What Is ‘Overfitting’ in a Trading Model?
Overfitting occurs when an algorithmic trading model is too closely tailored to the specific noise and random fluctuations of the historical data used for backtesting. This results in a model that performs exceptionally well on past data but fails to generalize and performs poorly when introduced to new, live market data.
Glossar
Noise
Phenomenon ⎊ Noise in financial time series refers to random fluctuations in price or trading volume that lack predictable patterns and are not attributable to fundamental information or systematic trading activity.
Historical Data
Provenance ⎊ Historical data, within cryptocurrency, options trading, and financial derivatives, establishes a verifiable record of asset origins and transactional history, crucial for assessing counterparty risk and regulatory compliance.
Random Fluctuations
Noise ⎊ Random fluctuations represent short-term price movements that do not reflect fundamental value or underlying market trends.
Algorithmic Trading
Execution ⎊ Algorithmic trading within cryptocurrency, options, and derivatives markets leverages pre-programmed instructions to initiate trades, automating order placement based on defined parameters and market conditions.