Financial econometrics · Time series · R programming
Building an R financial-econometrics pipeline and a custom procedure to measure the return of volatility toward a stable regime after a market shock.
Context
During a four-month assignment completed in December 2020, I supported doctoral research by developing a substantial part of an R-based financial-econometrics pipeline. The client identity, location and institution remain private.
Problem
The work required modelling conditional volatility across financial series, identifying high-volatility episodes and determining when volatility returned sufficiently close to a stable reference level after a shock.
My role
I developed most of the analytical R code, covering series preparation, return calculation, descriptive statistics, ACF/PACF analysis, Ljung-Box and stationarity tests, GARCH specification and estimation, diagnostics, conditional-variance extraction and identification of high-volatility episodes.
Method
The procedure identified conditional-variance peaks, recovered their dates, defined a reference volatility level, scanned observations after each shock, detected the first sufficiently close return to that level and calculated recovery duration.
Analytical distinction
The procedure measures volatility returning toward a reference or stable regime. It does not claim that prices return to equilibrium, establish causality or demonstrate unsupported model performance.
Outcome
The assignment produced an analytical workflow and an additional procedure for quantifying the temporal persistence of volatility shocks. Detailed empirical results are not published.
Technologies
R · GARCH · Time Series · Financial Econometrics · Conditional Volatility
Confidentiality
The client identity, country, institution, doctoral topic, data, detailed results and original source code are not published.