PureBytes Links
Trading Reference Links
|
FYI,
Rob Engle and Clive Granger are the rock stars in the field of
econometrics. Rob Engle developed the ARCH process in 1982 which is an
econometric method to model time varying volatility. Why is it of use ?
Well, anything that you use that uses standard deviation assumes the
volatility is constant in the entire time series lookback that you are
using. Obviously, most traders know that this is not the case in real life.
Vol changes from high vol periods to low vol periods, i.e. market data
exhibit volatility clustering. Look at equities right now. In 1986
Bollerslev enhanced the ARCH to a GARCH process which is an econometric
curve fit to model time varying volatility. A Garch(1,1) process is
effectively an EWMA of your volatility. The problem with GARCH is that it
models your historical vol as a curve fitting process which means it can be
a poor method to model your vol going forward. Most studies on GARCH and
the practitioners that I have communicated with that have modeled or used
Garch tend to show that it is flaky when estimating future vol.
Rob Engle and Clive Granger also developed the concept called Cointegration
using the 2 step Engle-Granger methodology in back in 1987. Cointegration
is another story though........
If you want to learn about ARCH and GARCH you need to learn econometrics
and the books available on econometrics are not for the faint-hearted.
There is no "Econometrics for Dummies" book out there unfortunately. For a
cheap introduction to econometrics try Schaum's "Statistics and
Econometrics" 2nd Edition by Dom Salvatore and Derrick Reagle. The Schaum
book does not have ARCH or GARCH but it is an introduction to
econometrics. A more expensive book which covers ARCH and GARCH in detail
is "Market Models" by Carol Alexander, and a top econometrics book worth
its weight in gold is "Applied Econometric Time Series" by Walt Enders. I
warn you though, all 3 books are not easy reads. This subject matter is dry
and heavy reading, but it's all about time series analysis, which is what
trading is all about. Best of luck.
Rob Bianchi
|