example strategy to be backtesteddefmomentumStrategy(df, short_window=50, long_window=200, entry_threshold=0.02, exit_threshold=0.01):
# Strategy code here# Inputs: Pandas Dataframe of price data, any relevant parameters for the strategy# The Dataframe will have columns 'Close', 'Open', 'High', 'Low', and 'Volume'# Output: Pandas Series of signals which are integers -1, 0, 1# -1 : Sell, 0 : Hold, 1 : Buy# Index of Series should be dates# The wrapper should only take in a DataFrame and output a Series of signals# This is essentially one version of the strategy with a specific set of parameters.deftestWrapper(df): returnmomentumStrategy(df, long_window=100)
bt=Backtester(testWrapper)
# This backtests on a particular ticker with given start and end datebt.testTickerReport('AAPL', '2010-01-01', '2020-01-01')
# You can also backtest on a custom DataFrame of price databt.testCustomReport(customDF)Start 2010-01-01
End 2020-01-01
Duration 2516
Exposure Time 470.5
Net Worth [1000000, ... ,8166774.230371475]
Equity Final 8166774.230371475
Equity Peak 8166774.230371475
Return 7.166774230371475
Buy and Hold Return 10.038871419853216
Max Drawdown -0.1029208755830342
Avg Drawdown -0.09627259509420738
Max Drawdown Duration 19
Avg Drawdown Duration 8.857142857142858
# Trades 4
Win Rate 1.0
Best Trade 0.984095270845883
Worst Trade 0.48189821881601236
Max Trade Duration 669
Avg Trade Duration 470.5
Sharpe Ratio 33.99413578326285
Sortino Ratio nan
Calmar Ratio 69.11692296006356
Common Questions and Issues
● Q: What if I encounter an error regarding missing data?
● A: Ensure that all required data fields are present in your dataset. Missing data can often lead to errors during the backtesting process.
● Q: How do I handle a strategy that requires multiple stock tickers?
● A: Modify your strategy function to accept and process multiple tickers. Ensure that your backtester is provided with the correct data format.
● Q: The backtester is running very slow. How can I improve its performance?
● A: Performance can be improved by optimizing your strategy code. Consider reducing the complexity of calculations or using efficient data structures.