Robert F. Engle

Robert F. Engle is an American economist who developed autoregressive conditional heteroskedasticity models for changing financial volatility. He received the 2003 Nobel Memorial Prize in Economic Sciences.

Time series: A sequence of observations recorded over time, often analyzed for patterns and dependence. Engle developed methods for modeling time-varying variance in economic time series.
Autoregressive conditional heteroskedasticity: A model in which a time series’ conditional variance depends on past squared errors. Engle introduced ARCH to let recent shocks raise or lower the variance forecast.
Clive Granger: A British econometrician known for work on cointegration and the analysis of economic time series. Granger shared the 2003 economics prize with Engle for complementary advances in time-series analysis.
Value at risk: A measure estimating the loss a portfolio may exceed over a specified period and probability level. Conditional volatility forecasts from Engle-style models can inform portfolio risk estimates.
Heteroskedasticity: A condition in which the variance of a variable or statistical error changes across observations. Changing error variance is the feature Engle’s ARCH models were designed to represent.
Generalized autoregressive conditional heteroskedasticity: A volatility model whose conditional variance depends on past squared errors and past conditional variances. GARCH extends Engle’s ARCH framework by allowing volatility itself to persist directly.
Nobel Memorial Prize in Economic Sciences: An annual prize recognizing outstanding contributions to economic sciences, established by Sweden’s central bank. The 2003 award recognized Engle’s methods for analyzing time-varying economic volatility.
Portfolio optimization: The selection of asset weights to balance expected return against risk under specified constraints. Time-varying covariance estimates help portfolio methods account for changing market risk.
Autocorrelation: Correlation between values of a time series and its values at different time lags. ARCH captures dependence in squared returns even when returns themselves show little autocorrelation.
Engle's ARCH test: A test for autoregressive conditional heteroskedasticity based on dependence among squared regression residuals. The test made ARCH effects detectable in residuals from ordinary regression models.

Linked from 1 page