Examining the Maxwell asset allocation below, we note that developed international and emerging markets are still priced below their respective 195-Day Exponential Moving Averages (EMAs). We are also below target in commodities as we are waiting on a price rebound. Exposure to growth stocks is accomplished by holding shares in the blend column.
The Importance of Managing Portfolio Uncertainty
If 2008 taught us one lesson it is that portfolio risk is just as important as portfolio return. In the following analysis I want to examine some of the risk details of the Passive Portfolio (PP) or the Schrodinger Portfolio. At the end of 2007, this portfolio had a value of $234,160 and by the end of 2008, one year later, the value dropped to $162,513 or a negative return of 30.6% (-30.6%). Were we prepared for such a drop and if so, how likely is this to happen again? Those are a few of the questions we want to analyze as we take a careful look at this portfolio.
While the Schrodinger continues to outperform the VTSMX benchmark by approximately 3% points annually over an eight-year span, for many years the portfolio bested the benchmark by 8% points on an annual basis. Here is the asset allocation plan for the passive portfolio when this analysis was last conducted.*
- Large-Cap Value = 14%
- Mid-Cap Value = 15%
- Small-Cap Value = 13%
- Large-Cap Growth = 13%
- Mid-Cap Growth = 13%
- Small-Cap Growth = 5%
- International = 15%
- Emerging Markets = 7%
- REITs = 5%
- Cash = 0%
The asset allocation percentages were not altered during the year and the portfolio did not require any rebalancing. Cash was added and a few ETFs were purchased throughout the year. Since the Schrodinger was launched in late 2000, there has been a minimum of activity, and this was true during the bear market of 2008.
Back on 12/31/2007, a Quantext Portfolio Planning (QPP) analysis showed this portfolio to have a projected return of 8.3% with a standard deviation of 17.3%. Looking back, this was a rather high SD when measured against the projected return. Within one standard deviation or a one sigma event, the portfolio would range from a high of (8.3% + 17.3%) or 25.6% to a low of (8.3% – 17.3%) negative 9.0% (-9.0%). However, the portfolio dropped -30.6% during 2008. For the Schrodinger, it was nearly a two sigma event.
A two sigma event would see the range of variance expand from the average to include 95% of the data. The standard deviation must be doubled. We now have a two standard deviation or two sigma event. The range of outcomes for the two sigma event would be 8.3% plus or minus 2 x 17.3% or range from 8.3% + 34.6% to 8.3% – 34.6%. Now a two sigma event captures 95% of the outcomes so the two sigma range is from 42.9% on the upside to -26.3% on the downside.? Once more, remember the Schrodinger dropped more than 26.3%. Therefore, what happened to this portfolio in 2008 was a two to three sigma event. A three sigma event includes 99% of the outcomes. Actually, it is 99.73% of all events, so you gain an appreciation of how unusual it was for a drop of -30.6% to occur.
The question we should have been asking ourselves in late 2007 was – "How much risk are we capable of handling?" Is a three sigma event acceptable? If not, what do we need to do to make sure this does not happen again? Is your portfolio positioned properly for today’s market?
It is important to hold down the projected risk and that is why we work hard to bring it under 15%. Since the bear market of 2008 and early 2009, we have also added the ITA Risk Reduction model for several portfolios. This is an experiment to see if we can enhance return while attempting to constrain portfolio uncertainty as we find two and three sigma events to the downside to be unacceptable.
* This blog was first written about two years ago.