Loading
A time series property where the joint distribution does not change when shifted in time.
A time series property where the joint distribution does not change when shifted in time.
A strictly stationary series has the same joint distribution at all time points; weak (covariance) stationarity requires constant mean, constant variance, and autocovariance depending only on the lag. Most price series are non-stationary (integrated) and require differencing or cointegration modeling for valid statistical inference.
Augmented Dickey-Fuller (ADF) and KPSS tests are the standard stationarity tests.
Non-stationary series produce spurious regression results with high R² and invalid t-stats.
Mean reversion strategies require stationarity of the spread, not the underlying series.
Quantitative research is the disciplined process of turning a financial hypothesis into a validated, deployable trading model. It spans data engineering, signal design, backtesting with proper out-of-sample validation, and ongoing production monitoring. The discipline's central problem is overfitting: most strategies that pass backtests fail in production.
Financial machine learning is the application of supervised, unsupervised, and reinforcement learning methods to financial prediction, classification, and decision problems. The defining methodological constraint is that financial data are serially correlated, not independently and identically distributed, which means that the standard machine learning toolkit must be substantially adapted. The dominant practitioner reference is López de Prado (2018).