Portfolio Construction: Establishing Goals

by: Lowell Herr

Constructing a portfolio is always a challenge, but less so if one establishes goals before putting together a group of stocks, index funds, ETFs or some combination. A software tool I've found useful in building portfolios is Quantext Portfolio Planner (QPP) developed by Geoff Considine. QPP provides guidance in future projections, risk management, and diversification. Below is a list of goals I extract from QPP when developing a Strategic Asset Allocation plan for a portfolio.

  • The projected return should exceed the projected return of the S&P 500 (NYSEARCA:SPY) by one percentage point. In this example I project the S&P 500 will grow at a rate of 7.0% over the next six to twelve months. I also used the last 48 months (8/1/2007 - 8/1/2011) as a time frame.
  • The projected standard deviation should be sufficiently low so the Return/Uncertainty ratio is greater than 0.60. This ratio varies in different market conditions and 0.60 for this period.
  • The Diversification Metric (DM) should exceed 40%. DM is one of the metrics produced by the QPP software.
  • Another metric from QPP is the Portfolio Autocorrelation (PA) percentage. While I don't pay a lot of attention to PA, I do want it to be under 20% if possible.

The following combination of twelve investments meets each of the above standards. This portfolio is projected to return 8.3% over the next few months or 1.3% greater than that projected for the S&P 500. In the blue background, one can see the standard deviation projection of 12.8%. This percentage gives rise to a Return/Uncertainty ratio of 0.64 or just above our goal of 0.60. Scrolling further down the screen shot one sees a Diversification Metric of 43% and a Portfolio Autocorrelation of 12.3%. This is a rather low and desirable PA value.

(Click charts to enlarge)

Click to enlarge

The following screen shot shows the correlation matrix for this portfolio of ETFs and two stocks. Pay particular attention to the Portfolio column as it tells the manager how the different investments are correlated to each other given these particular weights. Readers may be surprised to see the high correlation between the U.S. market and international or emerging markets.

Click to enlarge

While this combination of ETFs and stocks meet the required goals, it is a useful exercise to step back and criticize either the choice of investments or weight assigned to each. For example, does it make sense to allocate only 9% to the U.S. equities market? Is the allocation to emerging markets (VWO) too low while the commodity asset class receives 22%? Should the percentages allocated to TLT and TIP be switched? If this move is made, the results are quite different. This also raises the question of how much confidence should one place in any stock market software tool given the erratic nature of the market?

The QPP software is a little like a mean variance optimizer in that one needs to place constraints on the various allocation percentages lest one be drawn into portfolios that don't meet the specific requirements of the individual. While no software tool is perfect, the QPP does help investors interested in the principles of asset allocation to focus on diversification, portfolio risk or uncertainty, and potential returns. Judgment is always required as data extrapolation rarely hits the mark.

Disclosure: I am long VTI, VEU, VWO, VNQ, GLD, SLV, TIP, NLY.