Excessive Crowding of “Low Risk” Stocks: Size & Leverage Analysis

Executive Summary:

• Applied Finance produces robust intrinsic value estimates to capture changing relationships between market prices and firm value.  A valuation gap refers to notable differences in the valuation-based characteristics observed on large baskets of stocks.
• A significant valuation gap has formed between “high risk” and “low risk” equities.  “Low risk” stocks currently trade at historic premiums last observed at the peak of the tech bubble.
• Previous seasons with similar valuation gap characteristics have been followed by significant outperformance of “high risk” stocks, while “low risk” stocks and passive alternatives lagged due to high levels of correlation between each investment approach.
• Investors should consider strategic allocation that seeks out valuation-based active management.

In the 2020 Q2 Applied Finance quarterly write-up, we reviewed recent performance of “high risk” and “low risk” portfolios based on relative size and leverage characteristics across the Russell 3000.  Namely, for the two decades prior to 2017, a notable risk premium had been evident for smaller, highly levered firms.  Since the start of 2017, however, “low risk” equities have drastically outperformed “high risk” equities, which is a notable observation alongside the significant increase in zero- and negative-yield debt that may be distorting asset allocation decisions in search of returns and yield.

To fully explore this theme, we have aggregated valuation data to understand the investment characteristics of “high risk” and “low risk” portfolios.  Based on this, investors would be wise to take note of the significant valuation differences based on these commonly applied measures of risk.  “Low risk” stocks appear to be trading at significant valuation premiums, while “high risk” stocks appear to be trading much closer to normal market prices.  This creates an interesting paradox: “low risk” equities are trading at market prices that actually make them quite risky, while “high risk” equities are trading at market prices that make their investment prospects appear much more attractive than their “low risk” peers.

Russell 3000: Median Valuation Upside/Downside (Size vs. Leverage), 7/31/20

AFG Research Database: Russell 3000 Quintile Analysis, Median Percent to Target – Current, 7/31/20

The first data table displays median valuation levels relative to current market prices for baskets of stocks based on Size (Market Cap) and Leverage (Debt / Enterprise Value) as of July 31st.  On a single factor basis, the lowest leverage quintile currently has a median valuation upside of -60.4% compared to 12.6% for the highest quintile.  Market cap trends are muted, yet similar: The quintile of largest stocks currently has a median valuation upside of -31.6%, compared to -9.4% for the smallest quintile.  The 5×5 display in this table further highlights the intersection of size and leverage quintiles; for example, the largest cap / lowest leverage portfolio has a median valuation upside of -68.0% compared to 9.2% for the smallest cap / highest leverage portfolio.

We then further aggregate these quintile-based portfolios into “high risk” and “low risk” composites, based on the light green/light red shading in this table.  “Low risk”, in green, captures stocks that tend to be a combination of lower leverage and larger market cap.  “High risk”, in red, captures stocks that tend to be a combination of higher leverage and smaller market cap.  Based on this broad definition, each portfolio covers more than 40% of the Russell 3000 forecast universe.  The immediate decay assumption applied in the Applied Finance intrinsic value calculation is inherently conservative, but it is noteworthy that such a significant valuation gap exists between these broadly defined baskets of “high risk” and “low risk” equities on a relative basis.  The “high risk” portfolio is trading at market prices fairly close to intrinsic value estimates, while the “low risk” portfolio has intrinsic value estimates more than 40% below current market prices.

Russell 3000 Z-Score Analysis: Valuation Upside/Downside (Size vs. Leverage), 7/31/20

AFG Research Database: Russell 3000 Quintile Analysis, Normalized Percent to Target – Current, 7/31/20

Additionally, we can normalize the 7/31/20 raw data against the array of historic median levels for each portfolio.  Based on this z-score approach, the “high risk” portfolio is trading at market prices that suggest mild upside against long-term normal with a z-score of 0.16.  “Low risk” stocks, however, currently reflect an intrinsic value to market price relationship more than 2 standard deviations below normal.

High Risk vs. Low Risk Composites: Median Valuation Upside/Downside, 9/30/98 – 7/31/20

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

Plotting out the raw median valuation characteristics of these “high risk” and “low risk” composites can help us better understand a few meaningful themes.  In most market environments, “high risk” and “low risk” portfolios have identical aggregate valuation characteristics.  Size and leverage are inputs into firm-specific discount rate adjustments, meaning that relevant differences in firm riskiness should be captured in the intrinsic value estimate.

Also, several notable valuation gaps have been observed around the tech bubble, the financial crisis, and the more recent trend that intensified around the COVID crisis.  We can explore each of these in more detail.

As a baseline to our regime study, we can note that “high risk” stocks have outperformed “low risk” stocks, as well as the S&P 500 and Russell 3000 benchmarks, over the complete study horizon.  This highlights the aggregate risk premium of size and leverage risk factors.  In addition to this, we can note that cap-weighted benchmarks tend to be highly correlated to “low risk” portfolio performance due to the cap-weighting similarities and the inclusion of largest market cap stocks in the “low risk” construct.

AFG Research Database: High Risk and Low Risk Portfolio Analysis, Cap-weighted total returns, 9/30/98 – 7/31/20

High Risk vs. Low Risk Valuation Gap – Tech Bubble

During the tech bubble, the peak valuation gap (based on calendar month end data) was observed on 2/29/2000, when the “high risk” portfolio reflected a median valuation upside of 25.3% compared to

-43.1% for the “low risk” portfolio.  From that point forward, “high risk” stocks outperformed “low risk” stocks significantly over the next seven years.  It is worth noting that this observation prefaced the peak of the tech bubble by only 23 days.

AFG Research Database: High Risk and Low Risk Portfolio Analysis, Cap-weighted total returns, 2/29/00 – 5/31/07

Immediate outperformance was delivered over the following 6 months beyond 2/29/00, as the basket of “high risk” stocks appreciated by 17.9% on a cap-weighted basis while “low risk” stocks only appreciated by 8.6%, lagging the “high risk” portfolio by 9.3%.  This became even more pronounced over the 12- and 24-month horizons following 2/29/000, as “high risk” stocks continued to appreciate while “low risk” stocks sold-off as the tech bubble collapse continued.

This performance trend continued through mid-2007; across this 7-year cycle, “high risk” stocks appreciated by 219.3% while the “low risk” portfolio only appreciated by 14.4%.  S&P 500 and Russell 3000 index returns were slightly ahead of the “low risk” basket of stocks, but the correlations between the S&P 500, Russell 3000, and “low risk” portfolio approached 1.00 over this study regime, while the correlation of “high risk” equities was much lower at 0.81

High Risk vs. Low Risk Valuation Gap – Financial Crisis

Shifting our focus to the financial crisis, the peak valuation gap was observed on 2/28/09, when the “high risk” portfolio reflected a median valuation upside of 68.0% compared to 19.1% for the “low risk” portfolio.  Like the tech bubble theme, the “high risk” portfolio drastically outperformed the “low risk” portfolio on a cap-weighted basis from that point forward over the next two years.  It is similarly noteworthy that this observation prefaced the pivot to market recovery by only 9 days.

AFG Research Database: High Risk and Low Risk Portfolio Analysis, Cap-weighted total returns, 2/28/09 – 4/30/11

In the initial 6 months following the peak valuation gap observation, “high risk” stocks outperformed “low risk” stocks by 53.0%, and this expanded to 70.2% after a full year had passed.  This theme lasted slightly longer than two years, with “high risk” stocks eventually outperforming their “low risk” peers by 126.0%.

A similar correlation theme as noted in the tech bubble data also emerges here.  S&P 500, Russell 3000, and “low risk” portfolio correlations reach 1.00, while the “high risk” portfolio correlation to the S&P 500 is only 0.85.

High Risk vs. Low Risk Valuation Gap – COVID Crisis

Most recently, we have observed a widening valuation gap between “high risk” and “low risk” equities since the start of 2017, but this trend significantly intensified as COVID fears gripped investors in Q1 and “low risk” equities were viewed as being much more capable of withstanding the uncertain duration of an economic shutdown.  The peak valuation gap was observed on 3/31/20, when the “high risk” portfolio reflected a median valuation upside of 50.2% compared to -19.4% for the “low risk” portfolio.  This valuation gap of 69.5% between “high risk” and “low risk” portfolios is the largest on record (slightly surpassing the 2/29/00 gap of 68.4%).  It is also worth noting that this peak valuation gap was observed only 8 days after the market bottom had formed on 3/23.

AFG Research Database: High Risk and Low Risk Portfolio Analysis, Cap-weighted total returns, 3/31/20 –7/31/20

Over the four months following the valuation gap peak on 3/31, the “high risk” portfolio has outperformed the “low risk” portfolio by 8.8%.  This relative outperformance is likely due to the extreme valuation upside offered for “high risk” equities at the end of Q1, which has largely normalized as “high risk” stocks recovered in Q2.  As of the end of July, “low risk” stocks appear to still trade at excessive market prices (-42.3% median valuation upside) relative to their “high risk” peers (-5.1% median valuation upside).

Despite limited observations since March, a similar correlation theme is emerging.  The correlation of S&P 500, Russell 3000 and the “low risk” portfolio is also 1.00, while the “high risk” portfolio correlation to the S&P 500 is lower at 0.92.

Russell 3000 Z-Score Analysis: High Risk vs. Low Risk Composites, 9/30/98 – 7/31/20

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

As a final point, we can revisit the time series plot of valuation upside median data on a normalized basis.  The current z-score of -2.17 for the “low risk” portfolio is a level that has not been observed since the early 2000s, which prefaced a massive multiyear run in favor of “high risk” equities.  “Low risk” stocks sold off as the tech bubble collapsed, also creating significant drag on cap-weighted index returns.  The market prices achieved by “low risk” stocks in early 2000 had appreciated to such unsustainable levels, leading to meager performance for “low risk” stocks and benchmark returns through mid-2007.

This resonates with valuation trends in the current market place, as a number of stocks that have performed extraordinarily well through the COVID crisis now trade at market levels that make it difficult to assume further absolute gains on a forward-looking basis.  If a sell-off of these stocks were to occur, their higher benchmark weights would drag on overall benchmark performance; even without a sell-off, investors would be wise to temper return expectations for these stocks (with the caveat that irrational behavior may continue to lead to further crowding in the near-term as pandemic fears persist or zero- and negative-yield debt continues to proliferate).

Conclusion

This research should serve as a bellwether for relative performance expectations on a forward-looking basis.  Investors should consider actively managed portfolios based on robust valuation techniques to exploit this historic valuation gap.  Passive investors should also take note that common index construction tends to align towards this “low risk” classification, and cap-weighted broad market investment vehicles will lag as these valuation characteristics normalize.

Contact your Applied Finance representative to discuss meaningful ways to incorporate our valuation expertise into your firm’s equity allocation.

Definitions 

• Intrinsic Value -A firm’s estimated equity value derived from Applied Finance’s Economic Margin® valuation framework. Applied Finance has used its Economic Margin® valuation framework to calculate over 20 million out of sample valuations since 1998, incorporating Applied Finance’s proprietary risk and Economic Profit Horizon™ estimates… more
• Percent to Target – Current – Compares the intrinsic value estimate to the most recent closing price for each firm.