Cryptocurrency trading has emerged as the killer app of blockchain technology. Billions of dollars are exchanged daily on digital asset exchanges. Yet, it appears that most of the volumes are from retail investors. For example, earlier research showed institutions accounted for only 10% to 24% of the bitcoin trading volumes.
Big institutional investors that dominate traditional asset classes are just starting to dip their toes. Before they arrive en masse, trading technology has to catch up to what institutions are used to in equities and other established asset classes.
In this article, I’m going to outline the missing technology elements that we noticed while developing our in-house trading platform for managing our own digital assets treasury.
The ‘Traditional’ Setup
“Institutional investors” is a very broad term that includes a wide range of market participants, from passive fund managers overseeing index funds to niche hedge funds trading obscure securities.
Most investors have more or less the same trading technology infrastructure provided by big investment banks that act as prime brokers. For example, Goldman Sachs, Morgan Stanley, and JPMorgan serviced 80% of all hedge funds launched in 2018, according to Preqin data. They take care of almost everything: trade execution, securities lending, custody, clearing and settlements, and back-office work. These services were previously manual and are now done mostly with software. This work is largely commoditized, and institutional investors usually can pick whichever brokers they like most.
Also, most institutional investors have access to the same financial market data, provided by, for example, Bloomberg and Thomson Reuters. Such services cost tens of thousands of dollars per user (for example, Bloomberg Terminal is $24,000 a year) but hardly provide an advantage, since everybody has the same information.
Data Sources For Digital Assets
Unlike in traditional asset classes, data for digital assets is disparate and not easy to use. Although, in crypto, data is mostly free. In particular, bitcoin and Ethereum’s public blockchains contain the full record of all transactions for anybody to read and analyze. Most cryptocurrency exchanges also publish their volumes.
However, digital assets are a new phenomenon, and there are no dominant pricing models, for example, for fundamental analysis. Most digital assets do not have cash flows, and there are no obligations to disclose information in a structured way. In particular, U.S. federal disclosure requirements do not explicitly mention digital assets, and nontraditional assets require only “brief” descriptions.
Since retail investors and traders make up most of the market, it is important to have a measure of the prevailing sentiment. Media coverage and social media chatter could be used to estimate the market mood.
To get an edge in the market, you need to either find a new way to use the existing data or create new data sources. This may involve building your own analytical platform, just as my company has, but it’s also possible to uncover insights in public data by combining and integrating different sources.
Automatic Strategy Testing And Deployment
Before trading real money in the market, a trading strategy is first backtested on historical data and then forward tested using live data.
For equities and other established assets, this can be done automatically using existing software, either developed in-house or provided by trading infrastructure specialists. Data is readily available for all instruments, although it could be very expensive.
This is a lot less straightforward in digital assets. First, there are fewer years of market data observations. Bitfinex is one of the oldest bitcoin exchanges, was founded in 2012. Binance, one of the biggest altcoin exchanges, launched in 2017. There are dozens of other exchanges, and integrating all market data — and adapting it for strategy testing — is not an easy task.
Because programmatic strategy testing in digital assets is difficult, a significant portion of portfolio managers in digital assets currently rely on passive holding and infrequent discretionary trading. During the bull run of 2017, this worked well, and crypto hedge funds on average generated 1,700%. But, in 2018 they made -71%, according to Eureka data.
Quantitative algorithmic strategies are rare in digital assets, and there are few testing solutions for them. We had to develop our own trading platform to test strategies, optimize them and deploy them in production while managing risks. It’s important to be able to input and monitor strategies throughout the whole cycle: backtesting, forward testing and live trading. Be sure that your company’s solution checks these same boxes.
Order Execution For Cryptocurrencies
In traditional asset classes, brokers guarantee the best execution. It is their legal responsibility to execute orders on behalf of their customers (for example, institutional investors) to get the best possible result in terms of price, costs, speed and other considerations. This is usually done programmatically, and it’s not something most investors need to think about.
The whole idea of best execution in digital assets is still new. First, there is no regulation in crypto that would ensure that investors receive the best price execution. In fact, front-running is rampant on many exchanges, according to Bloomberg. Second, in our experience, many large trades are done via OTC desks, since even the biggest exchanges might not have the liquidity to execute orders quickly and without moving the price. This means the real costs of trades could be quite high.
The solution is to develop basic order execution capabilities in-house. Your goal here should be to minimize slippage through executing orders programmatically. This should eventually become standard practice, but right now, even a simple programmatic execution can help to increase returns.
Today, the digital assets market is still too underdeveloped for most institutional investors. Technology infrastructure that they expect is not yet built — possibly because there are too few paying clients who need that. This also means that there are plenty of opportunities for new players that are willing to invest in developing their own trading infrastructure capable of more than just passive holding.