Factor Performance Review
We track the performance of six factors (growth, quality, low-volatility, momentum, size, and value) as part of our multi-factor strategy. Over the past month, the best factors were growth and size (small-caps), with each factor outperforming by 2.2%. The worst factors over the past month were low-volatility (down 2.8% vs. the S&P 500) and value. Performance for each of the six factors over the past month is shown as the gray bars in Fig. 1.
Looking back over a 3-month period (blue bars in Fig. 1), size and growth have seen the best performance. Although the size factor has seen strong performance in recent months, it still lags the other factors, as it has underperformed the overall index by 8.1% on a trailing 12-month basis.
Multi-Factor Portfolio Performance Review
We track a dynamic multi-factor portfolio that tilts weight toward the factors with the best recent performance, and away from the factors with the worst recent performance. Fig. 2 shows the cumulative performance of this dynamic multi-factor strategy relative to the S&P 500 since 1997.
From the start of 2020 through August 5, 2022, the dynamic multi-factor strategy returned 34.8%. Over that same period, the S&P 500 gained 28.3%, for 6.5% of outperformance for the dynamic multi-factor strategy. Fig. 3 below shows the monthly performance of the dynamic strategy vs. the S&P 500 since the start of 2020. The dynamic strategy saw its streak of outperforming the S&P 500 end in July, as it underperformed the overall index by 1.9%. An underweight of growth and an overweight toward value contributed to the underperformance for the dynamic factor strategy in July.
Dynamic Model: Factor Weights for August
Fig. 4 below indicates the latest weights assigned to each of the six factors in the dynamic multi-factor strategy. For the next month, the dynamic strategy is overweight the growth and size (small-cap) factors while being underweight value and quality.
Baseline Stock Selection Model: Performance and Discussion
We have developed a stock selection framework that uses composite factors across five dimensions (value, quality, momentum, estimates, and investment) to predict stock performance. The model produces a list of 100 favored investments from across the S&P 500. Fig. 5 below shows the historical performance of the basket of favored stocks, rebalanced monthly (orange line) and the S&P 500 (black dotted line).
Fig. 6 (next page) shows the performance during July for each of the 5 composite factors that make up the stock selection model (blue bars), along with the performance of the overall model (orange bar at right). After underperforming in June for the first time in 2022, the model struggled mightily in July, underperforming the S&P 500 by 3%. Despite the recent underperformance, the model’s basket of preferred stocks is up 2.2% vs the S&P 500 year-to-date.
Of the five of the composite factors that make up the model, four underperformed in July, with only the quality factor eking out a minimal gain relative to the index. The value, estimates and investment factors underperformed the S&P 500 by between 1.0% and 1.5% for the month, but momentum was by far the worst factor in July, losing 4.4% relative to the index.
For perspective, the 4.4% underperformance of momentum in a single month has only been exceeded two other times in our historical database – January 2001 and November 2002. Those months coincided with unexpected rate cuts by the Federal Reserve. One interpretation, therefore, of the poor performance in momentum during July is that the market considered the most recent Fed commentary to be akin to a surprise rate cut.
While it is difficult to draw conclusions from a sample set of two observations, it is worth noting that after the S&P 500 rallied in both January 2001 (by 3.5%) and November 2002 (by 5.7%), performance over the subsequent 3 months was poor. Specifically, the S&P 500 lost 8.5% from February through April 2001 and lost 10.2% from December 2002 through February 2003.
Market Valuation: Residual Income Model
We use a residual income model to value the market. The residual income model produces an estimate for the equity risk premium, or the additional return that equity investors are compensated over the risk-free rate. The history of the equity risk premium is shown in Fig. 7. At the end of July, the equity risk premium implied by the model was 3.63%. This value, while it falls within the recent historical range of 3-5%, represents a decrease from the end-June observation of 3.97%. A decrease in the equity risk premium indicates a reawakening of “animal spirits” among equity investors.
Using the equity risk premium, we can evaluate the relative attractiveness of equities compared to investment grade fixed income via the ratio of their yields. Historically, when equities are expensive compared to fixed income (i.e., equities have a relatively low yield) the stock market experiences smaller average returns and higher volatility over the subsequent quarter (see Fig. 8).
At the end of July, the yield ratio indicated that equities continued to remain in the overvalued state, though the degree of overvaluation has come down compared to levels seen earlier this year. Based on the above relationship, we continue to expect muted returns and higher equity market volatility over the next 3 months, though we will closely monitor this gauge for signs that the equity valuation has moved to a more constructive level.
 See Market Valuation Report