ES / SPY / SPX behavior on FOMC announcement days: tends towards bullishness.

• yesterday, i referenced a setup (S5) by Trading the Odds, which suggested that the day before the FOMC announces their latest interest rate targets, tended to be bullish during the 1990-present time frame...
• in an attempt to potentially build a systemic trading model based on this bullish behavior, i modeled the behavior of the SPY on the day before the FOMC announcement for the 1996-present time frame...
• i too came to the conclusion that since 1996, the SPY has shown an overall bullish bias...however that bullish bias peaked in 11/2001 and since then, the SPY has traded bearishly on the day before the FOMC announcement...i did not build a systemic trading model based on this behavior...
• today, Trading the Odds posted another FOMC related study...this setup (S2), from 1990-present, goes long the ES at the close of the day before the FOMC announcement, and then exits the trade at the close of the day of the FOMC announcement...this simple strategy generated 156 trades, with a Profit Factor=2.13...
• this study peaked my curiosity, so i attempted to build a trading model based on the bullish behavior of the SPY on the day of the FOMC announcement...my study runs from 1996-present, and at the close of the day before the FOMC announcement, buys $100,000 worth of the SPY and upon the close of the next day, which is the FOMC announcement day, sells the entire position...there are no additional filters and there are no stops...
• my study, generated 107 trades, a Profit Factor=2.17, and an average trade return of +0.36%:

• a Profit Factor above 2.00 is pretty good...the next thing i want to see is what the trend of model has been over the life of the model, so i looked at the Equity Curve:

• the equity curve has shown extremely good performance lately...
• Quantifiable Edges also performed a similar study on this behavior, using SPX data from 1982-present...their equity curve chart shows similar performance...
• the next step for me was to see if i could build a systemic trading model based on this behavior...i divided up the SPY datastream into two slices: an in-sample slice from 1/1/1996-8/30/2006, and an out-of-sample slice from 9/1/2006-present...on the in-sample slice, i tested a simple moving average filter, a RSI filter and a stop...when i found a filter level that met my criteria, i would then run the model on the out of sample data...prior to adding any filters or a stop, i knew the Profit Factor was already 2.17...i like my models to have a Profit Factor of at 3.00...however in this case, neither a moving average filter nor a RSI filter combined with a money management stop yielded a model with a PF over 3.00...i was hard pressed to find something a model that yielded more than the baseline Profit Factor of 2.17...
• i was not able to build a systemic trading model based on the bullish bias that the SPY/ES/SPX tends to have on the day of the FOMC announcement...however, this bullish bias is good data in and of itself, and it has already factored into my thinking for my discretionary swing trading of the NQ...
• thanks again to Trading the Odds and to Quantifiable Edges for posting very relevant and useful studies of stock market index behavior....
• prior to the next FOMC announcement (11/4/2009 i believe), i will model the NDX and see how it performed on the day before the FOMC announcement and the day of the FOMC announcement...if there is a bullish bias, then i will attempt to build a systemic trading model...