Quote from total_keops:
Gordon Gekko (I think your gehko actually),
any monthly update on your OPG paper trading?
Here is a rundown of my current status.
August:
I spent the whole month trying to tweak my system to have a better win loss ratio. I also modified my spreadsheet significantly for news, earnings and (up/down) grades.
Therefore the results of the system were skewed in such a manor that it would not behoove myself to try to explain an entire month, when the settings changed every few days. To put it another way the results of the system is only as good as the data that is being collected (Shit In = Shit Out). Which leads me to September.
September:
With the massive overhaul to the spreadsheet in August and the lackluster results that came from constantly tweaking entries, exits and p/l ratios I devised a plan for September, October, and November that would concentrate on each of those points.
*Point 1 - Generate data on the morning (0930 to 1030) swing of stocks, by sector. This gives me a feel for the volatility of each sector during the morning hour and lets me get a more accurate picture of a stop-loss for that group. It is interesting to note that the stocks in my basket (which is loosely based on the S&P 500) will meet profit targets 73.3% of the time if they are given all day to do it. The downside to that is the losers REALLY lose and will wipe out all of the gains made from the winners. So it just goes to show you could have a system that works 73% of the time, but without proper money management you will run your account to 0 in no time. Therefore I chose to stick to the first hour because it was my personal preference to not have to wait all day on a strategy that is supposed to be quick and it did not make sense from a money management standpoint.
*Point 2 - Generate data for each stock, based upon previous performance, which focuses on the likelihood that it will move from opening print to the desired profit point without hitting the sector weighted stop. This step is moving my openings to more of a pair based setup that treats the sector as an 'opening basket' instead of relying on the performance of an individual stock. This allows me to give a sector based profit target, which can then be paired against other sectors, in conjunction with the information collected from Point 1. The interesting tidbit that I found with this point was, though the banking sector (is/has been) very volatile, it produced some of the most consistent winners over the pas few months:
*AIG - 86%
*COF - 81%
*MER - 88%
*MS - 72%
*Point 3 - Generate data based on pre-opening sector volatility that correlates known market events with likelihood to (reach profit target AND not hit the stop AND within the first hour). For instance, will the oil sector be more or less likely to fill in the first hour on Wednesdays when the inventory report is due out? How does the consumer discretionary sector behave when the economic data for personal spending is released? It has been generally accepted to cancel out of any openings which are news driven, but I want to find out how that relates to pairing the sectors as previously mentioned.
That said, the Gustav downgrade at the beginning of September started a 'month that was' and then coupling that with the short ban list and all the 2+ sigma swings it made it very difficult (if not impossible) to collect what I would consider normal data. I have suspended most of my opening strategy data collection until the short ban is lifted, and have been concentrating on the data I already have to make a more defined approach.
Thanks,
~g
