Dolphins Blowing Bubble Rings

Relaxing huh?

Robot Evolution

I wish my ATS could learn and evolve based on experience like these bots. The concept of an artificially intelligent auto trading system that could learn from mistakes and increase yield over time is very appealing. Of course I wouldn't complain if it did other things. Maybe a beer every once and a while? A snow shoveling, dish washing, laundry folding, floor cleaning, fish feeding, bill paying, back massaging, auto trading robot would be a friend of mine.

Current Rankings

The rankings after the first week are out and damn, that is some fierce competition. I told you I expected to be +20% by the end of this week but the top six places were already better than that at the end of last week! Much of my strategy depends on option decay so I will be seeing the biggest gains during options expiration week (this week), but still. Congrats to those top contenders I recognize a few of you from last year. I have a link to the current rankings on the left hand side of the blog if you want to see the standings directly from IB. They will be updating the rankings every Monday afternoon.

On a sour note, my program had a major catastrophe while I was in class today. SOLF tanked again and the program once again got stopped out of its position (at a huge loss). But the order took so long to fill that the software sent out another order to cover which resulted in a net long position in arguably the worst options in the market. There is a glitch in the program where if market orders take more than 20 sec to fill another order gets sent. I need to resolve it but I just don't have the time right now. I have an algorithm for exiting any long option position at a 50% losss or a 150% gain, so perhaps that will get hit (most likely the 50% loss). All around SOLF has cost the program $56k + $71k in realized losses and another $30k in unrealized losses on this new long position it acquired due to the glitch. If it weren't for that stock my program would be up $70k or 7% right now...

Ron Paul talks about "Arab Nations" and all McCain hears "al-Qaeda"

McCain is an angry war hungry old man and it scares the heck out of me that he has so much support. Frankly, if people voted with their conscience and not with their emotions Ron Paul would win by a landslide.



Who associates the beach boys with starting a war? The world would be a scary place with someone like McCain running the US. On the issue of bombing Iran, check this video below out. Ok, Ritter is a little dramatic but he makes some damn good points about bombing Iran. Its time to fix our broken foreign policy.

Olympiad Week One

Here's a quick update on the competition. This was a very tough week for the program as essentially a perfect storm hit it in the first three days. Without going into details, let me just say that very volatile stocks moved much more than is typical then kept on moving. In particular SOLF dropped around 20% on Monday, my program wrote naked puts, then it kept dropping. The software initially wrote $30 strike puts but got stopped out of those at a $55,000 loss (5%) ,then the program initiated a position in the $25 strike puts later in the week. In other words, the puts that were very expensive Monday got more and more expensive all week and my program won't get out of its second SOLF position until it breaks $25 and some change.

The software lost big on that position but made it up on other trades to close the week down slightly. There was a big swing for the account value being down more than 10% at one point but its now only down 2%. Here are the week ending Jan. 11th stats and holdings:

Liquidation Value: $980,064.37 -$20,000 on the week
Options Value: -$196,501 (I can reasonably expect +20% next week)
Net Stock Value: $55,523
Unrealized Profit: $49,760

Positions:

Longs:

XTO
SEPR
PPDI
KGC
IBN
GOLD
GDX
DBA
BVN
BAX
AUY

Shorts:

TIF
SYY
NIHD
DIS
BBH
AXP

Naked Option Positions (options sold short):

UAUA Jan 25 puts
SOLF Jan 25 puts
LDK Jan 40 puts
CSIQ Jan 20 puts
CROX Jan 25 puts
CAL Jan 25 calls
ABK Jan 20 puts
ABK Jan 15 puts

I'll post a link to the standings as soon as its up but I don't expect to be in it this week. The program takes an intial loss on every option position so I would expect it to not be up so much after writing all of those options this week. Next week however, those options will hopefully all be out of the money and expire worthless, meaning the software will rake in 100% profits.

May the best robot win! (IB Olympiad 2008)

Tomorrow the 3rd annual Interactive Brokers (IB) Collegiate Trading Olympiad gets underway and will last eight weeks. The competition involves writing a auto trading software (ATS) that syncs with the IB trader workstation (TWS) to retrieve market data and execute a trading strategy. In the first annual Olympiad the top three programs returned 119%, 83% and 46.5% over two months. Last year the top three returned 294%, 161% and 110% while my software placed me in 10th with a two month return of 27%. See previous winners here. So the competition is fierce to say the least. If any real fund manager could consistently generate these kinds of returns they would be getting paid millions.

This year the returns are certain to be smaller because the starting cash has increased by a factor of ten, from $100k to $1M. The simulated market that programs will be trading on is sophisticated enough to include market size from the top of the book, so liquidity is even weaker than in the real market. Now I'm not sure what strategies others were using last year but a small cap or options strategy might run into trouble scaling up their system x10 because those markets are less liquid. At any rate, it will be very interesting to see how much the top yields change.

The competition is only open to students (undergraduate and graduate (me)) so the prize money is a big deal and certainly represents a major augmentation to income if you get in the top 13 places. See the prize money breakdown below. Some have argued that this competition is a sour deal for contestants because they agree to give up the license to their strategy by competing and these programs would potentially return much more than the prize money if adequately funded. This point is valid I think but the fact of the matter is that in order for students to build the kind of reputation that you would need to be given a $1M investment they are going to have to have some nice auditable performance numbers. This competition does that (I can always put on my resume now that my software made 27% in two months) and provides a significant award for students who do well (I got a $10k check). So I'm not complaining.

I actually started this blog a year ago as a venue for posting my program's trades and progress. Those of you who were reading last year may have noticed that I removed all the posts about my software's activity. I did that because I started getting alot of emails from people who were clearly dissecting my software's strategy in hopes of doing well in this year's competition. I really want to do well this year and beyond, so given my success last year I decided to temporarily remove those posts while people developed their software this year. Plainly put, I think I have a very profitable and reliable strategy and I didn't want to give myself trouble by competing with others who are doing the same thing. Those posts are merely hidden from view and they will be put back up in the future.

This year I won't be posting every trade like last year because its very time consuming and my software will be trading more frequently than last year. I will however post my liquidation value every few days or so with some of the pivotal trades. IB will be doing a weekly ranking so I will post a link to that when they begin doing that on January 14th.

This year should be very interesting with the increased account sizes and the bear market in stocks. Last year the competition was running during the "glitch day" where the markets saw their biggest drop since 9/11. I expect that we will have some equally dreadful market swoons over the next few weeks to separate the floor cleaning robots from the gigantic city destroying mech warriors. Good luck everyone, may the best robot win!!!

Occam's Razor