I'll post a follow up to this with more detail shortly because I have to run, but last night's analysis uncovered a simple way to bring in 37% CAGR since 2004.

I use ETF's similar to those from Grossman's GMR: MDY, FEZ, EEM, ILF, EPP, TLT, and SHY.

Very simply, you calculate a regression line to 3 day's data and find the slope. You can do this with the LINEST or SLOPE functions in Excel. Then you buy the ETF with the lowest slope. Rinse and repeat every day. Usually, you're in the same ETF for a few days at least, so you are not trading every day, but you are trading generally once or twice a week.

Returns vs MDY:

SYSTEM | MDY | |

2004 | 22.46% | 14.31% |

2005 | 37.54% | 15.41% |

2006 | 30.91% | 8.60% |

2007 | 54.64% | 2.13% |

2008 | 56.12% | -31.60% |

2009 | 65.89% | 35.45% |

2010 | 44.02% | 23.83% |

2011 | 22.93% | -1.52% |

2012 | 9.38% | 20.49% |

2013 | 19.15% | 28.35% |

That is CAGR to 37.1%, total returns of 2404%. If you have a commission per trade of .04% ($20 on a $50,000 principle), then the CAGR goes to 28.5%. I'll take 28.5% CAGR any day.

Given the frequency of trades, this is probably best done with a leveraged strategy (i.e. options).

Max drawdown was 29% but still well below the S&P's max drawdown during the same time period. Standard dev was 28.2% annually with a Sharpe ratio of 1.32.

More details provided later...

**Disclosure: **I am long SSO.