what is white noise in stationary times series : part 2, advanced techniques

hey hey data folks !

in the last article about white noise in finance (what is white noise in stationary time series ? part 1: introduction), we explored the fundamentals of white noise, why it matters in time series analysis, and how to detect deviations using traditional methods like adf, ljung-box, and jarque-bera. but let’s be real, if you want to play at the top level, you need more firepower. time series are complex, and if there’s any exploitable structure hidden beneath the randomness, we need better techniques to uncover it.

in this part 2, we’ll dive into advanced techniques for detecting white noise and identifying subtle deviations

beyond classical tests

1. singular spectrum analysis (SSA) — signal extraction 2.0

traditional tests are good at detecting trends and autocorrelation, but what if the patterns are buried deep within the noise? ssa is a powerful tool that decomposes a time series into its fundamental components: trend, periodicity, and noise.

how it works:

• builds a trajectory matrix from lagged copies of the time series
• applies singular value decomposition (svd)…