When a comparatively young field like portfolio management is paired with a groundbreaking new technology like blockchain, it is unsurprising that not everyone agrees about what the best way to build a crypto portfolio is. But this leaves practitioners with plenty of room to innovate and find an edge in a nascent and evolving market. Whatever strategy you choose to adopt, extensive and reliable crypto data, encompassing both on-chain and market data, will be the key to executing it.
It’s easy to forget that even in traditional finance, portfolio management is a relatively new concept. While the pioneers of economics like Adam Smith lived in the 1700s and earlier, portfolio management as a formal discipline scarcely existed until Harry Markowitz published his seminal paper on the topic in 1952 which detailed the benefits of portfolio diversification. While most of Markowitz’s methodology has since been replaced with more sophisticated algorithms, he left finance with a valuable insight: that portfolio managers should strive to find the sweet spot between risk and return when building a portfolio and that quantitative analysis of market data is the key to finding it.
What looked so elegant on paper has proven highly complex in reality, with many competing portfolio tracking systems and data analytics models vying for dominance. But what types of crypto data are required to apply these models in practice and what distinct features of cryptocurrency markets need to be taken into account? These questions will be explored in this article.
Most financial professionals can agree on the high-level principles of portfolio management.
The first assumption is that by combining multiple assets into a single portfolio, the overall risk is reduced. While this might seem a statement of the obvious in traditional finance, it is perhaps not as self-evident in a market where a single asset, Bitcoin, commands such attention and market cap. Fortunately, however, research in the field supports the proposition that adding multiple assets into a crypto portfolio tends to reduce risk. There is also broad agreement on the overall objective: to construct a portfolio that maximizes returns at a given level of risk. So far so good. Unfortunately, however, this is usually where the broad consensus ends and the methodological debates begin.
While the classical approach to portfolio management used the standard deviation of stock or asset values as a proxy for risk, this has now been superseded by other metrics. As standard deviation captures both upside and downside outcomes, it is actually a measure of uncertainty, not risk. As a result, many other measures of risk that focus on potential losses rather than gains have been developed, with VaR (Value-at-Risk) being one of the most prevalent.
When analyzing the diversity of a crypto portfolio, managers will look for signs that asset prices are linked using metrics like covariance, which measures the extent to which prices tend to move in the same direction, and correlation, which measures the strength of those relationships.
While there are many different nuances to the calculation of risk and return, almost all approaches — such as the historical method, variance-covariance method, and Monte Carlo simulation — require historical crypto data detailing price movements over a specific review period. But while cryptocurrency prices are one of the basic inputs of any model, not all data on cryptocurrency prices is made equal.
Unlike conventional stocks which are traded on a central exchange like the The New York Stock Exchange (NYSE) or Nasdaq, cryptocurrencies are traded through many centralized and decentralized venues. This spread of liquidity has been shown to give rise to price differences across crypto exchanges, particularly during periods of volatility. In addition, in countries like South Korea that have capital controls in place, cryptocurrency tends to trade at a premium or discount relative to the rest of the global cryptocurrency market.
A further complication is created by less mature digital asset exchanges, which may provide inconsistent crypto data or have unreliable data APIs. Add to this the bogus transactions and washtrading of unscrupulous actors, and it becomes clear why crypto asset pricing is so complex.
As a result, there is no definitive, centralized pricing in digital asset markets and data from multiple different sources needs to be aggregated before separating the signal from noise. Thankfully, there are a growing number of reputable specialists who focus on doing the legwork to create accurate pricing data for centralized market sources and when it comes to transaction activity, you can always consult the underlying blockchain for a truly global view. Thus, portfolio managers should ensure that they have a portfolio tracker in place that provides a broad view of the market, encompassing both market and on-chain data and analytics.
But crypto is not just different in terms of trading venues, it also differs in terms of trading hours. Traditional markets have fixed trading times, typically from 9 am to 4 pm, and transaction activity tends to increase around opening and closing times. Traders tend to attach a great deal of importance to closing prices. Conversely, cryptocurrency can be traded 24 hours a day, 7 days a week. Thus, while crypto data providers provide a “close price” for digital assets, this is not exactly the same as its namesake in traditional markets and there is some debate about the implications for crypto portfolio management.
When conducting research on assets for potential inclusion in a crypto portfolio, or when setting up your data analytics or portfolio tracker, it is worth considering whether close price is an appropriate metric for your needs or whether a volume weighted average price over a different time span might be more appropriate.
When building and optimizing a crypto portfolio, price data doesn’t tell the full story. It is also necessary to understand the underlying tokenomics of an asset. At a basic level, it is important to know the size of the circulating supply to ascertain whether there is sufficient liquidity to support your crypto portfolio strategy. In addition, any processes that could affect future token supply like staking processes, vesting periods or lockups should also be considered. You should not have to do an internet search for every asset to check this type of tokenomic information. When setting up data analytics, make sure that such data is integrated directly into the frontend of your portfolio tracker and readily available at a glance with one click.
It can also be useful to diversify exposure across distinct use cases or sectors. Recent research into crypto portfolio optimization suggests that maintaining a balance between sectors like finance, exchanges, and business services, can increase diversity and reduce risk. To save yourself the time of coding sectoral categories manually, look for data providers and portfolio tracking tools that enable multi-level filtering of assets by sector.
Finally, in a crypto market where retail investors still play a considerable role, social media and online content have an undeniable influence. An emerging approach to portfolio optimization harnesses publicly available data from sources like Twitter and Google Trends to conduct sentiment analysis in an effort to predict patterns of market behavior. While this is a nascent field and may become less significant in the long run as institutional involvement in cryptocurrency markets increases, it is currently an area ripe for innovation.
A new market needs new data analytics tools. Nuant is a new, specialized portfolio management, analytics and data intelligence platform tailored to the needs of digital asset managers and crypto analysts. It allows all exchange accounts and wallets associated with your portfolios to be unified into one, global data analytics dashboard and portfolio tracker. Gain insights from both on-chain and market data providers and perform due diligence on wallets and tokens within a few clicks. Schedule a demo with us now by clicking the button on the top right to become one of our early users.
Bloomberg Professional Services (2022) ‘Crypto never sleeps - The challenges of data aggregation, pricing and risk’.
Brauneis, A. and Mestel, R. (2018) ‘Cryptocurrency-Portfolios in a Mean-Variance Framework’, SSRN Electronic Journal [Preprint].
Cifuentes, A. (2022) ‘Timeless relevance of Markowitz’s ideas for finance and investing’, Financial Times, 18 March
Čuljak, M., Tomić, B. and Žiković, S. (2022) ‘Benefits of sectoral cryptocurrency portfolio optimization’, Research in International Business and Finance, 60, p. 101615.
Makarov, I. and Schoar, A. (2020) ‘Trading and arbitrage in cryptocurrency markets’, Journal of Financial Economics, 135(2), pp. 293–319.
Markowitz, H. (1952) ‘Portfolio Selection’, The Journal of Finance, 7(1), p. 77.
Vidal-Tomás, D. (2021) ‘An investigation of cryptocurrency data: the market that never sleeps’, Quantitative Finance, 21(12), pp. 2007–2024.
Zhou, Z. et al. (2023) ‘Multi-source data driven cryptocurrency price movement prediction and portfolio optimization’, Expert Systems with Applications, 219, p. 119600.
