This past weekend I spent time developing an investing model investors might find interesting. I started the analysis on June 30, 2006 as HedgeHunter did with his Feynman Study. However, I narrowed the number of ETFs down to: VTI, EFA, VWO, VNQ, AGG, TLT, LQD, TIP, and SHY. While I thought about adding more, such as RWX, when the historical data became available, I think I will continue the analysis through to the end with these few securities. Here are the rules for analysis.
- Analyze the portfolio every quarter.
- Run a Cluster Weighting Momentum (CWM) analysis with the weighting set to 5. Due to the highly correlated ETFs, the number available for investing rarely tops 4.
- Follow the percentage recommendations from the CWM even though the portfolio is frequently concentrated.
- Sell shares when the ETF drops below the performance of SHY. The software automatically recommends this will happen.
- Sell shares when the price of the ETF drops below the 195-Day Exponential Moving Average (EMA) price.
- If CWM calls for a buy and the Absolute Acceleration is <0% or negative, do not buy the shares. This is different from other studies conducted here at ITA.
- If CWM calls for a partial sale and the Absolute Acceleration is negative, follow the recommendation of the spreadsheet rather than selling all shares.
- Round the buys to the nearest 50 shares.
I am approximately halfway through the period of investigation and the results look promising. It is noteworthy that the CWM software had the portfolio completely in cash in the middle of the Great Recession. The Risk Reduction model worked in this example, something I wanted to see in action. As the bottom of the market approached, the securities moved to bond and treasury instruments, another signal the software works well in a time of crisis. Check in later this week for the final results.