Feb

24

I've been casually participating in Yahoo!'s predictive markets — search engine buzz popularity — and gaming mechanical investment and trading strategies there for perhaps 18 months, but with renewed interest the last couple of months.

There are clear methods to 'cheat' the tool with classic pump and dump' and confederates, but I put that aside, as it has been useful to give me a venue to sharpen my 'perl' "counting" skills and to get more familiar with XML (Extensible Markup Language).

The market simulator uses an XML RPC (Remote Procedure Call) mechanism (full API details at the web site cited) whereby one uses computer retrievals through a virtual web view, to buy, sell, get holdings details, and to read market statistics. The results of trades, or market data are returned in the highly structured form of XML, which is well suited to further computer driven parsing.

The last few weeks, I've been pursuing a 'chase the best dividend yield' strategy, and moving up in the rankings; writing a 'momentum move follower' is on my docket. As part of this, I have written a full series of scripts to check for newly 'IPO'd symbols, to review the efficiency of my portfolio, and the potential top performing peers alternative available for switching into. The computer suggests the trades, but I can over-ride it.

And so, even though it was contrary to a profit maximizing approach, I added equal buys of the remaining primary contenders a few weeks ago — $100 each; the 'buzz' in search engine returns has moved them all a bit (1st column is the current dollar 'value' of each holding)

[herrold@couch xml]$ ./xml-parse-holdings.pl | \
     ./filter-holdings.php -gt 2 | grep PRIMARY
58 DEMPRIMARY CLINTON 173 0.34 295
128 DEMPRIMARY OBAMA 149 0.86 117
173 REPPRIMARY MCCAIN 231 0.75 134
[herrold@couch xml]$

I should have bought McCain hard when Romney left the race. There is no way to short. Ah well.

But each week, my tools have rejected the sentimentalism I used (in undertaking small 'positions' for a bit of blood in the primary gamesmanship) as none was in the cohort of the 'best yielding'. One of my script watches for 'laggard' holdings to 'cull' from the portfolio; each week, I have ignored its counsel on the PRIMARY markets, where it wanted me to sell all.

But this week something different happened:

[herrold@couch xml]$ ./xml-parse-allinfo.pl >  find_cap.data
[herrold@couch xml]$./expected_div.php > yield
[herrold@couch xml]$ cat yield | sort -n | tail -20 > top_twenty
[herrold@couch xml]$ ./xml-parse-holdings.pl > all_holdings
[herrold@couch xml]$ ./spot_laggards.php
./sell.sh 172 DEMPRIMARY CLINTON
./sell.sh 230 REPPRIMARY MCCAIN
[herrold@couch xml]$

** I was not scolded about holding DEMPRIMARY OBAMA **

Very curious, as I have been since I undertook it. Turns out OBAMA is in the top twenty 'yielding' investment opportunities.

That made me look at where the 'buzz' was, as reflected in 'market cap' (col 1).

[herrold@couch xml]$ ./xml-parse-allinfo.pl | awk {'print $3" "$1" "$2'} | \
     sort -n | grep PRIMARY | tail -5
1749814 REPPRIMARY HUCKABEE
2493690 REPPRIMARY PAUL
2891712 DEMPRIMARY CLINTON
14557726 REPPRIMARY MCCAIN
18761165 DEMPRIMARY OBAMA
[herrold@couch xml]$

Assuming a left to right continuum (HUCKABEE is in the grid, PAUL seems orthogonal in many ways; I list then as the counts appeared), there is an amazing buzz on the left (not overly surprising; for the last week, the R side is McCain's to lose), and an evenly divided 'electorate'

Amt        Cume
1749814        1749814        HUCKABEE
2493690        4243504        PAUL
14557726    18801230    MCCAIN
<<- median     20227053     ->>
2891712        21692942    CLINTON
18761165    40454107    OBAMA

Offered for what it is worth — I have already voted absentee, so it won't affect my 'real world' actions.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search