Intangible assets have become an important and fast-growing part of firms' capital stocks. Estimates of the proportion of intangible capital to overall corporate capital has increased from around one-third in the early 2000s (Corrado et al. 2009) to more than half in recent years (Eisfeldt and Papanikolaou 2013, Falato et al. 2020, McGrattan 2020, Belo et al. 2020, Ewens et al. 2020). However, since intangible assets are created through investment in employee, brand, and knowledge capital that are expensed, most of them do not appear in corporate balance sheets. This omission has resulted in the growing mismeasurement of firms’ book value of assets.

An important implication of this mismeasurement is the deterioration of book assets’ role as the fundamental anchor in firm valuation. Value investing strategies, such as that developed in the seminal work by Fama and French (1992, 1993), typically compare market capitalisation to book assets to identify undervalued stocks. This strategy has performed very poorly in recent decades, exhibiting negative cumulative returns post-Great Recession. The decline in performance has led many academics and practitioners to declare the value strategy dead.

In a recent paper (Eisfeldt et al. 2020), we argue that the value measure is simply in need of an update to incorporate intangible assets. We show that including intangibles to book assets completely reverses value’s poor performance while retaining the ability to explain the cross-section of stock returns. We also provide corroborating evidence that this adjustment leads to an investment strategy that buys firms exhibiting high productivity and with higher alternatives valuation measures such as sales or earnings relative to stock price. Lastly, we take stock of known accounting discrepancies across industries and show that our results are robust to alternative methodologies for calculating intangible assets.

**Figure 1 **Monthly cumulative returns of traditional value versus intangible value

*Note*: Each panel shows the returns of one dollar invested in each value factor at the beginning of years 1975 (full sample), 1995 (Internet Bubble), and 2007 (Great Recession), respectively.*Source*: Eisfeldt et al. (2020).

## The intangible value factor

In the US, intangible assets created internally are expensed and typically do not appear explicitly on the balance sheet. This means that the replacement cost of internally generated intangible capital must be calculated based on past investments. Following methodology first introduced in Eisfeldt and Papanikolaou (2013), we calculate intangible capital by accumulating past selling, general, and administrative (SG&A) expenses using the perpetual inventory method. While other works divide intangible capital further into knowledge capital and organisation capital and accumulate them separately (Falato et al. 2020, Peters and Taylor 2017, Crouzet and Eberly 2019, Park 2020, Arnott et al. 2021), we opt for the simpler measure because it requires the fewest assumptions. We add the resulting intangible capital to book assets and construct the book-to-market ratio (B/M ratio).

The Fama and French (1993) value factor (‘traditional value’) takes a long position in a portfolio of value stocks (high B/M ratio) and a short position in a portfolio of growth stocks (low B/M ratio). We follow this methodology to construct a value factor using the intangibles-adjusted B/M ratio (‘intangible value’).^{1} Our intangible value factor is highly correlated with the traditional value factor, exhibiting a correlation of 81%. Since the difficulty in measuring intangibles arises from industry differences in accounting practices (Koh and Reeb 2015, Enache and Srivastava 2018), we additionally construct an intangible value factor that sorts firms at the industry level.

## Returns and pricing performance

Between 1975 and 2018, we find that intangible value significantly outperforms traditional value. A regression of intangible value returns on traditional value returns for this period results in a statistically significant alpha of 3.32%. We also find that the outperformance is more pronounced in the post-crisis era during which the returns to traditional value have been particularly disappointing. In additional tests, we further isolate the effects of intangibles by constructing a factor that sorts firms on intangible assets as a fraction of market equity, as well as a factor that only includes firms that are uniquely in the long or short leg of intangible value. Compared to the baseline intangible value, these factors exhibit more independent variation and higher full-sample alphas relative to traditional value.

We also examine the ability of the traditional and intangible value factors to explain the cross-section of stock returns. As shown in Figure 2, replacing traditional value with intangible value in the Fama and French three-factor model with momentum reduces mean absolute pricing errors by 5% (Panels A and B), while the same substitution in a five-factor model with momentum reduces the errors by 8% (Panels C and D). In robustness tests, we find that the pricing performance of versions of intangible value that better isolate the effects of intangibles is at least as good as that of traditional value.

**Figure 2** Cross-sectional asset pricing tests for three-factor and five-factor models plus momentum

*Note*: The top row plots realised mean excess returns of 25 size and book-to-market-sorted portfolios and ten momentum portfolios against the mean excess returns predicted by the Fama and French three factor plus momentum model, where Panel B replaces traditional value with intangible value. The bottom row plots realised mean excess returns of 25 size and book-to-market-sorted portfolios, ten momentum portfolios, ten portfolios sorted on operating profitability, and ten portfolios sorted on investment, against the mean excess returns predicted by the Fama and French five factor plus momentum model. The sample is monthly from 1975 to 2018. *Source*: Eisfeldt et al. (2020).

## Firm characteristics

On average, the long leg of traditional value has a lower intangible-adjusted B/M ratio than the short leg. This means that traditional value may be buying ‘value traps’ that appear attractive because B/M is not properly measured. To extend this analysis, we examine the average characteristics of firms sorted in each leg of both value factors. We indeed find that the intangible value factor better identifies firms with superior fundamentals such as productivity, profitability, and financial soundness.

## Conclusion

The traditional strategy of value investing, which relies on using book assets as the fundamental anchor of firm value, has lost its edge in recent years. We argue that this trend is likely due to the omission of intangible capital, whose importance has significantly increased in recent decades. A value factor that incorporates intangible capital provides much higher returns throughout all periods and explains the cross-section of stock returns with similar or lower pricing errors. Our results are consistent with contemporaneous work (Park 2020, Lev and Srivastava 2020, Arnott et al. 2021). We make additional contributions by using a measurement strategy that minimises assumptions, providing results for portfolios that better isolate the effects of intangibles, and showing how traditional value fails to identify true value firms in modern times.

Our findings imply that asset pricing researchers should consider using an intangible value factor when studying the value effect. Practitioners can also incorporate intangibles to implement a profitable relative value strategy, especially in recent years when traditional value has underperformed.

## References

Arnott, R, C Harvey, V Kalesnik and J Linnainmaa (2021), “Reports of Value’s Death May Be Greatly Exaggerated”, *Financial Analysts Journal *77(1): 44-67.

Belo, F, V Gala, J Salomao and M A Vitorino (2020), “Decomposing Firm Value”, Working Paper.

Cecchetti, S and K Schoenholtz (2018), “Financing Intangible Capital”, VoxEU.org, 22 February.

Corrado, C, C Hulten and D Sichel (2009), “Intangible Capital and US Economic Growth”, *Review of Income and Wealth* 55(3): 661-685.

Crouzet, N and J Eberly (2019), “Understanding Weak Capital Investment: the Role of Market Concentration and Intangibles”, *Proceedings of the 2018 Jackson Hole symposium*, 87-148.

Eisfeldt, A L and D Papanikolaou (2013), “Organization Capital and the Cross-Section of Expected Returns”, *The Journal of Finance *68(4): 1365-1406.

Eisfeldt, A L, E Kim and D Papanikolaou (2020), “Intangible Value”, NBER Working Paper 28056.

Enache, L and A Srivastava (2018), “Should Intangible Investments Be Reported Separately or Commingled with Operating Expenses? New Evidence”, *Management Science* 64(7): 3446-3468.

Ewens, M, R H Peters and S Wang (2020), “Measuring Intangible Capital with Market Prices”, NBER Working Paper 25960.

Falato, A, D Kadyrzhanova and J Sim (2020), “Rising Intangible Capital, Shrinking Debt Capacity, and the US Corporate Savings Glut”, *Journal of Finance*, forthcoming.

Fama, E F and K R French (1992), “The Cross-Section of Expected Stock Returns”, *The Journal of Finance *47(2): 427-465.

Fama, E F and K R French (1993), “Common risk factors in the returns on stocks and bonds”, *Journal of Financial Economics* 33(1): 3-56.

Haskel, J and S Westlake (2017), *Capitalism without Capital: The Rise of the Intangible Economy*, Princeton University Press.

Koh, P-S and D M Reeb (2015), “Missing R&D”, *Journal of Accounting and Economics 60(1): 73-94.*

Lev, B and A Srivastava (2020), “Exploring the Recent Failure of Value Investing”, NYU Stern School of Business Research Paper Series.

McGrattan, E R (2020), “Intangible Capital and Measured Productivity”, *Review of Economic Dynamics *37(1): S147-S166.

Park, H (2020), “An Intangible-adjusted Book-to-market Ratio Still Predicts Stock Returns”, *Critical Finance Review*, forthcoming.

Peters, R H and L A Taylor (2017), “Intangible Capital and the Investment-q Relation”, *Journal of Financial Economics *123(2): 251-372.

## Endnotes

1 See https://github.com/edwardtkim/intangiblevalue to download the data and code.