PureBytes Links
Trading Reference Links
|
Hi Rudolf
Here is a more traditional explanation for what I was trying to get at. This
is the whole page without editing.
The following quote is from "Time Series Models for Business and Economic
Forecasting" Cambridge University Press ISBN # 0-521-58404-3 by Philip
Franses
Chapter 7 page 155
"Conditional heteroskedasticity
For many economic time series one may expect that additive and innovation
outliers do not occur too frequently. The main exception, however, concerns
financial time series. Since such data reflect the result of trading amongst
buyers and sellers at, for example, stock markets, various sources of news
and other exogenous economic events may have an impact on the time series
pattern of asset prices. Given that news can lead to various
interpretations, and also given that specific economic events like an oil
crisis can last for some time, we often observe that large positive and
large negative observations in financial time series tend to appear in
clusters.
One possible approach to take account of sets of outliers may now be to use
the methods described in the previous chapter to remove the impact of these
aberrant data in order to have a clearer view on the pattern of the
underlying time series.
However, in empirical finance we tend to be less interested in the level of
the asset price or stock market index since it is widely assumed that such
time series can be best described as random walks. Therefore, another
approach is to exploit the fact that outliers emerge in clusters by trying
to construct a time series model for the outliers themselves.
Since a sequence of outliers can be considered to reflect a more volatile
period, such a time series model can eventually be used to forecast
volatility. In other words, due to these sets of aberrant data, the variance
of financial time series varies over time and hence the forecasting
intervals for the levels should also vary. The intuition is that in volatile
periods there is more uncertainty about the next observation than there is
in less volatile periods, and hence in volatile periods the forecasting
intervals will be larger.
In this chapter, I discuss a few time series models for the type of time
varying variance that is regularly observed for financial time series. The
common property of these models is that they allow the squares of the level
series (after removal of autocorrelation) to have autocorrelation patterns
that are similar to those of, for example, standard ARMA time series.
...etc."
Best regards
Walter
----- Original Message -----
From: "rudolf stricker" <lists@xxxxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: Friday, January 28, 2000 6:55 AM
Subject: Re: Seasonals and technical analysis
|
| Hi Walter,
|
| On Wed, 26 Jan 2000 22:08:24 -0500, you wrote:
|
| >Yes ... I model winners and losers separately also long and shorts.
|
| Then your comment concerning leptokurtosis reduction during system
| optimization
|
| | Why? That's where all of action happens. No leptokurtosis ... no cash,
no
| | flame-outs.
|
| does not make much sense to me.
|
| mfg rudolf stricker
| | Disclaimer: The views of this user are strictly his own.
|
|