|
PureBytes Links
Trading Reference Links
|
Hi Barry ... no thanks to me, we owe all of our thanks to Gitanshu for
making this available.
======
To those who wrote about seasonality. In addition to the X-11 processes ...
here's what the "Help Index" in KyPlot says (a free "monster" stats program)
http://www.qualest.co.jp/Download/KyPlot/kyplot_e.htm
"... Time Series Decomposition: applies a seasonal adjustment model given by
Kitagawa and Gersch (ref. 1) to data. A time series is decomposed into
trend, seasonal, autoregressive and noise components. Trading day factors
can be also incorporated. State space modeling and Kalman filter are used
and parameters are estimated by maximum likelihood method. ..."
Sample File #12: "12Timser-S1" in KyPlot gives the following method:
"... Example 2: Time series decomposition and interpolation
The "Time Series Decomposition" in the [Math] - [Time Series Analysis] menu
decomposes time series data into components based on the model given by
Kitagawa and Gersch (ref. 1)
Procedures
1) Log transformation of the data.
Since the variability of the data increases for larger values,
logarithmic transformation is applied to the data.
a) Select the data BU4:BU149; Pressing Ctrl key, select the output cell
BV4.
b) Open [Data] - [Transform] menu.
c) Select "Base-10 logarithm" (and determine the output starting cell).
d) Click "OK".
2) Calculate AR coefficients by maximum likelihood method and make
prediction.
a) Select the data BW4:BW149.
b) Open the [Math] - [Time Series Analysis] menu.
c) Select "Time Series Decomposition".
d) Check "Trend" and set "Order"=1; check "Seasonal";
check "AR Order" and set "Order"=2; check "Trading Day Effect".
e) Input initial values.
e.g. Set "Variance Ratios" for "Trend", "Seasonal" and "AR" =10.
f) Click "Start Calculation". ..."
=========
This example is completely worked out for you in detail. It is a very large
spreadsheet and chartsheet. Note the pre-processing of the data. See
Ruggiero for more info on pre-processing data.
Best regards
Walter
----- Original Message -----
From: "Walter Lake" <wlake@xxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: Thursday, December 30, 1999 6:34 PM
Subject: Re: Heat Map explained
| Hi Jim and others who have written
|
| This is a nice "ATR stops for your portfolio.xls" with lots of VBA code
both
| in the .xls and as .txt ... .
|
|
http://content.communities.msn.com/isapi/fetch.dll?action=get_message&ID_Com
|
munity=MoneyCentralSuperModels&ID_Message=1111&ID_Last=0&Dir=0&ID_Topic=0&La
| stModified=942056506000
|
| From: Jon_Markman
| Sent: 10/28/99 7:36 pm
| Attachments: SuperModels_PRQ.txt (1.7KB), ATR_v3b.txt (9KB)
|
| From: Jon_Markman Reply 8 of 60 Add a Reply...
| Sent: 11/8/99 2:21 am
| Attachments: ATR Stops_Portfolio_v3.3.xls (968KB)
|
| From: Jon_Markman Reply 22 of 60 Add a Reply...
| Sent: 11/19/99 5:44 pm
| Attachments: PRQ MonthsAndWeeks v1_4.xls (263.5KB)
| here ist the new PRQ Monthly and Weekly analysis macro ....
|
| ====================
|
| To others ... seasonality in Gitanshu's workbooks is not the same as
| seasonally adjusted. These are two different concepts. Lets just
re-program
| these for our tradeables and worry about the other stuff later. Usually
the
| X-11 processes are used for seasonally adjusted time series data. The
Excel
| 3-D ribbon chart usually has a vertical ordinal scale. A tradable with no
| seasonality will have a divisor of 1.00. Seasonally-weak values have
| divisors of less than 1.00, and seasonally-strong numbers have values
| greater than 1.00.
|
| Best regards
|
| Walter
|
|
|
|
| ----- Original Message -----
| From: "Jim Greening" <jimginva@xxxxxxxx>
| To: <metastock@xxxxxxxxxxxxx>
| Sent: Wednesday, December 29, 1999 8:45 PM
| Subject: Re: Heat Map explained
|
|
| | Gitanshu,
| | Great work. I've been thinking along those lines myself and I'm
glad
| | someone has beat me to it and has already done all the leg work. I'm
| going
| | to think about how to incorporate your findings in the first part of my
| | trading strategy which is to pick the right sectors. In the past, I've
| just
| | waited for a sector to begin to trend strong and then jump on. If the
| | seasonality holds true, it will have the great advantage of being able
to
| | jump on early at lower prices before the trend starts. Keep up the good
| | work.
| | Once again, thanks for sharing.
| |
| | JimG
| |
|
|