Cryptocurrency investors are bombarded by a variety of investment strategies from a growing and increasingly complex asset class. In this article we present benchmark indices for four different investment styles in cryptocurrency markets that can be tracked at www.protosterminal.com/indices. These benchmark indices — value, trend-following, low volatility and small cap — are derived from the risk premia discussion in traditional finance and can be used to evaluate the risk in investment strategies. The benchmark indices have shown unique performances during the bull market of 2017 as well as during the recent 2018 Q1 downturn and may prove helpful to investors as comparables to their own portfolios (see Figure 1).
Figure 1: Performance of the market and bitcoin index versus the performance of four different investment styles during the recent cryptocurrency up and down cycle.
Benefits of having benchmark indices
Cryptocurrencies can be considered as a new asset class, but also differ widely in their application, market capitalization and distribution for example. According to www.coinmarketcap.com over 1500 different cryptocurrencies exist today and it becomes increasingly difficult to keep track of them.
In this article, we will propose several benchmarks indices to make sense of the performance in this new asset class. In particular, we will introduce concepts from traditional finance research and suggest applications to cryptocurrencies. The aim of these benchmark indices is to represent different investment styles that are as different as possible and can be used to understand investment performances.
The idea behind investment styles is backed by solid academic research, aiming at above-average risk-adjusted returns. In 1952, Harry Markowitz’s seminal paper “Portfolio Selection” marked the industry’s academic starting point and introduced the concept of diversification. This concept was broadened after the stock-market crash in 1987 and the bursting of the “IT bubble” to include various developed and emerging markets.
However, portfolio diversification did not protect investors from the financial crisis of 2008 and they started looking for alternative ways to improve risk-reward ratios. Until then, these investment styles were only familiar to the most sophisticated investors, like well-known hedge funds.
In 1997, William Fung and David Hsieh put forward the idea that risk and return characteristics can be structured dependent on trading strategies rather than on the diversification of asset classes. They showed that hedge funds follow few dominant systematic trading strategies such as value investing and trend-following.
Our indices explained
Value investing is probably the best-known style (see for example Asness, Ilmanen, Israel and Moskowitz, 2016). The implementation of the value style can be straightforward. Take a set of assets and sort them by some measure of fundamental value to price. Buy the assets that have high fundamental value to price (“cheap” assets) and sell the ones that have low fundamental value to price (“expensive” assets).
Another famous strategy is trend-following. Kallerhoff et al. (2016) and Hurst and et al. (2013) showed that the returns of very large hedge funds can be explained by simple trend-following strategies, specifically time series momentum strategies. Time series momentum is a simple trend-following strategy that goes long a market when it has experienced a positive excess return over a certain look-back horizon, and goes short otherwise.
A large driver of performance is also the market capitalization (Fama and French, 1992). Small caps have several advantages that large caps can’t match. That’s because smaller companies usually have less visibility in the investment community and therefore often experience a disconnect between their prices and their fundamentals. This discrepancy between price and fundamentals presents a tremendous opportunity that small cap investors can take advantage of.
Finally, a more recent investment style is to follow low volatility. Low-volatility strategies follow the observation that portfolios of low-volatility stocks have higher risk-adjusted returns than portfolios with high-volatility stocks in most markets studied (Frazzini and Pedersen, 2014). An explanation for this effect is that many investors are constrained in the leverage that they can take, and they therefore overweight risky securities instead of using leverage. Hence, they are paying a premium for high volatility stocks, that can be extracted from the market.
How are the indices calculated?
In order to implement these strategies for cryptocurrencies, we have collected large amount of historic data such as prices, trading volume on exchanges, transaction volume on blockchains, news, active addresses, hashrates, token supply, github commits and others. This data can be monitored daily at Protos Terminal (www.protosterminal.com). From this data we are calculating benchmark indices of the different investment styles — value, trend-following, small cap and low volatility.
The value index attempts to capture returns that arise from investing in undervalued assets. It is calculated based on the Network-to-transaction volume (NVT) that we discussed here. This measure was originally suggested by Willy Woo as an estimate for the price-to-earnings ratio. It is derived from the fact that networks can be modelled similarly to Software-as-a-Service (SaaS) companies, which are traditionally not measured by their earnings but their transaction volumes.
The trend index attempts to capture returns that arise from momentum in the price of an asset. It is calculated based on the price returns using time series momentum as suggested for traditional asset classes by Hurst et al. (2013) and is long-only. We discussed this at length here.
Small Cap Index
The small capitalization index attempts to capture returns that are unique to smaller sized assets. It is is calculated based on the market capitalization of cryptocurrencies, taking an overweight position in smaller cryptocurrencies as suggested for traditional asset classes by Fama and French (1992).
Low Vol Index
The low volatility index attempts to capture returns that are unique to assets with the lowest volatility across its peers. It is calculated based on volatility of cryptocurrencies as suggested by Frazzini and Pedersen (2014) for traditional asset classes .
The figure above shows the performance of the different benchmark indices. The benchmark indices implement each investment style systematically as described above. These indices are compared against an index only investing in bitcoin and investing in a market cap weighted index (Market Index).
All indices outperformed the market and bitcoin index during the recent up and down cycle. Most notably the trend-following index stepped out of the market during the recent downturn and has only recently started selecting cryptocurrencies again. The table below shows the return, volatility, sharpe ratio and beta for each of the benchmark indices. The sharpe ratio compares the risk and return of each index. The beta is calculated as the correlation of each benchmark index versus the market index.
The trend-following strategy has both the highest Sharpe ratio, the lowest volatility and the lowest beta. Hence, it can clearly be seen as superior to the other investment styles. However, the trend-following strategy also differs in that it is the only strategy that can step out of the market. The other investment styles — small cap, low vol and value — are always fully invested. Hence, these investment styles followed the market cycle to a large extend.
All benchmark indices have a correlation to the market above 0.70 since they are only implemented as long only and do not sell the opposite selection short.
In this article we have introduced four different benchmark indices for cryptocurrencies — value, trend-following, low volatility and small cap — that are derived from the risk premia discussion in traditional finance and can be used to evaluate the risk in investment strategies:
- While cryptocurrencies are strongly correlated, it is possible to make out clusters that show unique returns like low volatility, small cap, trend or value.
- Trend following may offer beneficial diversification for portfolios, with comparatively low beta and high sharpe.
- While volatilities are generally extremely high across cryptocurrencies — the low volatility index appears to capture unique asset characteristics that result in a high sharpe by reducing volatility.
- While small caps have shown the largest absolute returns, they also come with particularly high volatility.
Of course the implications of our backtests are limited due to the short testing period of this new asset class.
All of these models require more work to refine. Please reach out.
If you like what you read and like to stay up to date with future postings — be sure to press clap and follow.