The StarMine team has published the most recent StarMine Monthly Performance Report which looks at how certain stock market factors performed in September. We observed a sharp reversal between momentum and value, where value has significantly outperformed momentum across all regions. This is noteworthy as momentum has been outperforming value across most regions over the last 12 months. As shown in Exhibit 1, momentum is represented by the Analyst Revision Model and Price Momentum model. Value, which identifies both expensive and undervalued companies, is represented by the Relative Valuation and Intrinsic Valuation model. Exhibit 2 also highlights this picture, where the Russell 1000 Value Index increased 3.4% in September 2019, whereas the Russell 1000 Growth Index declined -0.08%.
Exhibit 1: September 2019 Factor Performance Across All Regions
Exhibit 2: Russell 1000 Growth vs. Value Index
Seeing a mood change
One possible explanation for this reversal is investors have temporarily stopped chasing growth in light of on-going trade talks, political unrest in Hong Kong, and a Brexit deadline approaching at the end of the month.
The U.S. was the strongest beneficiary of this mood change, as it achieved a decile spread of 9.4% and 15.8% respectively across Intrinsic Valuation and Relative valuation. A decile spread is a long/short strategy where investors buy (go long) the top decile of companies with the strongest model score and sell (go short) the bottom decile of companies with the worst model score. The desired outcome is for the long portfolio to gain in price and for the short portfolio to decline in price, which allows investors to ‘double-dip’ and generate returns on both sides.
If we anticipate value continuing to outperform momentum, we can use Eikon by Refinitiv to create our own long/short portfolio. Focusing on the Relative Valuation Model, this looks at how cheap or expensive a company is based on six ratios: P/E, P/B/ P/CF, EV/EBITDA, EV/Revenue, and Dividend Yield. As a reminder, a high Relative Valuation model score indicates that a company is inexpensive (cheap), whereas a low model score indicates that a company is expensive (rich). The cheaper a company, the higher score it will receive.
Exhibits 3 and 4 identify constituents with top/bottom decile scores within the S&P 500 which can form the long and short portfolios. Exhibit 3 represents the long portfolio (top decile constituents with a Relative Valuation score of 91-100) and Exhibit 4 represents the short portfolio (constituents with a score of 1-10).
Exhibit 3: Top Decile of S&P 500 Constituents with Strongest Relative Valuation Score
Exhibit 4: Bottom Decile of S&P 500 constituents with worst Relative Valuation Score
Going long or short
From the above exhibits, we can create two portfolios: a long portfolio (equal-weight of 50 constituents in Exhibit 3) and a short portfolio (equal-weight of 50 constituents in Exhibit 4). We can also see the company level ranking at a sector and industry level to provide additional granularity as companies are valued differently depending on the industry they are in.
Exhibit 3 shows U.S. retailers looking cheap on a Relative Valuation basis, as this group has been punished over the last year due to the rise of e-commerce. Macys, L Brands, Gap, Kohls, Tapestry, and Nordstrom have all suffered large price declines over the last year. Exhibit 4 includes high-tech growth companies including Netflix, Amazon, and Salesforce.com. We caveat this by mentioning that companies may appear cheap, but for good reason. To help eliminate potential “value trap” stocks, we would also turn to our Earnings Quality and Credit Risk models to filter out those with low scores.
It will be interesting to see if the rotation from Momentum into Value persists. If you wish to stay informed on factor performance, please register for a complimentary subscription at http://solutions.refinitiv.com/LipperAlphaInsightsubscription.
