RiskAR1
|
Description |
RiskAR1
An AR1 process is a common model for a time series because it is simple and often provides a good fit. It is characterized by an autocorrelation function (ACF) that decreases geometrically and a partial autocorrelation function (PACF) that cuts off to 0 after lag 1.
|
|
Examples |
RiskAR1(100, 40, 0.8, 490) generates an AR1 process with mean 100, variance 402 / (1 - 0.82) = 66.72, autoregressive coefficient 0.8, and value 490 at time 0. RiskAR1(C10, C11, C12, C13) generates an AR1 process with parameters taken from cells C10 to C13.
|
|
Guidelines |
|
|
Technical Details |
Define
Then
The mean and variance are
and
|
generates a first-order autoregressive (AR1) process with mean 
, volatility parameter 
, autoregressive coefficient 
, and value 
at time 0.
is a necessary condition for stationarity.
a sample from a Normal(0,1) distribution
