As the markets sell off today in the wake of poor China data and the expectation of a December rate hike (last week's good news becomes this week's bad news) I though I'd share a comparison of the M6 momentum (MM) versus mean regression (MR) models when looking at the broader market indices.
The portfolios are identical...SPY, QQQ, IWM, XLF, XLU and SH...the usual suspects in our market study models. Our goal today is simple to determine whether a momentum or mean reversion tactic (buy the 3 day high or buy the 3 day low) provides any advantage in terms of risk/reward.
We apply the usual .7% daily limit stop to all positions and we rotate into equal dollar amounts of the top 2 ranked positions at the end of each day.
Both models have superior linearity to the benchmark SPY as well as muck lower drawdowns.
Want to cover your bases? Here's an idea....trade both models simultaneously.
Tomorrow we'll look at the larger metrics panels of each model to help assess "in paradigm" versus "out of paradigm" conditions.
Things are continuing to evolve here and Mosaic and I will be offering the actual M6 software to regular readers in the near future so they can explore investment themes using the MM/MR skew to test their own ideas with alternate inputs.