Asset Pricing and Investment Styles in Digital Assets

Frontmatter

Chapter 1. The Emergence of a New Asset Class: Digital Assets

Abstract

Cryptocurrencies are a relatively young and emerging phenomenon within the financial world. The use cases of such virtual tokens are manifold, ranging from technical to legal to financial applications. In the years 2016 and 2017, for example, start-ups discovered initial coin offerings (ICOs) to raise funds, circumventing traditional venture capitalists. Since the number of possible applications is vast for cryptocurrencies, I prefer the term digital assets as a broader conception of the whole asset class.

Several groups of digital assets can be distinguished:

• “Traditional” cryptocurrencies such as bitcoin (BTC) or ethereum (ETH) that are mediums of (value) exchange and are connected to their own blockchain.
• Other virtual coins, such as ERC-20 tokens, are situated on such a blockchain and give the issuer of those tokens the possibility to operate on the underlying blockchain. ERC-20 tokens, however, are mainly used for capital raising purposes and could therefore, also be regarded as some kind of “equity” token.
• Some virtual coins serve as the so-called “utility tokens” that allow their holder to access or use a specific service or platform connected to the token.
• Furthermore, it is possible to link the value of a digital asset to a specific underlying index, e.g., the Standard and Poors (S&P) 500 or the USD. Hence, this group could be called “commodity tokens” or “stable coins.”

Tobias Glas

Chapter 2. An Asset Pricer’s Toolkit

Abstract

The methods used in today’s asset pricing and investment styles literature mostly developed during the twentieth century. These quantitative approaches can also be applied to nearly any asset class. Initially, most of the methods were intended to be used for US stock data. Then, the scope was extended to international stock markets and other asset classes such as foreign exchange (FX), commodities, or bonds. The latest asset class to which the same methods can be applied are digital assets. Here and in other asset classes, (almost) no theoretical adjustments need to be made to the original concept. This chapter is therefore not necessarily related to digital assets specifically but provides a methodological background for the analyses used in the remainder of this book.

Tobias Glas

Chapter 3. Investment Styles in General

Abstract

The before mentioned models and methods have been used extensively in the asset pricing and investment styles literature. By doing so, research has found that some characteristics achieve similar good results (i.e., positive excess returns, low correlations to long-only strategies) across different asset classes. If those characteristics are robust over asset classes such as equities, fixed income, commodities, and others, it could be assumed they can be applied to digital assets in the same manner. Since it is not yet clear if digital assets can be considered as their own asset class, characteristics in popular asset classes are described to draw connections in subsequent chapters.

Tobias Glas

Chapter 4. Related Literature

Abstract

Due to the short (data) history of digital assets, research in this asset class is still in its early stages. Meanwhile, many different peer-reviewed works have been published, which mostly investigate various different topics in only a handful of virtual assets. The full cross-section of digital assets is only investigated sparsely. Nevertheless, some context regarding this young stream of the finance literature is provided, especially with regard to asset pricing and investment styles.

Tobias Glas

Chapter 5. Digital Assets

Abstract

It is of importance to understand that digital assets do not carry any intrinsic value such as stocks or bonds. Digital assets are more comparable to gold, for example. Gold has no intrinsic value either, and its use cases are limited. However, gold serves as a haven asset in financial markets and has been attracting mankind for ages. Digital assets, on the other hand, do not even exist in physical form. Coins such as bitcoin or ethereum can only be created (i.e., “mined”) and stored in virtual form as computer code. After some technical introductions, the digital asset dataset(s) is described in the form of stylized facts.

Tobias Glas

Chapter 6. Digital Assets in a Multi-asset Context

Abstract

Digital assets are a relatively heterogeneous asset class with widely dispersed returns. This could lead to increased diversification when allocating digital assets to a portfolio of traditional assets. Therefore, we need to consider whether digital assets add value when compared to traditional asset classes. A similar issue was of importance for the hedge fund literature around the turn of the millennium. To investigate possible benefits, when adding an (new) asset (class) to a reference portfolio of other asset(s) (classes), Huberman and Kandel (J Financ 42(4):873–888, 1987) proposed a “mean-variance spanning” test that is based on the seminal work of Markowitz (J Financ 7(1):77–91, 1952). This test answers the question if it is beneficial to add digital assets to a portfolio consisting of traditional asset classes.

Tobias Glas

Chapter 7. On the Effectiveness of Technical Indicators in Digital Assets

Abstract

A controversial approach in the finance literature is the application of technical indicators or technical analysis in general. Practitioners rely heavily on such tools, but research has not yet agreed on a final conclusion regarding their effectiveness. Technical analysis makes predictions of future prices of an asset based on the past performance information. Therefore, it can be regarded a time-series instrument. On the other hand, with hundreds of investable digital assets, it is still difficult to obtain two things. It is almost impossible to determine (i) the “fair” value of a digital asset, and consequently (ii) an asset selection based on fundamental data (compared to equities) is not applicable. As such, it is even more important to generate buy and sell decisions through other methods.

Generally, technical indicators are used for exactly such decisions. Depending on the indicator’s specific value, buy and sell decisions can be derived and tested in digital assets.

Tobias Glas

Chapter 8. Investment Styles in Digital Assets

Abstract

In the early days of the asset pricing literature, researchers applied portfolio sorts to explore the cross-section of (equity) returns. In the 1970s, this approach was used to investigate whether the predictions of the CAPM hold in empirical tests. Since the ex ante beta of an asset is not observable, and data issues make tests difficult for single assets, portfolio sorts combine multiple assets into a portfolio of assets to mitigate such issues. Thereby, the number of test assets is reduced significantly, but data quality increases tremendously. One outcome of these tests is the so-called low-beta anomaly (Black et al., Studies in the theory of capital markets, 1st edn, Praeger, New York), which proclaims the same or even higher average returns for low-beta stocks when compared to high-beta stocks. Motivated by their work, such effects are investigated in the returns of digital assets as well.

Tobias Glas

Chapter 9. Risk Factors in Digital Assets

Abstract

Following on from the previous chapter, the search for risk factors in digital assets is extended by several means. First, statistical tools such as principal component analysis are used to condense the information contained in the digital asset dataset. After investigating the principal components more deeply, it emerges that this first approach yields no statistically significant results. Therefore, as a second step fundamental as well as other data is considered to intensify the search for risk factors. As that approach still results in rather disappointing outcomes, I turn to the characteristics versus covariances story as a last step. By doing so, three characteristics that tend to explain digital asset returns are identified while being priced significantly in the underlying data.

Tobias Glas

Chapter 10. How to Incorporate Characteristics into the Portfolio Construction Process

Abstract

Recent studies (see Glas (J Altern Invest 22(1):96–113, 2019) among others, and the previous chapters of this book) have uncovered the first indications of characteristics which are able to describe and explain the cross-section of digital asset returns. The standard procedure in the asset pricing and investment styles literature to harvest such characteristic premia is to construct hedge portfolios based on a specific characteristic. In digital assets, however, short selling is almost impossible. Therefore, alternative (long-only) approaches that incorporate characteristics into the portfolio construction process are described for digital asset investors.

Tobias Glas

Chapter 11. Are the Results Robust and Still Valid?

Abstract

Between the original time of writing of this book and its publication, the results appeared to be a bit “out of date.” Therefore, I obtained another dataset, totally independent from the initial dataset. The intention with this second dataset was to prove that the described results are correct, since it includes a much larger set of digital assets than the previous dataset. It is also longer (in terms of data history) and does not suffer from a survivorship bias. Since the second source of data is not connected to the first one at all, it serves as an additional (data) robustness check.

Tobias Glas

Chapter 12. Conclusion

Abstract

This chapter empirically has explored several issues regarding asset pricing and investment styles in the young and emerging asset class of digital assets. For comparability reasons the results of the digital asset outcomes are set against more traditional asset classes such as equities, bonds, and others. Thereby, I have investigated several well-known and popular topics from the investment and asset pricing literature and applied them to digital assets. One main finding is that digital assets are still relatively heterogenous and shaped by the smallest digital coins.

Tobias Glas

Backmatter