Infinite Power Factors
An unbiased multi-factor model built on
academic research and powered by AI and ML
Allowing you unique insights into the cryptoasset markets
Gain unique insights
into the cryptoasset market.
Let artificial intelligence and machine learning integrations guide your experience within Evai’s constantly evolving platform.
The ease with which an asset can be bought and sold without impacting the market price. Liquid assets are normally stable, less influenced by market slippage and easily transferred, making them more appealing to traders.
The community consensus toward an asset or market. Sentiment can be qualified via psychology-based indicators such and the Fear and Greed index. We seek to capture asset sentiment and identify bullish and bearish biases.
Return per unit of risk. This is a fundamental measurement of investment performance. The higher the risk-adjusted return, the better the investment. We use risk measures to identify assets that have performed well compared to those that have underperformed.
The rate of change of an asset's price movement over time. Financial assets experience uptrends and downtrends. Measuring momentum helps ascertain the strength and the continuation of an existing market trend.
Quantifying utility involves determining asset application, adoption, community engagement and development activity. We can measure asset utility and overall quality by analysing metrics such as the number of active addresses, GitHub pull requests, supply distribution and on-chain volume.
Measuring the effect of behavioural biases on asset value. Behavioural economics state that various cognitive biases can introduce market anomalies and impact the intrinsic value of assets. We exploit this hypothesis to measure potential divergences from fundamental value.
Weighing excess to determine asset bubble risk. An asset bubble caused by over-enthusiastic investors increases the chance of reversal. We seek to gauge the probability of a bubble forming and determine the odds of a significant pullback in price.
How correlated an asset is to the broader cryptoasset market. When the market risk is high the asset will magnify the performance of the market. This can reflect positively on price in an uptrend but will pose investment challenges in poor economic conditions.
Determined by the current price multiplied by the circulating supply, market capitalisation is an important determinant of cryptoasset value and demand. We rank assets according to capitalisation and use this as an indicator of potential market performance.
your experience within Evai’s constantly evolving platform.
How Evai's Multi-Factor Model Works
From raw on-chain data to a sophisticated global cryptoasset rating:
A step-by-step breakdown of the Evai ratings process.
Crypto Exchanges (Markets) for data
Rating based on a single factor
MULTI-FACTOR RATING BASED ON THE MOST STATISTICALLY SIGNIFICANT FACTORS
What sets Evai apart?
The Evai platform combines Artificial Intelligence with a machine learning self-correcting and autonomous ratings protocol, removing human bias. Evai continuously learns, maintains and remains decentralised to produce unbiased ratings.
Evai’s groundbreaking model has been developed from a synthesis of academic research and financial market experience. Powered by AI and ML the autonomous model constantly evolves each day increasing its overall accuracy.
Its range of risk-adjusted return ratios alongside our proprietary illiquidity measure identify fundamental factors crucial to determining crypto ratings and long-term asset value.
Unleash the power of AI and Machine Learning
Artificial Intelligence selects and combines indicators with optimal weightings to determine the greatest accuracy of each risk factor. The AI software runs real-time tests on all indicators, selecting the most efficient with results recorded.
Machine learning techniques are then implemented to assess the daily test results and modify the parameters of the AI software. The tweaking of the multi-factor model’s selection criteria continuously improves the overall accuracy of the ratings over time.
Our ratings are designated on an alphabetical scale ranging from A1 to D. The Evai A1 rating (“A1”) denotes cryptoassets of the highest technical and quantitative characteristics.
Evai’s D rating (“D”) indicates an unstable technical foundation and high risk, while a rating of U denotes an unratable cryptoasset, typically due to insufficient data.
The constituent rating scores are combined to return the overall asset rating (A1-D). The overall rating is an equally-weighted average of the indicators.
For instance, if two indicators were employed with the first returning a rating of “B3” (6), and the second returning an “A1” (2) they would combine to produce an average overall rating of 4 (B2).
Our selected crypto indicators drive the Evai ratings platform, informing our power factors, highlighting risks and providing an indication of long-term asset value.
Evai draws on a variety of financial indicators, ranging from risk-adjusted return ratios and illiquidity measure to pricing factors and sentiment analysers.
We use Artificial Intelligence to select and combine indicators with optimal weightings to determine the greatest accuracy of each risk factor.
Turnover ratio is the volume of cryptoassets traded relative to the outstanding assets. The higher the turnover ratio, the more frequently the cryptoasset is being exchanged. The easier it is to exchange, the more liquidity and valuable the asset.
The Turnover Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
This measure is defined as the standardised turnover-adjusted number of zero-trading volume days over one month. A cryptoasset with a higher number of zero daily volume is less likely to be traded and, thus, less liquid. (Liu, 2006).
The Adjusted Turnover Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
Amihud Illiquidity Ratio (Amihud, 2002) represents liquidity premium that compensates for price impact. It is measured as cryptoasset returns relative to volume. Cryptoassets with high Amihud ratio have a large price impact as buying and selling will move the price by a relatively large amount. These cryptoassets are considered relatively less liquid than cryptoassets with low Amihud ratio.
The Amihud Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
Evai’s exclusive illiquidity measure based on the 2011 research of Professor Andros Gregoriou.
The Gregoriou Ratio is a modification of the Amihud ratio that compares price impact against turnover rather than volume. It, therefore, overcomes some of the disadvantages of the Amihud Ratio like size bias. The lower the ratio, the smaller the price impact of orders and the more valuable to cryptoassets.
The Gregoriou Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
Roll (1984) measures the extent to which market-making will cause cryptoasset prices to move in response to the bid-ask spread. The larger this zig-zag movement, the lower the liquidity and the more difficult to exchange cryptoassets at a stable price. This measure is usually used over high frequencies (intraday).
The Roll Spread is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
This will measure the actual cost of trading as two times the spread between the bid-ask midpoint and the actual price traded. It will require information about the price at which transactions are executed as well as the bid-ask spread. The difference between the quoted and effective spread can be positive or negative, providing information about the true cost of trading.
The Effective Spread is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
Quoted bid-ask spread is the difference between the best bid and best ask price. A narrow spread implies lower trading costs and more liquidity.
The bid-ask spread is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
The Crypto Fear & Greed Index is evaluated as an equally weighted index of five indicators including volatility, market momentum, volume, cap factor and crypto social media history.
The indicator quantifies a simple but important investing psychology, i.e. most investments happen mainly due to greed or fear and most sell-offs similarly happen mainly due to either greed or fear.
The Fear and Green Index is constructed to run from zero to 100. A score of 100 is a top rating and a score of zero equates to the lowest rating.
Evai understands the importance of social media ranking. The Evai cross-platform cryptoasset online performance index represents a holistic view of cryptoasset performance in Facebook, Twitter, Reddit and GitHub.
This index allows investors to track the specific information they need to drive their unique investing strategies.
The Social Development Index is constructed to run from zero to 100. A score of 100 is a top rating and a score of zero equates to the lowest rating.
The Sharpe Ratio measures the return on cryptoassets over the risk-free rate relative to the standard deviation of those returns. It is the return per unit of risk and the higher the measure the better the investment. Sharpe (1966).
The Sharpe Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
The Sortino Ratio also measures risk per unit of risk, but now risk is measured as downside deviation. This is the deviation below the minimum accepted rate of return (MAR). Therefore it only combines worse-than-expected outcomes in the measure of risk. Sortino (1991, 1994).
The Sortino Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
A Moving Average will capture the underlying trend for the cryptoasset. It is calculated as the simple moving average of closing prices or an exponentially weighted moving average. These can be of varying lengths.
The Closing Price relative to the Moving Average is using to identify ratings. Strong uptrends give strong ratings and strong downtrends give low ratings. Signals are to be optimised using machine learning.
The Rate of Change is one of the many indicators that try to capture momentum. It is calculated as the percentage change over a specific period. The rate of change over a period of say 10 days, would be monitored for evidence that momentum is increasing or decreasing. It would also be compared against the price to identify divergence of price and momentum.
The price relative to the Rate of Change index is used to identify the momentum. A higher price with more momentum relates to a positive rating while a higher price with momentum that is not moving higher is a sign that the trend may be vulnerable to reversal and attracts a lower rating. Specific levels are to be optimised using machine learning.
Moving average convergence divergence (MACD) is a trend-following momentum indicator that shows the relationship between two moving averages of a security's price. The MACD is calculated by subtracting the 26-period exponential moving average (EMA) from the 12-period EMA.
The relationship between the price and the momentum indicators determines the rating. Higher prices with momentum confirm the trend is intact. Higher prices without momentum are a warning sign that the trend may be about to change.
Return on Investment (ROI) relates net income to investments made in a cryptoasset, giving a better measure of cryptoasset profitability. Measuring ROI helps in making a comparison between different cryptoassets in terms of profitability and asset utilisation.
The level of profitability over the past year is used to assess the rating. The period to be assessed is to be optimised using machine learning.
This is the peak return that has been achieved over the last month and the final return for the month or preceding months for the lagged version.
It is a variable that seeks to capture well-known behavioural biases in decision-making related to the importance attached to peak and end experience by investors.
Positive peak and end readings provide a high rating. Negative peak and end readings provide a low rating. Machine learning will be used to fine-tune intermediate indications.
Fibonacci retracement levels can help to quantify levels of retracement risk. They are based on the Golden Ratio and are considered to be levels where profit-taking or reconsideration may naturally have been completed. The greater the retracement risk, the more vulnerable the cryptoasset.
Retrenchments across assets are ranked and used to assess the level of retracement risk. Where the potential retracement is small, the rating will be high. Where the potential retracement is high, the rating will be low.
Ichimoku cloud is part of the Japanese collection of technical tools. It is based on a combination of a number of moving averages and trading ranges. It provides information about trend as well as support and resistance levels.
Ichimoku cloud provides a number of indicators that feed into the rating process. These indicators are optimised for particular cryptoassets and market conditions.
The Bollinger Band is one method of identifying extreme price movements. It is a trademarked property of John A. Bollinger. The Bollinger Band consists of a moving average of the price combined with upper and lower bands that are based on multiples of the standard deviation of the moving average.
The band identifies price extremes and the relative performance of the bands (converging or diverging) will help to determine market conditions, consolidation or trending. They can be used to identify risk, reversal potential or the relative weight to apply to consolidation-trending tools.
Bollinger Bands can identify the initiation of a trend (a positive rating if positive and a negative rating if negative). They can also show the risk of a pullback if the extreme is reached and not sustained (negative rating for a negative pullback and a positive rating for a positive pullback). Extremes and parameters are to be identified and optimised with machine learning.
The Stochastic Oscillator is another momentum indicator. This indicator compares the current price to the price range over a given period. High readings show strong upward momentum and low reading strong downward momentum. The divergence between the cryptoasset price and the momentum indicator is also used.
The price relative to the Stochastic Oscillator is used to identify the momentum. A higher price with more momentum relates to a positive rating while a higher price with a momentum that is not moving higher is a sign that the trend may be vulnerable to reversal and attracts a lower rating. Specific levels are to be optimised using machine learning.
The Relative Strength Index (RSI), developed by J. Welles Wilder, is a momentum oscillator that measures the speed and change of price movements. The RSI oscillates between zero and 100. Traditionally the RSI is considered overbought when above 70 and oversold when below 30.
The relationship between the price and the RSI indicator determines the rating. Higher prices with the RSI above 70 confirm the trend is intact and a positive rating. Higher prices with the RSI moving lower are a warning sign that the trend may be about to change. This will reduce the rating. Levels can be optimised for particular cryptoassets and market conditions.
The Market Factor shows the relationship between the return on a cryptoasset and the return on a basket of cryptoassets. It is a measure of systematic risk and it shows how much this cryptoasset would be affected by shocks that affect the whole cryptoasset market - this is sometimes called the beta. A beta of one indicates that the return on this cryptoasset is very similar to the overall market. A beta above one means that the cryptoasset is very sensitive and will react more in both positive and negative way to changes in the cryptoasset market. A beta below one shows that the reaction to the market is muted. A higher beta is considered to be higher risk.
The Market factor is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
The Treynor Index is like the Sharpe Ratio and the Sortino Ratio as a measure of the return per unit of risk. However, in this case, the risk is measured as the Systematic Risk, the Market Risk or beta. Treynor, 1965).
The Treynor Ratio is compared to a moving average of past performance to assess the rating. The periods to be considered are to be optimised using machine learning.
Size is a measure of capitalisation. There is a size factor for equities with strong evidence that firms with lower capitalisation have a relatively high return even when risk has been accounted for. Our research suggests that for cryptoassets, there is a positive size effect with the cryptoassets with a larger capitalisation making higher returns, even when other risks have been accounted for.
The rating is assessed by comparing market capitalisation across cryptoassets. Those with the highest Capitalisation have the highest rating and those with the lowest have the lowest rating.
Build your own
Identify value and minimise risk with
unbiased and actionable insights.
Live, on-chain data fed into a highly customisable dashboard.
Build the market around you.
Stay one step ahead with access to multiple cryptoasset data sets.
Design your own dashboard with fully customisable parameters.
Set real-time alerts and never miss a move.
Evai's Research Team are constantly working to improve the accuracy of the multi-factor model by testing new indicators every week.
We currently support the following list of market indicators and metrics:
Turnover Ratio, Adjusted Turnover Ratio, Amihud Ratio, Gregoriou Ratio, Roll Spread, Effective Spreads, Quoted Spreads, Fear & Greed index, Social History Index, Sharpe Ratio, Sortino Ratio, Moving Average, Rate of Change, MACD, Proxy ROI, Peak End, Value Demand, Fibonacci Retracement, Ichimoku Cloud, Bollinger Band, Stochastic Oscillator, Relative Strength Index, Market Factor, Treynor Index, Market Capitalisation.
You choose the metrics. Build the market around you.
Choose from a selection of on-chain metrics and financial market indicators to gain exclusive insights into the cryptoasset market.
Design your own intuitive dashboards using the metrics and indicators that matter most to you. Visualise and chart cryptoasset investments with some of the most extensive and unique on-chain data available.
Data dashboards are set up to accessibility guidelines and available in dark & light mode which you can set via the preferences menu.
Each data visualisation has customizable parameters allowing you to set your criteria for each dashboard.
Drag, drop, and move your data visualisation charts around your dashboard and design the optimised display according to your preferences.
Arrange Data Hierarchically
All data visualisations can be arranged according to your preferences, using metrics and indicators you value most.
Data visualisations are updated every 5 minutes directly from live on-chain data.
We provide real-time, actionable alerts on crypto market movements, enabling you to time market tops and bottoms.