FRACTIONAL BROWNIAN MOTION WITH STOCHASTIC VARIANCE: MODELING ABSOLUTE RETURNS IN STOCK MARKETS
2008 ◽
Vol 19
(08)
◽
pp. 1221-1242
◽
Keyword(s):
We discuss a model for simulating a long-time memory in time series characterized in addition by a stochastic variance. The model is based on a combination of fractional Brownian motion (FBM) concepts, for dealing with the long-time memory, with an autoregressive scheme with conditional heteroskedasticity (ARCH), responsible for the stochastic variance of the series, and is denoted as FBMARCH. Unlike well-known fractionally integrated autoregressive models, FBMARCH admits finite second moments. The resulting probability distribution functions have power-law tails with exponents similar to ARCH models. This idea is applied to the description of long-time autocorrelations of absolute returns ubiquitously observed in stock markets.
2021 ◽
2012 ◽
Vol 2012
◽
pp. 1-20
◽
Keyword(s):
2013 ◽
Vol 13
(04)
◽
pp. 1350010
◽
2014 ◽
Vol 51
(1)
◽
pp. 1-18
◽
2019 ◽
Vol 522
◽
pp. 215-231
◽