Picking the Right Cryptocurrency Exchange
A Dilemma of Choice
If you want to own the latest new tech stock or purchase an ETF, where do you go? Most likely you use an online broker (ie: Robinhood, E-Trade, Fidelity, and the likes). Your chosen broker will then execute your desired trade on a stock market. This is where choices for trade execution become severely limited. A majority of stock trades are executed on a select few marketplaces. In the U.S., the NYSE and NASDAQ dominate the majority of stock trading volume and are the largest two exchanges in the world.
This doesn’t end up being too big of an issue though, because while there are only a few options, the exchanges are regulated and are predominantly trusted by the overall investing community. The average stock trader isn’t too concerned about how their trade gets executed, they just want the desired end result to be achieved in a quick and secure manner. In the world of cryptocurrency, traders face a vastly different reality. Currently, there are more than 250 exchanges listed on the popular crypto-data aggregator site, coinmarketcap.com. Other sources, such as bitcoin.com, have put the number of cryptocurrency exchanges at over 500. While cryptocurrency brokers are growing in popularity, traders mostly have to execute trades on the exchanges themselves. This has led us, along with the cryptocurrency community at large, to ask the following question:
How should we decide where to trade cryptocurrencies?
This is a question that our firm set out to solve. There are a number of exchanges out there and not all of them are great venues for executing trades. There are plenty of voices that have chimed in on this same issue in the form of impactful research and regulations. Bitwise, an asset management firm specializing in cryptocurrency investing, sent an in-depth report to the SEC that includes a method for finding exchanges that report accurate trading volume measurements. Then there is the stringent NY Bitlicense, issued by the New York Department of Financial Services (NYDFS), that requires cryptocurrency-related companies to meet a series of standards in order to operate in New York. Together, Bitwise and the NYDFS have established well-respected standards for judging the quality of cryptocurrency exchanges. However, both focus on specific aspects of exchanges and tend to leave out issues that we find relevant.
Bitwise touched on the trading volume metrics that are often misleading and skewed by exchanges to attract publicity and gain an edge on competitors. The NY Bitlicense requires adherence to strict standards regarding consumer safety and security of deposits. Both groups, along with other leaders in the space, raise important questions that we took into consideration. For our purposes, we want to expand on previous efforts to establish standards and incorporate factors that are important to the trading needs of our firm.
Introducing the Exchange Ranking Formula
With 200+ exchanges there is no shortage of choices. There is also no shortage of negative news stories regarding exchange hacks and poor operational practices. We find that with such a large number of options, we have to be increasingly diligent with our selection process. In order to find a concise way to establish our own exchange standards, we use an equation that weighs all the exchange characteristics that our firm considers important. We start broadly by identifying five categories that we judge exchanges on:
- Security
- Platform Downtime
- Liquidity (Volume)
- Operations
- Trading Capabilities
It is our belief that by applying a score out of 100 to each exchange for each characteristic, and then taking a weighted average, we can arrive at a numerical value to represent an exchange’s ranking.
The next logical question to consider is how much should you weigh each factor? Every platform has its unique strengths and weaknesses. For example, the Poloniex exchange offers over 100 different currency pairs and is a preferable choice for a trader looking to diversify across many digital assets. However, that exchange has experienced hacks in the past. While there is no doubt that firms such as Poloniex put tremendous efforts into operating in a secure fashion, past hacks are an important factor. On the other side, there are exchanges that offer only a few trading pairs, but with a cleaner history.
Our Preferences
Internally, we breakdown our evaluation of a cyptocurrency exchange into five distinct components. Most importantly we look at security features and the amount of time that the platform is operational. In addition, there are still three additional categories that we look at (trading, operations, and liquidity). We go into further detail about what exactly we look for later in this piece.
How it Works
Finding an ‘optimal’ exchange isn’t an objective science. The system we developed works for us and works best when choosing between several different exchanges. To that end, it is imperative to start with a select number of exchanges, instead of looking at all 200+ at once. We look at a dozen or so exchanges at a time and by using this methodology, can designate an optimal choice. Then it is a matter of going through each exchange’s details and establishing a score out of 100 for each category. After that, we multiply each score by the pre-determined percentage weight and then arrive at an average score for each exchange. Those scores are then ranked and a clear distinction can be made between preferred and less preferred exchanges. The end result is a clear ranking of each exchange we are considering compared to a larger pool of platforms. To see the exact details we look for in each category click on the links embedded in each line of the below outline.
The Categories explained
Security
We break Security into three categories:
- Front-End Security (33% of Security Score)
- Back-End Security (33% of Security Score)
- Hacks (33% of Security Score)
This is the first category we look at, with good reason. At the end of the day, we want to execute our trading strategies with confidence knowing that our funds are well secured. We broke this one category into front-end security, back-end security, and hacks so that we could further quantify the total level of security present on different exchanges.
Front-End Security
Front-end security represents the features that provide secure access to, and withdrawals from, exchange accounts. We assign 20 points for each feature present. These features include:
- two-factor authentication
- email verification
- trade and withdrawal verification
- multi-user admin privileges
- Ability to whitelist API keys by IP address**
** We define multi-user admin privileges as the ability to delegate different levels of access (i.e., view-only, non-withdrawal/trade only, etc.) to multiple users or sub-account holders, either through varied Application Programming Interface (“API”) permissions or through the traditional web user interface.
Back-End Security
This sub-category refers to the actions an exchange has taken to keep users’ funds/assets secure, such as security certifications (ex. SOC, PCI, etc.), percentage of user assets held in cold storage, penetration testing, etc. We found that most platforms are using the front-end features mentioned above. The biggest variance came from looking at the documentation that explains all the practices instituted by exchanges on the back-end. For this section, we relied on information posted by the individual exchanges. We found that most exchanges are fairly open about the measures they take to secure funds. For example, these are links to the security practices of Coinbase and Binance, two major exchanges in the space.
Hacks
We define hacks as events in which a significant number of exchange accounts lose a large amount of value. While unpleasant at the moment, hacks can shine a light on bad technology and practices. In the end, though, we want to check and see how many times an exchange has been hacked as it can still be a sign of faulty security measures. When looking at each exchange’s history with hacks, or lack of, it is easy to see which exchanges really stand out from a security standpoint. If an exchange has never been hacked we give it a 100, while exchanges that have been hacked once or twice may receive a lower score. We don’t consider exchanges with more than two major hack incidents.
— Some notable exchanges that have never been hacked are Coinbase, Gemini, and Itbit.
In our methodology, we value transparency so we think more favorably of exchanges that share how they are protecting user funds. Assigning a score for this sub-category is subjective and we assign a perfect 100 to multiple exchanges that have thorough documentation.
Platform Downtime
This category is comprised of the following sub-categories:
- Withdrawal Halts (50% of Platform Downtime Score)
- Exchange Monitoring (50% of Platform Downtime score)
For this category, we look at how often exchanges go through periods of degraded service. In our experience, we are looking at two sub-categories: withdrawal halts and exchange monitoring.
Withdrawal Halts
Occasionally, exchanges need to halt withdrawals. When these events occur there is often a public outcry which makes finding out information about past incidents as simple as looking through Reddit or Twitter for large groups of disgruntled traders. While overall subjective, we ranked exchanges for this sub-category in the following manner:
- No halts (100)
- Occasional halts, with advanced notice; (80)
- Occasional halts, with no advance notice; (60)
- A large number of halts, with or without advanced notice. (0)
We kept this section open-ended on purpose. After looking at numerous exchanges we were able to see that a “large number of halts” meant more than 4 per year. This conclusion can only be drawn after investigating an exchange and comparing it against similar companies. What we really look for is the process in which exchanges communicate and plan withdrawal halts. Greater transparency results in increased access and control over user funds, therefore receiving a higher score.
Exchange Monitoring
It is fairly common for exchanges to occasionally go offline to fix errors, implement upgrades, perform security tests, etc. For our purposes, it is crucial that we are notified within a reasonable time frame prior to an exchange going offline so that we can monitor the status of that exchange and properly control our exposure and risk. Hence, we decide to use the following metrics when considering this sub-category:
- The exchange alerts users about planned downtime in advance (100)
- The exchange provides a way to monitor its online and offline status (50)
- Past incidents of the exchange going offline without prior notice or due to one-off technical issues (0)
A lot of the criteria for the platform downtime category is subjective and for a good reason. Investors carry varying thoughts on what is an acceptable amount of time for an exchange to be offline or have degraded service. For that reason, it is important to realize the unique preferences of traders and institutions. Once a score is established for each sub-category, the appropriate percentage weights are applied which results in the final platform downtime score per exchange.
Trading
For this category, there are a few distinct exchange characteristics that we consider. The sub-categories we use are:
- Currency options (50% of Trading Score)
- Trading fees (50% of Trading Score)
For this category of exchange analysis, the results end up being more quantitative, which results in clear distinctions.
Currency Pairs/Options
Depending on the desired number of tradable cryptocurrency pairs, preferences can vary greatly. We look at how many trading pairs an exchange supports, how many quote currencies are supported and if the exchange supports fiat currency. We prefer to have the ability to trade in the top 25 cryptocurrencies by market cap and want our trades to involve fiat currencies. These preferences differ from a trader who wishes to trade in less known cryptocurrencies. For that reason, it is important to again realize that this methodology is meant to be altered to fit specific trading needs. For our purposes, we assign a score of 100 to exchanges that offer the top 25 cryptocurrencies and fiat pairs. From there we give lower scores to exchanges that have fewer currency options.
Fees
There are numerous fee structures that exchanges utilize to capture revenue. For our purposes, we look at specifically whether a platform uses a maker/taker fee structure. For reference, a ‘maker’ is a trader who adds liquidity to the market, and a taker is the one who then takes that liquidity. Since exchanges want liquidity, they will often incentivize makers with lower trading fees, or in some cases, no trading fees. We want to have the ability and incentive to trade with as little commissions as possible, given the high turnover of our strategies. For that reason, we prefer the maker/taker fee model. In addition, when comparing any number of exchanges, it is beneficial to find the average trading fee across all fee tiers and use that as a comparison tool to find a low expense platform. Since this sub-category is strictly quantifiable, the exchange with the lowest trading fees receives a 100 and then exchanges with higher fees rank lower and lower depending on how high their fees are.
Once the scores for both these sub-categories are found, the total Trading Capabilities score can be found by averaging the two scores.
Operations
This category looks specifically at the overall quality of service provided by an exchange, and the accompanying fees. This is another pretty broad category and required us to develop several sub-categories:
- Fees (33% of Operations Score)
- Customer service (33% of Operations Score)
- API capabilities (33% of Operations Score)
Fees
Most exchanges charge varying levels of fees when withdrawing assets, and a few even charge a fee when making a deposit. For us to compute a number that can be compared among other exchanges we average the fees that are required to deposit or withdraw ETH and BTC. Doing so, allows us to arrive at a distinct number and establish a ranking of platforms by the magnitude of deposit and withdrawal fees. The exchange with the lowest fees receives a score of 100, and each subsequent exchange is given its percentile value, with the highest fee-tiered exchange given a zero. For example, if there are 10 exchanges then the exchange with the lowest fees would receive a 100 and the next lowest would receive a 90 and so on.
Customer Service
We prefer to work with exchanges that dedicate resources to customer support. When things inevitably go wrong, it is important to have the option to quickly receive answers. To gauge the level of customer support on exchanges we track how long it takes for a customer service representative to respond to a submitted inquiry. Once again, we created this methodology to reflect our own preferences and find that we overwhelmingly favor exchanges that can respond to a customer support ticket within 12 hours. A lot of exchanges will respond much faster, sometimes within minutes, but we believe 12 hours to be the minimum acceptable time frame. For us, an exchange that responds to customer request tickets within a few minutes receives a 100 while an exchange that takes multiple days will receive a 0.
API Capabilities
The last sub-category we track for operations is API capabilities. APIs allow users to program actions such as buys, sells, limit orders, withdrawals, etc. If you have never heard of APIs and have no interest in using them for your trades, you can safely ignore this parameter. However, if you’re like us and are interested in the quantitative and automated environment that APIs facilitate, this can be an important parameter in your equation. It is common practice for exchanges to offer API capabilities but the quality of service varies greatly. For that reason, we classify exchanges into a couple of buckets that are subjectively judged.
- Excellent endpoints and documentation (100)
- Usable endpoints and documentation (80)
- No documentation (60)
- Out of service/No API. (0)
Once all the sub-categories are calculated, the percentage weights are applied to find the total Operations score per exchange.
Liquidity
Lastly, we look at the trading volumes on each exchange. For this metric, we don’t need multiple sub-categories or subjective rankings. What is required is a valid data source for exchange trading volume. As shown in the Bitwise report mentioned earlier, a large number of exchanges inflate trading volume statistics in order to seem larger than competing exchanges. We look for exchanges that are honest in their reporting methods so that our trades are being executed with a valid number of other counterparties. To that end, we compare exchanges based on concrete 24-trading volume numbers. Out of the set of exchanges being considered, the one with the highest trading volume receives a 100 and then exchanges with less trading volume receive incrementally lower scores based on their percentile ranking. It is important to note that the data used to compute the average 24-hour trading volume should come from a reputable source and represent a trusted exchange. As an example, there are 10 exchanges (outlined in the Bitwise report) that are exceptional when it comes to reporting honest trading volume numbers. Those exchanges can be found here.
Looking forward
As our firm continues to grow and change, so will the way we look at cryptocurrency exchanges. The features we focus on now may change as will the options available in the market. New developments in the cryptocurrency space, such as decentralized exchanges and traditional brokers like Fidelity adding cryptocurrency trading services means we will continue to iterate on our methodology. Traders will always have to make a concerted effort to find the most beneficial venue to execute their trades. We are confident that by focusing on the factors listed above, we can pick out the best platforms to execute our trades and reach our return goals.
We want to hear from you!
If you made it this far, then you are just as passionate about picking great exchanges as we are. We are aware that while this methodology and factors work for our firm, they aren’t perfect. If you feel like we are leaving something out or are including something unnecessary, reach out to us here. Also, if you wish to receive a weekly newsletter with blockchain and cryptocurrency related news and updates, subscribe below.
Disclaimer:
Blockforce Capital is providing this publication for informational purposes only. This is not investment advice, nor an offer or solicitation to purchase or sell any investment or attract any investment of funds. The data and opinions in this publication are for informational purposes only and no representations or warranties are given or implied — the information may change, be inaccurate, incomplete or outdated. You should do your own research and analysis before deciding to invest in digital assets/cryptocurrencies and/or use an exchange platform. Nothing in this publication is an endorsement or recommendation of any exchange platforms or digital assets/cryptocurrencies.
