AFG Quantitative Review – Q3 2020

The transition into Q4 is accompanied by a notable milestone. The Applied Finance Group turns 25 years old today!

We want to take this opportunity to thank our research partners that appreciate the merit of a valuation-based discipline when developing longterm investment track records and creating a competitive edge in their investment process. As a firm, the Applied Finance Group has now performed more than 20,000,000 valuations, and we continue to sharpen our expertise and understanding regarding economic measure of corporate performance and intrinsic value. We look forward to building upon our research partnerships and strategy track records as we transition to the next 20,000,000 valuations.

Focusing on the previous quarter, US equity markets continued to appreciate, peaking in early September before a mild sell-off for most stocks that was more sharply pronounced for the top five stocks by market cap (AAPL, MSFT, AMZN, GOOGL, FB). In general, the stock market continues to be dominated by uncertainty over potential COVID flare-ups and economic lockdowns. Investors are continuing to show strong preference towards megacap technology and stocks that benefit from the ongoing digitization of our economy.

Many of the themes that we reviewed in last quarter’s review are still relevant today regarding the top five concentration of the stock market, the valuation gap between “high risk” and “low risk” equities, and the sharp contrast in valuation characteristics between value and growth stocks. This quarter, we have also taken note of a valuation gap that has formed between high yield stocks and their no yield peers based on the investment trends that have unfolded over the course of this year.

On a forward-looking basis, the relative performance of various investment themes will be directly impacted by health care data and policy decisions, so we want to explore recent COVID trends that will influence policy and shape investor preference as either uncertainty continues, or promising signs of normalization arise. This directly applies to portfolio construction focused on the significant valuation gaps that have formed on style, yield and risk themes. In the long-run, these valuation gaps will close, but it may also be dangerous in the near-term to take a contrarian view against the recent trends favoring growth and megacap technology. We are hopeful that ongoing monitoring of these datasets will help our clients continue to confidently apply a valuation-based strategic allocation, while considering tactical allocations towards these valuation gaps as promising signs that would encourage a shift in investor preference appear more likely.

We also have updated the performance and constituents of portfolios formed on the valuation/cheapness gap between Percent to Target Current and Book to Price. This write-up also includes Q3 attribution reports for AFG’s turnkey strategies, the AFG 50, AFG High Dividend and Strategic Valuation portfolios, as well as an overview of AFG metric performance globally.

Concentration of Top 5 Market Cap Companies in S&P500 (Update)

A dominant theme of investor preference in the COVID era is defined by a “flight to safety” with preference towards megacap stocks assumed to better navigate a prolonged economic shutdown. This trend persisted in Q3, as the five largest stocks appreciated 13.3% on a cap-weighted basis compared to 10.6% for the next 95 largest stocks and only 6.8% for the remaining 400 stocks in the S&P 500.

AFG Research Database – Russell 3000 Size Analysis: AFG Data as of 9/30/20.
Market Cap and Market Weight data based on Market Cap data as of 6/30/20 & 9/2/20

Last quarter, we reviewed the increasing concentration of the top 5 stocks on a cap-weighted basis in the S&P 500, and their accompanying valuation characteristics. The five largest stocks (AAPL, MSFT, AMZN, GOOGL, FB) continued to appreciate faster than the remaining basket of their S&P 500 peers through early September to reach nearly 25% of the overall index before lagging over the last several weeks of the quarter.

As of 9/30, this group of the five largest stocks compromised 23.2% of the S&P500 (falling from 24.6% as of 8/31). Since 1998, concentration previously peaked at 18.3% during the tech bubble, with CSCO and GE notably trading at implied sales growth levels in excess of 55%. For additional reference, we have also mapped the top 5 stocks at the end of each calendar year in the table below.

S&P 500 Top 5 Market Cap Constituent Analysis, 9/30/98 to 9/30/20. Line Graph reflects monthly data.
Table ranks based on year-end market cap data; 2020 ranks based on 9/30/20 market cap data.

Last quarter, we reviewed these largest stocks and noted that they offered reasonable valuation characteristics on an implied sales basis despite plunging percent to target valuation levels since the end of March. Default percent to target levels have fallen further as these stocks appreciated in Q3, with a median percent to target of -36.2% observed at the end of August. Valuation prospects mildly improved as these five largest stocks lagged in September.

Revisiting implied sales analysis, we can shift our focus to 2021 implied expectations. Incorporating new information from pricing and analyst forecasts, we still see GOOGL and FB trading at implied sales growth well below consensus forecasts. AAPL now appears closer to fairly-valued following its appreciation in Q3, while MSFT and AMZN continue to trade at implied sales levels outpacing their forecasted growth.

Value Expectations: Implied Sales Growth vs. 2021 Consensus Growth
(Implied Sales Growth based on 9/30 prices. Forecast sales growth is Zack’s consensus as of 10/2/20)

Dividend Yield Analysis

Dividend Yield Valuation Gap

Over the last nine months, high yield stocks (dividend yield >= 3%) underperformed no yield peers (dividend yield = 0%) globally by 23.7%. This trend has been strongest in North America where no yield stocks outperformed high yield stocks by 31.2%, while spreads have been 28.1% in Asia and 15.9% in Europe. No yield stocks in the S&P 500 have outperformed high yield stocks by 28.9%; these return spreads are 33.5% in the Russell 1000 and 26.5% in the Russell 2000.

AFG Research Database from 12/31/19 to 9/30/20, Monthly Rebalancing, Average Returns
Universe Size: 5,300 to 5,600 Companies – Global Database | All Sectors | 500M+ USD* | ADRs Excluded
Universe Size: 900 to 1,000 Companies – Russell 1000 | All Sectors | All Cap | ADRs Excluded
*Size Tiers include all firms that meet the market cap or index criteria.

At first glance, it is convenient to assume that the underperformance of high yield is attributable to the lagging performance of sectors with higher concentrations of high yield stocks, including Financials, Energy, Real Estate and Utilities. Narrowing our focus on individual sectors in the Russell 1000, however, we can confirm that high yield stocks have underperformed sector peers on an equal-weighted basis across every sector except Energy.

Plotting a time series of the normalized Percent to Target levels of high yield stocks against their no yield peers offers evidence of an interesting valuation gap that has recently formed regarding dividend yield characteristics.

Using the Russell 3000 universe, the median Percent to Target of high yield stocks is currently 5.2%, which is 0.95 standard deviations above normal levels. No yield peers, on the other hand, have a median Percent to Target of -44.9%, which is 1.63 standard deviations below normal.

Alternatively, we can view this on a cap-weighted basis, where the high yield Percent to Target of 16.8% is 0.20 standard deviations above normal while no yield stocks offer a -36.4% Percent to Target, which is 1.56 standard deviations below normal. Historical trends of normalized valuation levels are presented in the charts below on a median and cap-weighted basis.

AFG Research Database from 9/30/98 to 9/30/20, Monthly Rebalancing
Normalized Median / Market-Cap Weighted Percent to Target Current Levels
Universe Size: 2,000 to 2,700 Companies – Russell 3000 | All Sectors | All Cap | ADRs Excluded

Style Analysis

Style Valuation Gap

Over the last nine months, the highest quintile of Market Value / Net Invested Capital (MVIC), or growth, stocks outperformed the lowest quintile by 26.6% globally. This trend is also clearly observable across all regions, US indices, and the ADR universe. Style-based themes based on MVIC align with the recent performance trends of the commonly-used style approaches applied by most index methodologies.

AFG Research Database from 12/31/19 to 9/30/20, Monthly Rebalancing, Average Returns
Universe Size: 5,400 to 5,700 Companies – Global Database | All Sectors | 500M+ USD* | ADRs Excluded
Universe Size: 900 to 1,000 Companies – Russell 1000 | All Sectors | All Cap | ADRs Excluded
*Size Tiers include all firms that meet the market cap or index criteria.

MVIC grades reflect an inverse rank. A-graded stocks would imply a low MVIC sector rank, while F-graded stocks imply a high MVIC sector rank. The year-to-date outperformance of the lowest MVIC quintile (based on US sector ranks) is observable across all sectors in the Russell 1000 with the exception of Utilities.

We can form style classifications based on the top half (growth) and bottom half (value) of US MVIC sector ranks. Plotting a time series of the normalized Percent to Target levels offers similar evidence of the valuation gap that has formed between value and growth stocks.

In the Russell 3000 universe, the median Percent to Target of low MVIC (value) stocks is currently 9.5%, which is 0.42 standard deviations above normal levels. High MVIC (growth) peers, on the other hand, have a median Percent to Target of -45.4%, which is 2.22 standard deviations below normal.

We can also view this on a cap-weighted basis. The bottom half of MVIC currently has a Percent to Target level of 33.0%, which is 0.67 standard deviations above normal. The top half of MVIC stocks offer a -30.2% Percent to Target, which is 1.68 standard deviations below normal. Historical trends of normalized valuation levels are presented in the charts below on a median and cap-weighted basis.

AFG Research Database from 9/30/98 to 9/30/20, Monthly Rebalancing
Normalized Median / Market-Cap Weighted Percent to Target Current Levels
Universe Size: 2,000 to 2,700 Companies – Russell 3000 | All Sectors | All Cap | ADRs Excluded

Crowding of “Low Risk” Stocks: Size & Leverage Analysis (Update)

Last quarter, we also reviewed a notable valuation gap that had formed between “high risk” stocks (based on smaller market caps and higher levels of leverage) and their “low risk” peers. High risk stocks strongly outperformed in Q2 following significant underperformance in Q1. In the third quarter, the performance gap narrowed as high-risk stocks lagged their low risk peers by 1.7% on a cap-weighted basis.

Over the previous two quarters, the broad portfolio of high-risk stocks has appreciated by 40.9% while the low risk portfolio has appreciated by 34.1%. More importantly, the valuation gap, defined by the median percent to target of stocks within each basket, has diverged to nearly 40%, implying large cap stocks with low leverage are trading at market prices that are actually quite risky, while smaller cap stocks with high levels of leverage are trading in-line with long-term “normal” levels.

Comparing 9/30 median percent to target data to the array of month-end observations since September 1998, the current valuation levels of the low risk portfolio at -41.8% reflects a standardized z-score of -2.08. The high-risk basket, on the other hand, trades at a median percent to target level of -3.3%, reflecting a zscore of 0.28.

AFG Research Database: Russell 3000 Two-Factor Quintile Analysis (Independent
Ranking for Russell 3000 constituents that have valid data for each factor)
Market Cap Weighted Returns from 9/30/98 to 9/30/20

Investors should be aware of the eventual reversal of this multiyear trend that has been further exacerbated by the flight to large cap growth and large cap tech in response to COVID impacts on the economy. As investor preference eventually expands to include the real economy (hospitality, retail, office space) at some point, this will likely lead to lagged performance of cap-weighted benchmarks and “low risk” strategies. For example, in the four years following the peak of the tech bubble in early 2000 through the end of 2004, “high risk” stocks outperformed the Russell 3000 by 18.9% per year as megacap stocks sold-off sharply through 2002 and continued to lag in the ensuing recovery.

AFG Research Database: High Risk and Low Risk Portfolio Analysis, Median Percent to Target – Current, 9/30/98 – 7/31/20

AFG Research Database: High Risk and Low Risk Portfolio Analysis, Cap-weighted total returns, 3/31/20 –9/30/20

Covid-19 Consumer Update – US

Given the continued prevalence of Covid19 in our daily lives, we would like to share some current consumer data that may provide meaningful insight to gauge allocation decisions between firms focused on the digital economy vs. the tangible and service-based economy, which likely is the root of the significant valuation gaps that we have studied. The charts below display data for the US and international trends.

The map below is regularly updated in the Wall Street Journal and shows current hotspots for Covid-19 in the US. In recent weeks, the virus has moved on from larger urban areas to middle America. In the first half of October, North Dakota, South Dakota, Wisconsin, Montana, and Idaho are currently facing the highest outbreaks per capita. The virus began in urban areas in the Northeast & Northwest, moved to the South and Southwest, and is now hitting some of the final previously calm states.

Wall Street Journal published Johns Hopkins Heat Map: Covid-19 Cases per Capita

Hotel occupancy fell sharply beginning in March, hitting a low of 24.5% compared to 67.8% in 2019 for the month of April. The industry has seen a gradual recovery since then, but current figures are still well below last year’s data. It is worth noting however, that there is a pronounced trend towards normal occupancy levels, and an eventual recovery may be rapid.

American Hotel & Lodging Association Report – State of the Hotel Industry Analysis: Covid-19 Six Months Later

The airline industry saw one of the most dramatic collapses in history during the March panic. TSA Passenger Throughput data dropped precipitously in only 3 weeks. From the lowest points in April, recovery was initially quick through July, but has remained somewhat stagnant since then. Current passenger data shows a drop of 67% at the end of September for 2020 compared to 2019.

TSA Passenger Throughput Data: Rolling 7-day average, 3/1/20 – 9/30/20

Covid-19 Consumer Update – International

The chart below presents global Google Mobility data for tracking specific activities during the Covid-19 pandemic. In the initial panic in March, Retail/Recreation, Transit, and Workplace trends saw significant drops, down 50% or more on a mobility basis. Residential and Parks categories diverged, with Residential up ~25%, and Parks usage was down ~25%. Towards the end of April/early May, mobility behavior began to normalize with all categories trending towards their baseline, with the exception of Parks. As weather improved in summer months, park mobility increased drastically and is now tapering off as colder weather returns.

7 day moving average for each G7 country, then average of the 7 countries for each category.
G7 Countries include US, Canada, UK, Germany, France, Italy & Japan.
Baseline developed from median data 1/3/20 – 2/6/20. Data in chart from 2/21/20 – 10/4/20.

Two conclusions can likely be drawn from the chart: 1) people react quickly to new information and 2) there is a definite trend towards normalizing behavior (the existing baseline measure of activity). It is uncertain if these trends will continue over the next two quarters if we are faced with a spike in cases and hospitalizations this winter that lead to policy decisions that enforce sheltering in place, but it should be reasonable to assume that many individuals will eventually resume normal mobility as the COVID health crisis passes.

Recently, Europe has surpassed the US in new cases per day on a rolling 7-day basis. After having an extremely calm summer, European countries are now seeing fresh outbreaks and new restrictions are beginning across the continent. From an equity standpoint, Q4 data for Europe will likely be revised lower over the next few weeks as many cities in France and the UK are instituting partial lockdowns again.

Wall Street Journal: “Europe Overtakes U.S. in New Cases of Covid-19”, By Jason Douglas, Stacy Meichtry and Andrew Barnett (Oct 14)

Current EM Momentum grades for the Dow Jones Euro Stoxx index are around normalized levels, with most companies having minor upward or downward revisions. If restrictions become harsher over the next few weeks, we would expect the EM Momentum grade distribution to lean more towards the D and F buckets, which could negatively affect overall grades in Europe in the near term.

EM Grade Distribution: Applied Finance Database as of 10/16/20

Valuation Minus Cheapness in Bull/Bear Markets (Update)

Over the course of this year, we have reviewed Valuation Minus Cheapness themes across various bull and bear market regimes since 1998. The table below provides a performance update through the end of Q3. Since the COVID sell-off began in mid-February, the Valuation Preference portfolio has outperformed the Cheapness Preference portfolio 6.1%.

From the market bottom on March 23, Valuation Preference stocks have underperformed Cheapness Preference stocks by 260 basis points. In addition to this, the Strategic Valuation portfolio based on this research has outperformed the S&P 500 by 3.2% between March 23 and September 30th after lagging by 2.3% through the initial sell-off in February and March.

AFG Research Database – S&P 500 Market Cycle Analysis, Equal-Weight in Sector Returns from 9/30/98 to 9/30/20