Comments Off on An Exploration of Cryptocurrency Investing, Modern Portfolio Theory, and Portfolio Construction Fintech news

An Exploration of Cryptocurrency Investing, Modern Portfolio Theory, and Portfolio Construction

By Phil Glazer @MGV. Previously @KKR_Co & @UCBerkeley |

This piece provides an exploration of the role cryptocurrencies may play in a larger portfolio of assets (as always, this piece is not advice nor a recommendation for investment) including:

  • Modern Portfolio Theory (MPT)
  • The rationale for diversification within equity and bond investing
  • Diversification applied to cryptocurrency investing
  • Exploration of the role of cryptocurrency exposure in a portfolio

Modern Portfolio Theory (MPT)

Modern Portfolio Theory (MPT) provides a mathematical framework to build a portfolio of assets that maximizes expected return for a given level of risk taken. This creates an “efficient frontier”, where expected return can only be increased by also increasing risk level taken (shown below).

This shows the trade-offs of holding different allocation weights of stocks vs. bonds. Combinations that are above and to the right of the “75% Bonds, 25% Stocks” allocation are on the efficient frontier. (Image Source)

The example shown demonstrates the benefits of including a mix of equity and bond exposure in different proportions. As shown, holding only bonds is not risk/reward efficient and it is possible to increase equity exposure to achieve higher returns for a given risk level. This same model can be extended to include other asset classes to expand a portfolio beyond equities and bonds to hold other investments, like real estate and gold.

The key takeaway from the model is that it is advantageous to hold investments that have a positive expected value but are not perfectly correlated to each other (in fact, the lower the correlation between the investments, the better). Details about model inputs and derivation can be found here.

Rationale for Diversification within Equity/Bond Investing

At the portfolio level, it makes sense to hold exposure to different asset classes. Similarly, it makes sense to diversify within an asset class. Specific to equity and bond investing, companies face two types of risk: systemic risk and idiosyncratic risk.

Systemic risk (also known as market risk) refers to risks faced by all companies in a market or a market segment, like interest rate movements or political instability.

Idiosyncratic risk (also known as specific risk) refers to risks unique to a given company, like a decline in consumer interest for hamburgers or company mismanagement.

For equity investors, there is significant debate over what the appropriate number of stocks to hold is to be “diversified” (opinions differ and range from 15–20 stocks up to 100+, or an index that represents the entire market), but all converge on the idea that holding many different stocks is beneficial. Indeed, it is possible to create a more desirable risk/reward profile by holding many stocks with different qualities (geography, industry, market cap, etc.) to minimize the downside scenarios created by idiosyncratic risk while still enjoying the broader benefits of long-term business growth and profit.

Diversification within Cryptocurrency Investing

While diversifying within an asset class and constructing a portfolio of multiple assets have both been thoroughly studied, very little has been written about how these concepts might apply to cryptocurrencies.

Data exists for the historical prices of various cryptocurrencies. However, it is likely a mistake to use this pricing information to attempt to approximate correlations, variance bounds, and other related metrics to use as inputs in a model. Fragmented exchanges with low trading volume and a retail dominated market leads to significant noise and little signal to base models on that rely on getting longer-term trends right (additionally, historical data is very limited and restricted to a multi-year run up environment that may be change in the future). Using this historical data would mistake precision for accuracy and be misleading. Instead, it might be more useful to make assumptions and think about the nature of the risks qualitatively.

Diversification within cryptocurrency investing is more nuanced than equities or bonds because, for now, there are significant unknowns. One of the key unknowns is which token models will actually accrue value and appreciate in price over the longer-term. Outside of value accrual, there is significant uncertainty surrounding the nature of risks taken and how they might differ between coins, like:

  • Is having a large percentage of actively traded supply take place on one or only a few exchanges ($150 million worth of NANO was taken from BitGrail, where much of the supply traded, in a recent hack) safe? How much volume across how many exchanges is needed to be comfortable?
  • Will certain types of utility tokens be deemed securities and, therefore, be limited in their distribution (at least until regulation changes on this, if it does)? What would this mean for price/adoption?
  • What is the long-term value for a utility token that is used to do something like be a medium of exchange for file storage? Eventually will this native token not be needed and be phased out by interoperability? If so, over what time frame and when will the market adjust to reflect this?
  • Is the store-of-value use case the only form of cryptocurrency that will meaningfully appreciate in price and continue to do so in the future? If so, does it make sense to only diversify across projects attempting to solve for this use case? If not, which other forms should be allocated to and in what proportion to other use cases?

Theoretically, if there are multiple coins that have an estimated positive expected value and the risks involved in each are idiosyncratic in nature, then diversifying investment across multiple projects with unique risks would lead to a more desirable risk/reward profile. Additionally, as the nature of these risks changes over time it would be prudent to reevaluate and rebalance accordingly. Ultimately, diversifying within cryptocurrency investing will come down to how an investor views the market, especially which use cases and coins have the potential to accrue value over the long-term and what risks these coins face.

Does Cryptocurrency Exposure Belong in a Portfolio?

Reminder: this piece is not investment advice nor recommendation.

Whether or not cryptocurrencies belong in a portfolio will be determined by the outlook a given investor has on the market and, in particular, the value accrual theses he/she believes in as well as which idiosyncratic risks specific projects face. The assumptions that need to be considered for diversifying cryptocurrency positions was explored in the prior section. This section will suggest assumptions that need to be made for justifying cryptocurrency allocation within a portfolio, like:

  • What is the long-term correlation between cryptocurrencies and equities and bonds (or other assets held in a portfolio)?
  • Will the long-term success of cryptocurrencies change the way that public companies perform? Put another way: will cryptocurrency adoption lead to certain industries performing better or worse? This may be relevant for sizing exposure and understanding correlation before sufficient data can explain it quantitatively. For example (in the future), will cheap, massively distributed file storage change industries?
  • In assessing the risk of a portfolio, is it expected that the performance of certain groups of cryptocurrencies (like store-of-value) will be negatively correlated to equities in a bear market for the broader economy?
  • How will interest rate movements impact the value of cryptocurrencies?

The inclusion of cryptocurrencies within a larger portfolio of assets will come down to an investor’s estimation of expected value and the relationship between cryptocurrencies and other asset classes. If cryptocurrencies have positive expected value and low correlation to other asset classes over the long-run, then they may play a role in constructing a portfolio with superior risk-adjusted returns. For large institutions that manage $100+ billion (there are 10 that each manage more than $1.5 trillion) answering these questions will be critical before getting involved in the space. Improving returns or reducing risk by a fraction of a percent makes a meaningful difference at scale.

Conclusion + Open Questions

This piece concludes with far more questions than answers. There are many further points of interest on this topic that will be explored over the next few months, from the practical (trading volume distribution across exchanges) to the theoretical (long-run idiosyncratic risks). Stay tuned! It will be fascinating to see how the space develops.

If you’d like to chat about cryptocurrencies or anything else technology related Tweet at me here (DMs are open, too). You can also email me at

Comments are closed.