Dissertation and Other Status
December 2, 2005A-B-C-D Box Puzzle Update
November 11, 2005
Resume (C.V.) 3-page version
pdf, preserves formatting.
Resume, Full version               doc   |  pdf, use this one. Includes professional training, awards, etc.
January 3, 2005
Resume updated to show ongoing Adjunct Professor at University of St. Thomas, Cameron Graduate School of Business since January 2005.
November 14, 2004
Resume revised to show Rice University Faculty Lecturer in Statistics, originally posted July 13, 2004.
The Ph.D. oral defense examination was successfully completed on April 12, 2004. The dissertation manuscript, "Market Outperformance by Nonparmetric, Simugram-Based Portfolio Section", was submitted to the Office of Research and Graduate Studies on April 23, 2004. Excerpts are available here.
April 10, 2004
I have completed the unfinished portions of of the last two chapters. Some other changes have been made to the first part, but these can wait, but I think they changed the page numbers by 1 or 2. Just replace everything starting from table 3.4 (if you have it) until the end. Finished (again)
March 15-20, 2004
a. Have quantified and fixed the underperformance issue. Bottom Line: using the lamest allocation mix, the SP500 now outperforms the SP100 data, which of course outperforms the market.
b. Now its a matter of writing this thing up, while addressing the drawdowns in 2000-2002, and properly scaling the number of selected stocks in the grouped optimization solution.
c. We are in process of coordinating a defense date, probably around April 8-12 or so. Please check your calendars.
March 7-13, 2004
|Now that our new initialization scheme is running (for market cap instead of equal-weights), we have had the chance to check the SP100 vs. SP500 performance. Turns out the the SP500 underperforms the SP100 data starting with year 2. We are soon implementing the split-group technique to fix that.|
March 1-6, 2004
These are some comparative results for the S&P 100, with multiple samples of simugram returns providing measure of variability in the realizations. The terminal values are based on investing 1 unit at the beginning of 1970. sp100results.pdf.
|NEW! Return envelopes for alternate constraint sets are being generated now, and will be posted shortly. These will be for all years for the SP100, and selected years for the SP500. The examples shown here is for the SP100, in 1981, which was a down year for all the indexes (Dow -.10, SP500 -.10, simugram down average of .11 with +/-2% standard deviation.)|
We have here better returns from using a higher allocation per the above discoveries. As you can see from the table, annualized returns of 30% are possible. This assumes of course that there is not some HUGE non-causality in the software, which is currently being checked.
Tuesday, November 11, 2003
JRT had repeatedly asked "when do these futures contracts trade, and how far out can you trade them?" I put together 4 contracts for 2001 and 2003 July/December Wheat, so you can see the strips.
Click on the graphic to the LEFT to see the term PRICE structure of the 4 contracts; Click on the graphic to the RIGHT to see the VOLUME structure, or when one would actually trade these.
Note the WN01 is Wheat for July 2001 delivery; WZ is december
Monday, September 15, 2003
a. Workaround HTML compatibility issues by posting here.
b. JRT requested a big picture on the thesis concept to better place in context the postings hereon. I inventoried my in-process notes and "to-do' items; there are more things to develop and tie together than time available, at least for the truly significant result I was hoping to surprise you with.
c. The overview document is big_picture.pdf with 3 appendices:
Big Picture big_picture.pdf
Appendix 1. mm.geoindex.pdf
Appendix 2. mm.volume.tempchoice.pdf
Appendix 3. disstest.pm.pdf
The appendices are merely examples of writing style; appendix 1 and 3 are fairly complete, pending edits; appendix 2 is very rough but I had wanted to get something out before going to Houston for couple of days on 9/5.
d. Just printed my 700th piece of paper since 8/19/03.
e. IRS 1040ES submitted on time, sure could use a stipend.
While I am in New Mexico, telecommunications are difficult. I will try to check an alternate "civilian" email every day or two, but the lines drop out and are so slow. I cannot access Rice because of the hyper-security they implemented, which also eliminated the text-based dialin access, which these 9600-baud lines provide. The number at my parent's cabin is 505 425 8199, or you could contact me at 713 502 3894 and leave a voice message which I can check everyday.
September 5-9, 2003
In Houston for Hope's Happy 10th birthday ;')
September 4, 2003
a. Completed the geoindex section, posted at www.dobelman.net/dissertation.html, but there's some HTML compatibility problem with Mac browser, so we have to use this page. It is pending Jim's review prior to notifying Ed and Kathy, we've played phonetag and since then I've been busy on items b. and c.
b. Coding Fortran and splus various functions, routines and programs that should have been done 6 months ago. The goal this next reporting period is to complete the volume model. The volume model will be used in conjunction with the pricing and position model to simulate what an overall position looks like.
c. Since I'm going to Houston for Hope's 10th birthday this weekend, I have put some pictures up on the volume stuff, but it is not complete, nor even proper English; just to let you know some basic investigations and what my intent is.
d. Skeleton TOC and available drafts, the Apple html problem will probably be on these too... are HERE.
Tuesday, August 27, 2003
Do not use these, they are supereeded; mm.geoindex.doc
June 9-14, 2003
May 6, 2003
April 15 - May 2, 2003
Turned over trading accounts to fiduciary traders to permit concentration to finishing dissertation.
April 13-17, 2003
March 30 - April 2, 2003
August 24, 2002