Applied Finance has spent the last several years exploring the contrast between intrinsic value and other “value” classifications commonly referenced in equity investing. Our Valuation Beta research paper highlighted the overlooked role of intrinsic value in asset pricing, where a firm’s value in excess of the book equity reported on their balance sheet provides a source of unexplained alpha when regressed against book-to-price, or the “value factor”.
While the Valuation Beta study challenges book-to-price’s legacy application in asset pricing, deconstruction of other popular price multiples with intrinsic value highlights that valuation shortcuts tend to only be successful to the degree that they are confirmed by a comprehensive intrinsic value thesis. As stand-alone factors, price multiples do not provide alpha against a valuation-based asset pricing model, and “cheap” price multiple classifications unsupported by intrinsic value provide significant negative alpha. This “valuation vs. cheapness” contrast offers insight that there are no simple shortcuts to properly estimate the intrinsic value of a firm.
With convincing, out-of-sample proof of the role intrinsic value plays in asset pricing, we can shift our focus towards the other dominant secular trend in equity allocation over the last decade: passive investing. For this study, we will include an additional measure of “value” based on the market value of a firm’s equity, or its market capitalization. This will allow us to compare the performance between intrinsic value, book value, and market value weightings applied to the S&P 500’s constituents and determine which approach represents the optimal market portfolio, as well as which approach offers investors the largest strategic advantage against market cap weighted benchmarks.
Alternative “Value” Classifications:
- Market Value (Market Cap): Market cap is the most common measure of value applied in passive investing (or a mildly adjusted alternative based on free float). Its use in passive indexing is established through the Efficient Market Hypothesis, which states that market prices reflect all known information, while presuming consistent alpha generation is impossible.
- Intrinsic Value: Intrinsic value is an estimate of the present value of future cash flows less any opportunity costs derived from a firm over the life of the business claimed by its common equity owners. Applied Finance’s Economic Margin framework measures corporate performance from an economic perspective. We then assume that abnormal levels of economic profitability and organic reinvestment will be competed away to long-term normal levels, which avoids a reliance on terminal value assumptions.
- Book Value (Common Stock): Book value reflects the common stock reported on a firm’s balance sheet from its most recent 10-K. Book value is assumed to offer a stable measure of value useful in asset pricing literature and factor research. Its acceptance in asset pricing was popularized through Fama French’s expansion to three factors (HML “value” and SMB “size”) in addition to the the cap weighted market return.
- Excess Intrinsic Value: Each firm’s intrinsic value estimate can be bisected into the amount already reflected as book value (common stock) and its Excess Intrinsic Value that represents the amount of intrinsic value not captured on the firm’s balance sheet. Excess Intrinsic Value is created through profitable investment and helps capture the wealth compounding benefits offered to shareholders as profitable firms reinvest organic cash flows.
In this study, we will challenge the Efficient Market Hypothesis assumption that market prices adequately capture each firm’s intrinsic value characteristics. We will provide evidence that an intrinsic value weighted index outperforms a market cap weighted index on a risk-adjusted basis over Applied Finance’s horizon of weekly live production datasets between October 1998 and May 2021. We will also study changing active share levels over time to understand the varying degree in which market value proxies intrinsic value across different market regimes.
We will then deconstruct intrinsic value into its asset pricing components (common stock and Excess Intrinsic Value) to understand the unique role each offer in value weighted indexing, then overlay the Financing Yield factor to eliminate exposure to stocks diluting existing ownership with external financing. This path reconciles the observations from this passive indexing study with our asset pricing study, while also highlighting the enhanced valuation-based indexing characteristics offered by Applied Finance’s Valuation Stewardship strategy.
Testing the Efficient Market Hypothesis: Comparing Market Cap & Intrinsic Value Weighted Indices
We begin this study by comparing the performance and active share characteristics of an intrinsic value weighted index against a market cap weighted benchmark to test the underlying assumption of the Efficient Market Hypothesis, which presumes market prices are adequate proxies to intrinsic value. As of each calendar month end, we form two portfolios using the equation below; one that weights each S&P 500 constituent as a function of its market cap divided by the total S&P 500 market cap, and the other that weights each stock as a function of its estimated equity value divided by the sum of intrinsic value estimates for all S&P 500 stocks. To eliminate the use of short positions in this study, we replace negative values with zero.
As shown in the cumulative return chart above, a dollar invested in the market cap weighted index at the beginning of October 1998 appreciated to $6.51 by the end of May 2021. The intrinsic value weighted index appreciated by an additional $1.23, reaching $7.74 over the same investment horizon.
Further examining performance characteristics through a risk-adjusted lens, we observe annualized returns 83 basis points higher for the intrinsic value weighted portfolio, accompanied by a higher Sharpe ratio, an Information Ratio of 0.185, and a Jensen’s alpha of 0.84%. The intrinsic value weighted index also has slightly lower max draw down and beta characteristics, accompanied by slightly higher monthly return volatility compared to the market cap weighted benchmark.
Over the course of its worst season of underperformance, the intrinsic value weighted index lagged the cap weighted alternative by 11.5% (in the late 90s during the tech bubble). Cap weighting, on the other hand, has lagged intrinsic value weighting by more than 120% over the course of this study horizon; this relative draw down is worse than the inverse relationship by more than a factor of 10.
Regressing both of these portfolios against commonly studied asset pricing models, we can confirm that the outperformance of an intrinsic value weighted index compared to a market cap weighted alternative does provide incremental alpha. (+79 basis points against the CAPM, +102 basis points against the Fama French 3 Factor Model, +58 basis points against the Fama French 5 Factor Model, and +64 basis points against the AQR 6 Factor Model) In a six-factor context, an intrinsic value weighted index mildly increases the factor loading towards high book-to-price stocks, high operating profitability stocks, and negative momentum stocks, while market, size, and investment rate betas are immaterially different.
The intrinsic value weighted approach clearly dominates the market cap weighted alternative over our study horizon. Study of returns, risk-adjusted returns, and alpha characteristics do little to support the assumptions embedded in the Efficient Market Hypothesis which claim that market prices are an adequate proxy for intrinsic value.
One final comparison tracks the active share calculation across time between each index. Active share simply measures half of the sum of the absolute differences between portfolio weights and benchmark weights for each constituent. The following chart plots active share trends over time between the intrinsic value weighted index portfolio and market cap weighted benchmark.
A couple of observations from this chart are noteworthy. By breaking the time series into four regimes, we note that active share levels averaged 22.1% between 1998 and 2000, 16.0% between 2001 and 2009, 17.9% between 2010 and 2019, then 22.0% since the start of 2020.
We observe an interesting barometer regarding “valuation agnosticism” in markets when the active share between intrinsic value and market cap weighted portfolios exceeds 20% for extended periods of time. The first observation occurred between September 1999 and February 2001 during the tech bubble. More recently, this spike occurred at the start of April 2020 and is still ongoing, likely driven by the increase in thematic investing around COVID, digitization and reopening trends, further exacerbated by the S&P 500’s inclusion of TSLA in late December 2020.
Shifting our focus away from these active share peaks, we note a structural increase in active share coming out of the financial crisis. Since 2009, the adoption of passive indexing has increased significantly, which may offer insight to the degree in which price discovery has become suppressed by this trend. (16.0% average active share between 2001 and 2009, followed by 17.9% average active share between 2010 and 2019)
From this chart, we note that market cap weighting adequately captures 73-89% of the overall S&P 500’s intrinsic value weights over the last twenty years (not including the unknown error component embedded in our intrinsic value estimates). The remainder represents misallocation of capital that underweights attractive investment opportunities and overweights stocks with unreasonable expectations reflected in their market prices. The degree that active share appears to be variant in different market regimes, while drifting higher with the increase in passive investing over the last decade, provides compelling evidence in addition to the improved risk-adjusted returns and alpha to untether portfolio construction from market prices in favor of a disciplined intrinsic value alternative.
It is worth noting that the active share trend in the Russell 2000 over the same time horizon indicates a stronger upward trend in active share over the last decade, even prior to the early 2020 COVID-related spike noted in the S&P 500. This indicates that the market impact of suppressed price discovery has been much more dramatic in small/mid cap stocks. (Active share averaged 24.3% between 2001 and 2009, then averaged 27.6% between 2010 and 2019 before reaching 30.9% by the end of 2019)
Valuation vs. Cheapness in Value Weighted Indices
Equipped with evidence that investors benefit from an intrinsic value discipline when forming a passive index, we can now separate intrinsic value into its two distinct asset pricing components. Fama French’s three factor research posits that information from a firm’s book equity can be used to derive an additional “value premium” based on the study of historic cross-sectional returns. Applied Finance’s research indicates that the remaining component of intrinsic value not already reflected in book equity captures the present value of future economic profits derived from abnormal earnings and organic reinvestment, which we refer to as Excess Intrinsic Value. We can calculate two additional value weighted index methodologies based on these common stock and Excess Intrinsic Value measures of value.
As noted in our previous study, $1 invested in a market cap weighted index returned $6.51 over our study horizon, compared to $7.74 for an intrinsic value weighted alternative. A common stock weighted index based on each firm’s common stock level on their most recent annual balance sheet only returned $5.73, while the Excess Intrinsic Value weighted index returned $8.60.
The Excess Intrinsic Value weighted index returns 134 basis points per year over the market cap weighted index, with a Sharpe ratio of 0.73, an Information Ratio of 0.31, and a Jensen’s alpha of 1.73%. Excess Intrinsic Value weighting has also offered materially lower draw downs, lower beta characteristics, and lower volatility than the market cap weighted benchmark.
Common stock weighted portfolios, on the other hand, underperform the market cap weighted benchmark by 61 basis points per year with poor risk-adjusted characteristics and higher draw downs, beta and volatility measures. This also highlights the poor proxy that book value provides to intrinsic value, as the common stock weighted portfolio underperforms the intrinsic value weighted portfolio by 144 basis points per year.
Regressing each of these alternative value weighted approaches against commonly studied asset pricing models, we observe notable alpha differences between Excess Intrinsic Value and common stock weights. (The annualized alpha for Excess Intrinsic Value is 296 basis points higher than common stock when regressed against the CAPM, +281 basis points against the Fama French 3 Factor model, +175 basis points against the Fama French 5 Factor model, and +184 basis points against the AQR 6 Factor model).
In a six-factor context, the Excess Intrinsic Value weighted index has a market beta slightly less than 1, a book-to-price (HML) and investment rate (CMA) beta near zero, coupled with increased exposure to high operating profitability (RMW) and negative momentum (UMD) compared to the market cap weighted index. The common stock weighted index, on the other hand, has a market beta slightly higher than 1, with factor tilts towards high book-to-price, low profitability, and conservative growth compared to the market cap weighted index.
We can expand on our earlier active share study by benchmarking the various value weighted alternatives, now against our intrinsic value weighted benchmark. We noted previously that the application of the Efficient Market Hypothesis to justify market value as a proxy for intrinsic value leads to an average active share between each index of 17.8% over our study horizon. It is compelling that the improved risk-adjusted returns and alpha characteristics offered by migrating from intrinsic value to Excess Intrinsic Value are generally accompanied by a smaller level of active share (14.0% on average) than the levels accepted by investors using market cap as an intrinsic value proxy.
It is also noteworthy that the active share levels between common stock and intrinsic value weighting have continued to diverge since 2008. This deteriorating link between common stock and intrinsic value is likely attributable to rising levels of intangible investment and share repurchase activity rendering the information provided by book equity increasingly distorted. Active share averaged 25.8% between 1998 and 2008, but this has since averaged 32.6% from 2009 forward while recently reaching 41.2%.
Enhanced Valuation-Based Indexing with Valuation Stewardship
In the Valuation Beta research that introduced Applied Finance’s Excess Intrinsic Value factor, we also reinterpreted the dividend discount framework used by Fama French to motivate their operating profitability and investment rate factors, which allowed us to address the problematic assumption that high growth is unconditionally a predictor of negative future stock returns. This interpretation motivated the Financing Yield factor, which separates firms returning capital from those reliant on external financing on an enterprise basis; as a parsimonious single factor, it provides unexplained alpha against the Fama French 5 Factor model, while rendering the operating profitability and investment rate factors redundant.
In the context of passive indexing, we can incorporate the Financing Yield factor by simply excluding companies that diluted existing ownership with external financing in their most recent fiscal year, then apply the Excess Intrinsic Value weight to the remaining constituents. We will refer to this approach as “Excess Intrinsic Value + Stewardship” in this study.
In previous charts, we highlighted that an initial $1 invested in October 1998 appreciated to $7.74 with an intrinsic value weighting and $8.60 with an Excess Intrinsic Value weighting. The initial $1 investment in the Excess Intrinsic Value + Stewardship portfolio appreciates to $10.92 over our study horizon.
The performance data above highlights the further improvement to risk-adjusted returns when negative Financing Yield stocks are removed from the portfolio. The stewardship overlay lowers the portfolio count to roughly 300 constituents on average. Removing dilutive stocks that required external financing in their most recent fiscal year adds an additional 117 basis points per year in annual outperformance compared to the Excess Intrinsic Value weighting studied earlier. We also note drastic improvements in the Sharpe ratio, Information Ratio, max draw down, and Jensen’s alpha measures. The beta coefficient falls drastically below 1.0 to 0.874, and the standard deviation of monthly returns at 3.94% is dramatically lower than the market cap weighted alternative level of 4.34%.
Regressing the Excess Intrinsic Value + Stewardship portfolio against commonly studied asset pricing models provides significant alpha characteristics in both the single factor and three factor asset pricing contexts. The unexplained annual alpha over a market cap weighted benchmark is +322 basis points vs. the CAPM, +341 basis points vs. the Fama French 3 Factor model, +152 basis points vs. the Fama French 5 Factor model and +159 basis points against the AQR 6 Factor model. In a six-factor context, this portfolio has a market beta of 0.91, a mild load towards low book-to-price (HML), conservative investment rates (CMA) and negative momentum (UMD) accompanied by a significant load on high operating profitability (RMW) compared to the market cap weighted alternative.
Despite dropping our portfolio count to roughly 300 holdings (forming portfolios on 60% of the overall S&P 500 constituent list), active share levels over the past decade are only marginally higher than the active share assumed when market cap weighting is used as a proxy for intrinsic value. The active share of Excess Intrinsic Value + Stewardship was much higher during the tech bubble when many large cap technology stocks were likely reliant on external financing to fuel rapid growth expectations before falling to levels below 40% by late 2003.
This Excess Intrinsic Value + Stewardship approach to portfolio construction offers vastly improved risk-adjusted performance and alpha characteristics over our study horizon compared to cap weighted benchmarks, as well as the alternative value weighted strategies that allocate to the entire set of index constituents. This approach offers investors comfortable with modest levels of active share the largest strategic advantage against market cap weighted benchmarks.
While Applied Finance has long advocated for investors to consider the strategic advantages from incorporating a valuation-based discipline in portfolio construction and stock selection, this study provides compelling evidence that even passive allocations benefit when index weights are formed on intrinsic value characteristics instead of using market cap as a proxy. While the Efficient Market Hypothesis presumes that market prices are adequate proxies to intrinsic value, we can offer more than twenty years of compelling, out-of-sample evidence that highlights that this presumption misallocates capital and lowers overall market returns by more than 80 basis points annually in the S&P 500.
This theme is especially relevant in the current market environment. Active share levels between market cap and intrinsic value weighted indices have recently spiked due to a high degree of “valuation agnosticism” that has taken hold of equity investors over the previous year, directly impacting market cap weighted passive allocations.
This chart conveys the tactical merit of these active share observations in the current market environment. The market cap weighted index has aggregate intrinsic value upside of -30.0%. The intrinsic value weighted index has aggregate upside materially higher at 1.5%. The common stock weighted index has aggregate upside of -21.9%, while Excess Intrinsic Value has aggregate upside of 10.5%, further improved to 11.3% with the stewardship overlay.
While the two dominant trends in equity allocation over the past decade have emphasized passive and factor-based investing, we find it noteworthy that both value weighted indexing and asset pricing studies lead to a reconciliation of the Valuation Stewardship strategy formed on the Excess Intrinsic Value and Financing Yield factors. Investors comfortable with modest levels of active exposure can apply these valuation-based asset pricing insights to invest in an enhanced valuation-based passive alternative designed to provide a strategic advantage over cap weighted benchmarks.
Applied Finance provides valuation driven investment strategies to help our clients and research partners incorporate an intrinsic value discipline into their equity allocation. Please contact your Applied Finance representative to discuss which investment strategies are most appropriate for you or your clients.