News Focus
News Focus
icon url

OldAIMGuy

07/01/05 8:13 AM

#799 RE: aptus #798

Hi Mark, This looks very interesting for those interested in that aspect of stock selection. Thanks.

I've completely lost track of Barry. Last I heard from him was probably two years ago. He'd planned to come to WI for a pow wow bit it was cancelled when his wife had some health troubles.

Have you had more recent contact?

Best regards, Tom
icon url

WTMHouston

07/03/05 3:05 PM

#802 RE: aptus #798

A very nice tool and many thanks for taking the time to make it available.

A couple of ideas, thoughts, suggestions:

1. An option to make the volatility parameter in the algorithm adjustable. Thus, instead of the fixed 17%, one could enter any different value -- 5, 10, 15, 20, or more.

2. A function for a group of variable end dates to be computed over a variable time period. Thus, instead of just a single end date with a single result, one could acquire a result with say 48 end dates, one for each month for the last 48, one with 104 end dates, one for each week for the last two years, or even the last 712, each day's result for the last two years -- all run as a batch. If this were graphable, either in the program or through importing the data output into a program that would graph it, it would provide a picture of the volatility trend over the period. From a graphing standpoint, graphing the total as well as the high-low line would be particularly useful. Even if it were not graphable, just viewing the raw data output would give one a pretty good feel for the volatility trend. Ideally, plotting that on top of the moving average and current price could provide some truly useful data. Then again, as I am want to do, I may just be overanalyzing it.

3. Of course, combining the moving volatility with batch mode would be truly powerful.

Two questions:

At what frequency does the algorithm measure the volatility: daily, weekly, monthly?

From what online source does it retrieve the data? Obviously, the accuracy of that data controls the accuracy of the output.

Two observations:

1. It took me forever (well, okay, just 20 minutes, but it seemed like forever), to realize that somtickers.txt was not sometickers.txt. When I finally saw that there was no "e" after "som," the Duh! moment had arrived. It was "Screen-O-Matic tickers" not "Some Tickers."

2. I have probably not run enough mutual funds yet, but every one I have run has come back at zero. Same for SPY and QQQ over various time periods.