Strategy
This site uses artificial intelligence to make weekly
investment decisions based on global asset class allocation. In other words, each week the system will
make an informed decision to invest in a particular broad category of the global
economy. The goal is to outperform the S&P 500 (IVV).
Overview
Each week, the system chooses an ETF among a set of global
asset classes. Asset class allocation provides
many advantages including portfolio diversity and improved performance. Research suggests over 90% of a portfolio's return comes from asset class selection, and is preferable over individual stock selection.
Neural network technology is used to develop models that
identify successful trading patterns from historical data. This includes data such as price history and
derived technical indicators, economic data, indexes, options and commitment of
traders. Model inputs are derived using statistics and genetic selection to avoid curve fitting. The idea is to identify repeatable
profitable trades to develop a play book for the future. The play book consists of advantageous
positions that have statistically proven to be lucrative in changing market
conditions.
Week ahead forecast predictions are posted prior to Monday market open. Intermediate term forecasts are used in combination
with weekly model predictions to select the favorable asset class (ETF). System performance is calculated based on
weekly return excluding slippage and commission.
How to Trade the System
The predictions can be used to trade
in many different ways. Obviously, you can
simply go long each week on the selected weekly ETF. Or, augment your portfolio with the selected
ETF to add diversity. Additionally, you
can use the overall market predicted direction to assist you in your own
trading style.
