Introducing a Crypto Benchmark Framework
TL;DR. The purpose of this piece is to introduce a rough draft framework for benchmarking in the cryptoasset industry. According to Vision Hill Research, crypto hedge fund managers outperformed bitcoin by 560 bps during the second quarter of 2018. Vision Hill Research found the median quarterly return of its in-scope crypto hedge funds (35 data points used for this analysis) to be -4.3% during Q2’18, whereas bitcoin’s quarterly return amounted to -9.9% over the same time period. We plan to provide these reference points on a quarterly basis going forward segmented by fund strategy as well as fund type. As will be discussed herein, we do not believe we have a perfect solution for benchmarking, and are not trying to present this piece as such. Rather, we are proposing a framework, or starting point, and a call to action to the community to help us work on improving this framework so that it may one day become more widely adopted for benchmarking purposes by fund managers and institutional investors. We also discuss what attributes we believe a good benchmark should possess, as well as what challenges we are aware may arise over time in regard to benchmarking in this nascent asset class, such as self-reporting bias, survivorship bias, backfill bias, style drift, and more. Let’s dig in.
Oftentimes, when people refer to “crypto”, we find they are generally referring to the public side of the market; they are referring to bitcoin, ethereum and other publicly-traded, liquid, listed cryptoassets. What isn’t generally picked up in such discussions, is the fact there is a large (and growing) private side of the crypto market, and many different kinds of investment strategies that are underway in addition to the pioneering buy-and- hold strategies. What we seek to deliver in this piece is a framework, or taxonomy rather, that we believe can help the community organize the cryptoasset market a bit more clearly. We will separate the cryptoasset industry into three main market segments and discuss what appear to be common investment strategies in such segments. We will then discuss more comprehensively, how we think investors can properly benchmark their performance against these strategies.
Generally speaking, we believe the cryptoasset industry can be separated into three main segments:
· The public crypto markets;
· The private crypto markets; and
· Crypto-enabled markets.
With respect to the public crypto markets, as noted earlier, we are referring to the public side of the market; we are referring to bitcoin, ethereum and other publicly-traded, liquid, listed cryptoassets. As of the time of this publication, the total current network value of publicly listed cryptoassets is ~$200B, according to CoinMarketCap. On the private side, we are seeing capital formation around private token projects that have not undergone public sales: such examples include Filecoin, Blockstack, DFINITY and so on. Rather, similar to what is seen in the traditional venture capital markets, we are seeing investors participate in simple agreements for future equity and/or tokens (“SAFEs” and “SAFTs”, respectively), or in certain cases, Series A fundraising rounds, Series B, Series C and so on. At the time of this writing, we believe this private token market to have an aggregated network value in the approximate $10-$15 billion range. Lastly, we believe there are crypto-enabled markets that either bootstrap existing business models and tailor them toward crypto, or create new businesses in crypto. An example of something new would be a mining strategy or a staking strategy, two novel models enabled by the process of validating transactions on a network. Examples of something existing applied in a new way would be a traditional lending credit strategy utilizing crypto as a new form of secured collateral, or a research-oriented news platform, such as CoinDesk that has attributes of a legacy business model tailored toward this new asset class.
Now that we’ve addressed market segmentation, we can turn to discussing investment strategies. There are many “crypto funds” that currently exist, so we believe it is important to go a layer deeper here to distinguish them by strategy. An infographic of what we believe to be the most popular investment strategies in the cryptoasset ecosystem at present is illustrated below:
Fundamental Long Only — these are managers that employ extensive research and experiment with quantitative and qualitative metrics to determine what cryptoassets have the potential to accrue large amounts of value over time. These managers find assets they have high conviction in, and buy and hold those assets for a certain period of time. This is popular in both liquid and illiquid strategies (e.g., both public and private markets). We expect to see this strategy expand to a sub-sector focus (e.g, dApps, NFTs, DAOs, TCRs, uncensorable value transfer, and more).
Fundamental Long/Short — these managers similarly undergo extensive research in the space, but instead of only buying and holding assets they believe will accrue value, they also short-sell assets that they believe are expected to decrease in value. This is more popular in liquid markets.
Hybrids — in certain cases, managers possess a hybrid strategy where they are investing in high conviction projects in both liquid and illiquid (public and private) crypto markets. Some managers choose to only be long in both markets, while others choose to actively short-sell where they are able to.
Directional — these are managers that run highly sophisticated quantitative models that produce either “risk-on” or “risk-off” trading signals that direct these managers to execute their discretionary investment strategies in an aggressive or passive way depending on the perceived market cycle. Momentum, sentiment, trend and some systematic algorithmic trading strategies also tend to be popular here. This strategy can be focused on a single cryptoasset, a basket of cryptoassets, or relative value trading between assets, but is largely reliant on a quantitative model.
Arbitrage / Market Neutral — these are managers that utilize strategies that remove or limit market risk while capturing alpha returns. Exchange arbitrage strategies capitalize on the price discrepancies between different exchanges geographically while managed futures, developing derivative strategies, or HFT seek returns with minimized market risk. Systematic algorithmic trading strategies also tend to be popular here.
Credit — these are managers that are leveraging the traditional lending business model and applying it to this new asset class. Such managers lend out cash in return for cryptoasset collateral (typically bitcoin and in certain cases, ether and others, at present) and earn interest/yields on these cash loans. This has been growing in popularity lately as many holders of cryptoassets want liquidity but do not want to monetize their positions to trigger taxable events or give up the potential upside they believe these cryptoassets can achieve in the future.
Mining — in many cases, we are seeing investment flows that are used to purchase the necessary infrastructure, equipment and warehouses to run large-scale mining operations for proof-of-work protocols whereby miners receive cryptoasset block rewards (typically new bitcoin and transaction fees) when they solve the complex mathematical problems required to process transactions in the digital currency system.
Staking — similar to mining, but with proof-of-stake protocols. The transaction confirmations are not done with hardware but with already owned cryptoassets (generally via bonding). We are seeing investment flows that are used to purchase cryptoassets for staking purposes where stakers earn a percentage (a yield) of their coins as a reward for processing transactions in the digital currency system.
Active Ecosystem Participation / Delegate Work Entities — We are seeing this as a developing strategy in the ecosystem, whereby the line between investors and active ecosystem participants is blurring. Traditionally, investors have provided value through capital, governance, operational partnerships and more. In crypto networks, we see some investors emerging as active ecosystem participants in addition to the investor characteristics laid out above; we are seeing such crypto network investors also engage in mining, staking, validation, bonding, curation, dispute resolution, node operation, network routing, and more to help shape the direction of networks they are invested in. We are also seeing the emergence of delegate work entities, whereby average users may elect a specific ecosystem participant to perform work on their behalf (in other words, delegating the work (e.g., resource contribution) required by a crypto network to a trusted party in return for remuneration). We will continue to monitor this trend for future consideration as its own sub-category, but for the time being we consider it to be a developing opportunistic-related strategy.
Typically, in the private markets, we see a bifurcation between private token strategies and private equity strategies. On the private token side, we see managers investing in projects undergoing private presales, follow-on sales, and the like, whereas on the private equity side, managers are focusing on business models that do not warrant a token and instead invest in these businesses as you would in traditional venture equity. We typically see three sub-strategies here, but for the purposes of our framework, we refer to this strategy as simply venture.
Seed/Early Stage — these are investors that finance the very early development of new products or services. Typically we see simple agreements for future tokens (“SAFTs”), simple agreements for future equity (“SAFEs”), pre-sales, and early stage private sales (typically Series A) as the investment structures in this strategy.
Growth — these are investors that finance expansion and help projects position for critical mass. Typically we see private sales (typically Series B, Series C, etc.) as the investment structures in this strategy. On the token side, we see larger private accredited investor rounds to increase distribution and participation.
Late Stage — these are investors that finance positions that are in the works of positioning themselves for public offerings. The projects can demonstrate significant growth but may or may not be showing value accretion. Projects have usually been in existence for more than three years. Typically we see later-stage private sales (Series D, and onward) as the investment structures in this strategy. On the token side, this might take the form of a public token sale, launch of the network, airdrop, or public listing of tokens on an exchange.
Active Indexing — these are managers that index to a certain benchmark but are overweighting or underweighting certain selected cryptoassets actively in an effort to outperform the benchmark.
Passive Indexing — these are managers that create a market-weighted index or portfolio in an effort to offer a diversified, low turnover, low cost alternative (e.g., basket funds and the like) to actively managed fund strategies. Usually there is a “smart” element to its genesis structural composition before it becomes fully passive.
The Vision Hill Benchmark Framework
Now that we’ve addressed market segmentation and various types of investment strategies, we can discuss benchmarking. In many cases, we have seen funds benchmark against bitcoin, which is satisfactory in certain cases (e.g., if you are a fundamental, long only manager expressing the sound money thesis). However, benchmarking against bitcoin may not be satisfactory if you are pursuing some of the other strategies we discussed previously. For instance, comparing an exchange arbitrage strategy to bitcoin, in our view, is comparing apples-to-oranges, because exchange arbitrage strategies are meant to mitigate market risk, whereas bitcoin significantly encapsulates market risk. We also see active managers benchmark against the Hold10 (Bitwise), the Crypto20 (Bletchley Park) and other passive cryptoasset index funds. While we believe these index fund managers are doing some great work in the industry, we also do not believe such index funds to be an appropriate point of comparison. A long/short, alpha-seeking manager, in our view, should not be benchmarking its performance against a basket of “blue chip” cryptoassets that do not incorporate short-selling as a strategy or other active strategies in an attempt to outperform the market. The same can be said for any venture-like managers that are sourcing and investing in early-stage projects that they believe have the potential to capture outsized, risk-adjusted returns in the next several years. Comparing these managers to the aforementioned “blue chip” basket of cryptoassets is also comparing apples-to-oranges, in our view, given the different time horizons, risk profiles and liquidity levels associated with the differing strategies.
In consideration of the foregoing, what we propose here is a framework, or starting point, for comparing like-kind managers. In other words, we seek to provide points of reference for measuring performance among similar investment strategies, and we plan to deliver these performance statistics on a quarterly basis going forward.
It is important to note that there are seven widely-accepted essential properties a good benchmark should possess that can be remembered with the acronym SAMURAI [Source: Managing investment portfolios: A dynamic process (CFA institute investment Series), Third edition, John L. Maginn, Donald L. Tuttle, Jerald E. Pinto, Dennis W. McLeavey].
· Specified in advance — The benchmark is specified prior to the start of the evaluation period;
· Appropriate — The benchmark is consistent with the manager’s investment style or area of expertise;
· Measurable — The benchmark’s return is readily calculatable on a reasonably frequent basis;
· Unambiguous — The identities and weights of securities are clearly defined;
· Reflective of current investment opinions — The manager has current knowledge of the assets in the benchmark;
· Accountable — The manager is aware and accepts accountability for the constituents and performance of the benchmark; and
· Investable — It is possible to simply hold the benchmark.
While benchmarking, and the statistical measures that accompany it, can provide investors with insights into a portfolio’s risk/return profile, concentration levels and holdings tendencies, one consequence of the practice is that all measurements of a portfolio’s dimensions are evaluated in relative, rather than absolute, terms. Thus, it is important to discern whether a benchmark is actually appropriate for an investor based on that investor’s investment objectives, risk tolerance and time horizon. We believe good benchmarks should put returns into context and provide insight into the value of tactical decisions.
Accordingly, presented below (where sufficient data was available), are what we believe are comparison points of like-kind managers for the three-month period ended June 30, 2018. For confidentiality purposes, we are unable to disclose any specific performance numbers or specific manager names; we only present measures of central tendency and generalized brackets of performance to protect manager privacy. We also elected to solely present data for the full three months leading up to June 30, 2018 (April, May and June), and not other periods such as year-to-date given many funds launched at different points in time over the course of 2018.
Vision Hill Research Crypto Benchmarks: 2nd Quarter 2018 Performance*
*It should be noted that these performance metrics are quarterly metrics as of a particular point in time (April 1, 2018 through June 30, 2018), that represent only a limited snapshot and should not be considered as long-term indications of performance. This analysis should only be used as of this point in time as a framework. The cryptoasset class remains highly volatile and in the early stages of development, and past performance is no guarantee of future results. Additionally, for the purpose of this exercise, we had 35 data points to consider for the three-month period ended June 30, 2018. We expect this number to increase going forward as we diligence more funds and receive assistance from the community.
For the purpose of this exercise, we had 35 data points to consider for the three-month period ended June 30, 2018. Out of our internal database of 325 crypto funds and counting, we have diligenced approximately 65 funds so far as of the date of this report and found many of those funds were either pre-launch, or launched during the second quarter of 2018 (and thus didn’t have a full quarter worth of data). For data to be presented in the aforementioned quartiles, we decided a minimum of five data points must be available for a particular strategy. Accordingly, a breakdown of the data points available for each strategy is presented below:
* As of Q2’18, we have not tracked venture funds. We do, however, plan to begin tracking and incorporating such data in future releases.
As noted previously, we seek to provide points of reference for measuring performance among similar investment strategies, and we plan to deliver these performance statistics on a quarterly basis going forward. We are targeting mid-November 2018 for our release of third quarter 2018 (ending September 30, 2018) fund performance, which we hope includes much more data.
A Call to Action: Help Us Build a Better Benchmark
It should be noted that we do realize these benchmark comparison points are imperfect for a variety of reasons, hence we propose this as a framework at best. First, the number of data points we have used in this exercise (35) is extremely limited. Typically, statistically significant sample sizes warrant a minimum of 30 variables. It is our goal to work with the community to collect more data over time so that future quarterly releases can grow more robust in size and fund managers can start comparing performance against reliable data points of like-kind managers and strategies. Thus, if you are a crypto fund and are reading this, and want to contribute to helping us, please add us to your monthly and/or quarterly performance distribution list. We have set up email@example.com for this very purpose and will look to automate this over time.
We acknowledge that there are certain flaws and limitations surrounding our proposed benchmark framework. For starters, we understand there are certain risks to self-reporting, whereby managers may choose to only report performance figures during periods they believe they outperformed on a relative basis, and not report performance figures during periods they believe they underperformed on a relative basis. It is our every intention to combat this bias by only including managers that consistently report to us on a monthly and/or quarterly basis and that we believe are in good standing.
We also realize backfill bias (adding the performance of a fund months or years after benchmark inception) and survivorship bias (underperforming funds are more likely to be terminated over time) can both generally skew performance data. Given how early we believe it to be in the crypto asset class, as well as the fact we have diligenced approximately 65 funds so far as of the date of this report and found many of those funds were either pre-launch, or launched during the second quarter of 2018 (and thus didn’t have a full quarter worth of data), we believe backfill bias and survivorship bias risks to be relatively low. We nonetheless remain cognizant of these biases and plan to take appropriate action where necessary in the future to mitigate the impacts of such biases.
In addition to the foregoing, we understand that managers may also undergo style drift in certain cases. The definition of investment style may be ambiguous or inconsistent with the manager’s investment process and may also deviate from the definitions and categorizations we provided earlier. We plan to keep a close eye on this going forward, and will re-categorize managers as appropriate, but only after careful review and discussions with such managers about how to best classify their strategy as time goes on.
Lastly, we refrained from listing redacted peer rankings (e..g, “Fund A”, “Fund B”, etc.) because we believe these to lack many of the SAMURAI benchmark properties listed above; they are not known in advance, and are also subject to survivorship bias (underperforming funds are more likely to be terminated). We also do not believe this to be of high value at this early stage until the benchmark framework becomes more robust and participation by the fund manager community becomes more ubiquitous and fully comprehensive. Moving forward, we will look to develop a transparent opt-in system of reporting and ranking guided by community consensus and traditional market best practices.
In sum, we understand benchmarking is a challenging topic in the crypto asset class given how early it still is. We do not believe we have a perfect solution for benchmarking, and are not trying to present this piece as such. Rather, we are proposing a framework, or starting point, and a call to action to the community to help us work on improving this framework so that it may one day become more widely adopted by fund managers and institutional investors.
Thus, if you are managing a crypto investment fund and are reading this, and want to contribute to helping us, please add us to your monthly and/or quarterly performance distribution list. We have set up firstname.lastname@example.org for this very purpose.
Additionally, if you have any feedback or comments surrounding how we can improve this benchmark, we’d love to hear from you. You can email us at email@example.com with “Benchmark” in the subject line.
Vision Hill Advisors is a crypto asset and blockchain focused fund of funds. Through a proprietary fund manager selection process and institutional level due diligence, Vision Hill aims to lead investors into the future of digital assets. Vision Hill brings together a team with extensive experience in traditional financial markets, a deep passion and understanding of crypto and digital asset markets, and a history of risk and portfolio management.
The content provided herein should not be considered investment advice, and is not a recommendation of, or an offer to sell or solicitation of an offer to buy, any particular security, strategy, or investment product.