Analytics & Research Tools

Analytics Tools Overview

How Crypto Compare, horakl Quadrants and Market Intelligence measure performance and risk across tokens, categories and portfolios.

horakl provides three analytical tools built on the same market data and the same measurement principles: Crypto Compare, horakl Quadrants and Market Intelligence. Each answers a different question about the universe horakl tracks.

A shared measurement basis

The tools describe performance and risk in a consistent way, which is what makes cross-entity comparison meaningful:

• Normalized performance starts at an index value of 100 on the first day of the selected window.
• Every entity in a comparison is measured over the same calendar window, with the same reference dates.
• Return and risk indicators are calculated for the selected period only. They are not lifetime portfolio statistics.
• Market categories are represented by fixed, equally weighted baskets of currently eligible tokens that have complete daily price history for the period.
• Missing historical observations are never interpolated, so a gap in the data is shown as a gap rather than filled with an estimate.

Return and risk indicators carry the same definitions across the platform.

Choosing a tool

• To evaluate specific assets, categories or research portfolios against each other, use Crypto Compare.
• To see the shape of the market and where individual tokens sit relative to the benchmark, use horakl Quadrants.
• To read the aggregate state of the tracked universe for a period, use Market Intelligence.
• To examine one asset in depth, including its price history, use the Token Explorer table and token pages.

Open the tools

Crypto Compare

Compare cryptocurrency tokens, market categories and horakl research portfolios on normalized performance and risk.

Crypto Compare places several entities on a single chart and a single metrics table so their performance and risk can be read side by side. It compares three kinds of entity: cryptocurrency tokens, crypto market categories and horakl research portfolios.

A comparison has one primary entity and up to five additional entities, so a maximum of six entities can be displayed at once. Mixed comparisons are supported: a token can be compared with a category, a portfolio, or any combination of the three.

Three periods are available: 7D, 30D and 90D. The selected period determines the window used for every entity in the comparison.

What Crypto Compare helps you understand

• How different assets, categories and research strategies performed over the same window, expressed on a common percentage basis.
• How much risk each entity carried while producing that return, using drawdown, consistency, Sharpe Ratio and volatility.
• Whether a return was broad or concentrated, by comparing a single token against the category basket that contains it.
• How a model research portfolio compares with the assets and categories it is exposed to.
• Whether a return was consistent day to day, or the result of a few large moves.

How to use it

1. Select the primary entity. Search for a token, category or portfolio and set it as the primary entity. The primary entity is the reference for the relative-performance column, so it is the natural starting point of an analysis.
2. Add entities to compare. Add up to five more entities. The search can be filtered by type (tokens, categories, portfolios) and, for tokens, by category. Entities already in the comparison are marked and cannot be added twice. When the limit is reached the picker is disabled until an entity is removed.
3. Choose the period. Switch between 7D, 30D and 90D. Every series, indicator and coverage figure is recalculated for the new window; the reference dates shown above the chart confirm which window is displayed.
4. Read the chart and the tables. The chart shows normalized performance. Below it, the performance summary shows each entity return for the period and, where comparable, its difference against the primary entity in percentage points. The risk and return metrics table reports the five indicators for the same window, and the category coverage panel discloses what each category basket contains.
5. Share or save the view. Use Copy link to copy the current comparison. The selection and period live in the page address, so a shared link restores exactly the same view and browser back and forward move between earlier configurations.

Available indicators

Why no Score is shown

The comparison engine returns no universal cross-entity score, so Score is deliberately not displayed on this page. Portfolio Score is documented separately and applies to portfolios only.

How to interpret results

Read absolute performance and risk-adjusted performance as two different questions. ROI describes how much an entity moved; Sharpe Ratio, volatility and drawdown describe the path it took to get there.

• A higher ROI does not imply a better risk profile. Two entities can end the window with similar returns while one of them fell much further at its worst point.
• Max Drawdown describes the worst loss an observer tracking the same window would have seen, not the loss an investor is guaranteed to experience next.
• Consistency describes how often daily returns were positive. A high value with a low ROI describes a smooth but modest path; a low value with a high ROI usually points to a few large moves.
• A negative Sharpe Ratio means the entity returned less than a zero-risk reference over the window. It is a description of the period, not a judgement of the strategy.
• Volatility is a measure of variability, not of quality. It is most useful when comparing entities over the same window.
• Volatility and Sharpe Ratio are only meaningful when they are read from the same table, since they are computed on the same daily returns.

Limitations and considerations

• Missing daily observations are not interpolated. Gaps appear as gaps in the chart.
• Entities with incomplete history for the window may return no return and no metrics, or may not be comparable with the primary entity.
• Category results carry survivorship bias: the basket uses current category membership, not the membership that existed at the time. Coverage is disclosed per category.
• Portfolio lines describe tracked model portfolios, not executed trades. They do not include transaction costs, slippage, fees or the timing differences of a real account.
• Public comparison of a portfolio never reveals gated content: holdings, allocations, composition details and paid research are not requested by this page.
• Comparison windows end on the same reference date for every entity, but market data may be delayed, corrected or revised after it is published.
• Historical performance does not guarantee future outcomes.

Crypto Compare Methodology

How the Crypto Compare engine normalizes performance, calculates risk indicators and builds category baskets.

This article describes how the figures on the Crypto Compare page are produced. All financial calculations are performed by the horakl comparison engine; the page presents what the engine returns and never recomputes an indicator.

Entity types

One entity is designated as primary. Up to five additional entities can be added, giving a maximum of six per comparison. Identifiers are validated by the backend, and the same entity cannot appear twice in one comparison.

Comparison window and reference dates

A comparison selects a fixed calendar window of 7, 30 or 90 days. The response reports the exact start and end date used, and every entity in the comparison is measured over those same dates, which is what makes the lines directly comparable.

Normalized performance

Every entity starts at an index value of 100 on the first day of the window and is expressed as an index for each subsequent day.

The normalized index is an analytical representation, not an asset price. An index of 110 means the entity is 10% above its starting reference value; an index of 92 means it is 8% below it. Normalization exists so that assets quoted in different units, category baskets and portfolio net asset values can be placed on one scale.

Indicator calculations

All five indicators are calculated from the daily observations inside the selected window. They are period-specific: they are not the lifetime portfolio statistics shown on portfolio pages, and they are not calculated over any other historical window.

Category basket methodology

A category is not a published index. It is a basket horakl constructs for the comparison:

• The basket uses the tokens that currently belong to the category and are eligible for comparison.
• A token is included only when it has complete daily price history for the selected window.
• Included tokens carry fixed, equal weights for the whole window. Weights are not adjusted for market capitalization.
• Missing historical prices are never interpolated, so no synthetic value enters the basket.
• The basket is reconstructed retrospectively from current eligibility and membership.

Because the basket is retrospective, a complete calculated series does not mean the basket covers every token that currently belongs to the category. Coverage is disclosed separately for that reason.

Category coverage indicators

Coverage is reported to two decimals. A basket may, for example, report that 23 of 73 eligible tokens are included, which is 31.51% historical coverage and a partial status. That figure is an illustration of the format, not a fixed value: coverage changes with the category and the window.

Missing data

• Missing daily observations are not interpolated and are not carried forward. The chart shows a gap.
• An entity with too few observations may return no return and no metrics for the window; the interface shows a dash (—) with an explanation instead of a fabricated number.
• Relative performance against the primary entity is withheld unless both sides have complete, comparable histories for the window.
• When no selected entity has observations for a window, the page states that there are no observations instead of drawing an empty chart.
• The comparison window uses the same reference dates for every entity, so a shorter history is visible as missing points rather than as a shifted window.

Details that require confirmation

The following details are enforced by the horakl comparison engine rather than by the application code, and are stated here only to the extent the product displays them. They should be confirmed against the engine specification before being documented as exact values.

• The minimum number of daily observations required before each indicator is published.
• Whether the engine applies any outlier treatment or winsorization before calculating returns and volatility.
• The exact rule the engine uses to decide that an entity has complete history for a window (for example whether a single missing day is tolerated).
• The engine refresh cadence for market data and derived series. The application caches comparison results in the browser for five minutes, which is independent of the engine refresh.
• Whether the engine publishes additional internal statuses beyond the complete and unavailable cases the page handles.

horakl Quadrants

A market matrix that classifies tracked tokens by absolute performance and by performance relative to the market benchmark.

horakl Quadrants is a 2x2 market matrix on the Token Explorer page. It sorts the tracked token universe into four groups for the selected period, using two independent questions: did the token rise or fall, and did it do better or worse than the market benchmark?

The matrix complements the other analytics tools. Crypto Compare answers questions about specific entities; horakl Quadrants answers where the whole universe sits, and it links directly to the tokens behind each group.

What horakl Quadrants helps you understand

• How the tracked universe is divided between tokens that rose and tokens that fell over the period.
• Which tokens are ahead of the market benchmark and which are behind it, regardless of direction.
• Whether strength is broad across the universe or concentrated in a minority of tokens.
• Where a specific token sits relative to the benchmark, by selecting its group and reading the token table.
• How each group behaves in aggregate, through average Change, average Consistency and average Max Drawdown.

How the matrix is organized

The two axes measure two different things and cross at zero:

• Vertical axis: the absolute Change of the token over the selected period, in percent. The upper half holds tokens with Change at or above zero; the lower half holds tokens with Change below zero.
• Horizontal axis: performance relative to the market benchmark, in percentage points. The right half holds tokens at or above the benchmark; the left half holds tokens below it.
• The four cells keep a fixed order and meaning; they are never rearranged.

Each cell shows the number of tokens it contains and three aggregates for that group: Average Change, Average Consistency and Average Max Drawdown. Colours identify the quadrant; they are not a quality grade or a recommendation.

How to use it

1. Set the period and filters. Choose a period (24H, 7D, 30D or 90D, defaulting to 7D) and optionally narrow the universe by category, blockchain or market-cap tier. Market Intelligence, the quadrant matrix and the rankings on the page all read the same filter state, so every module describes the same universe and the same period.
2. Select a quadrant. Selecting a quadrant restricts the token table below to that group and shows how many tokens it contains. Selecting the same quadrant again clears the selection, and clearing it restores the full filtered table.
3. Read the tokens. The token table reports market capitalization, Change, relative Change, Consistency and Max Drawdown for each token in the group, so a group can be inspected member by member rather than only in aggregate.

How to interpret the quadrants

For a beginner: the matrix answers two plain questions about every token. Did it go up or down? Was that better or worse than the market average over the same period? A token can rise and still be a laggard, and it can fall and still be a relative defender.

For an analyst: quadrant membership is a description of a period, not a signal. Pair it with the aggregates in each cell, with the Category coverage and risk indicators in Crypto Compare, and with the token table when a group needs to be inspected token by token.

About movement between quadrants

The matrix is a snapshot of the selected period. It does not plot an individual token path between quadrants and does not store quadrant history, so movement is observed by switching the period and comparing the two readings, or by comparing the same token across periods.

Limitations and considerations

• Classification depends on the selected period. A token can sit in different quadrants across 24H, 7D, 30D and 90D windows.
• Results depend on active filters. Narrowing by category, chain or market-cap tier changes both the universe and the benchmark.
• Relative performance is measured against the benchmark of the filtered universe, not against a global market index.
• Tokens without sufficient data for the period are excluded from the matrix and reported separately; they are not treated as underperformers.
• Cell aggregates are averages of the members of that group and can be influenced by a small number of large movers. Read the member count together with the averages.
• Stablecoins are excluded from the tracked universe.
• Quadrant placement is not a prediction and not an investment recommendation.

horakl Quadrants Methodology

The inputs, filters, classification rule and aggregates behind the horakl Quadrants matrix.

horakl Quadrants is produced by the horakl market engine and presented unchanged. This article documents the inputs, the filters and the classification rule, and marks the parts that are engine-defined rather than defined in the application.

Input data and universe

• The universe is the set of tracked tokens that horakl follows with market data.
• Stablecoins are excluded from the tracked universe, as stated in the tool.
• Each token is classified using its Change over the selected period and its performance relative to the benchmark of the filtered universe.
• Tokens that lack sufficient data for the selected period are reported as insufficient data and are not placed in any quadrant.
• Consistency and Max Drawdown are also reported per token and aggregated per quadrant.

Filters and parameters

The market-cap tiers are translated into an inclusive minimum and maximum market capitalization before the request is sent, and the engine performs the selection. The filters apply to every module on the page simultaneously: Market Intelligence, the quadrant matrix, the rankings and the token table.

Classification rule

Each token is placed by the sign of two measures. The axes cross at zero, and the boundary is inclusive on the positive side:

• Change at or above zero places the token in the upper half; Change below zero places it in the lower half.
• Relative performance at or above zero places the token in the right half; below zero places it in the left half.
• Combining the two halves gives the four quadrants: Positive Laggards, Market Leaders, Underperformers and Relative Defenders.

Relative performance is expressed in percentage points against the benchmark and the tool reads it as being above or below that benchmark. The exact formula the engine uses (for example whether it is the difference between the token Change and the benchmark Change) is engine-defined and is listed under details requiring confirmation below.

Benchmark

The benchmark is the reference the matrix compares every token against. It is the mean Change of the eligible universe under the active filters for the selected period, calculated by the engine and displayed unchanged. Market Intelligence shows it as the Market <period> indicator with the label “Filtered benchmark · eligible universe”.

Because the benchmark follows the filters, it changes when the universe changes. A token classified as a leader while filtering by one category may be a laggard under another filter, even with the same Change. The application never recomputes or substitutes the benchmark.

Quadrant aggregates

These aggregates are calculated by the engine over the members of each quadrant. The application displays them as returned and never derives its own averages. A quadrant with no matching tokens is shown with a zero count and unavailable values rather than an invented figure.

Availability and refresh

• Consistency is not calculated for the 24-hour period. The rankings panel states this explicitly instead of showing an empty list.
• The page caches results in the browser for five minutes, so changing filters quickly does not repeat an identical request.
• Market data is refreshed by the horakl data pipeline rather than in real time. horakl does not claim live pricing for this tool.
• When the engine is unavailable, the module shows an error state with a retry action instead of a partial or estimated matrix.

Details that require confirmation

The classification, the benchmark and the eligibility rules are enforced by the horakl market engine, which is not part of the application code. The documentation states only what the tool displays; the following items should be confirmed against the engine specification before being documented as exact definitions.

• The exact definition of relative performance used for the horizontal axis, including whether the benchmark is compared as a difference in percentage points.
• Whether the benchmark is equally weighted across the eligible universe or applies any weighting scheme, and how tokens with partial history are treated inside it.
• The minimum observation requirements that make a token eligible for Change and Consistency, and the rule that produces the insufficient-data group.
• The definition and calculation window of per-token Consistency and Max Drawdown as returned by the engine.
• The engine refresh cadence for the universe, the benchmark and the derived indicators.
• Whether the engine exposes any quadrant history that the interface does not currently display.

Market Intelligence

Four aggregate indicators describing the state of the tracked crypto universe for the selected period.

Market Intelligence is the aggregated module at the top of the Token Explorer page. It summarizes the tracked universe for the selected period in four indicators, so the market context is available before any individual token is examined.

All four indicators are derived from the same aggregated response as the quadrant matrix, using the same filters. They can never disagree with the matrix or the token table.

Available indicators

What the percentages are calculated on

The two percentage indicators use the same denominator, and it is not the full tracked universe. Tokens that lack sufficient data for the selected period have no Change and are excluded from both percentages; Tracked Tokens still counts them, which is why the percentages describe participation among tokens that could be measured rather than among all tracked tokens. A dash (—) is shown when no token in the filtered universe has a usable Change.

Data sources and aggregation

• Market data comes from the third-party market-data providers horakl uses for price history. horakl does not name a provider-specific figure in the interface.
• Aggregation is performed by the horakl market engine over the filtered universe; the page displays the aggregates unchanged and never recalculates them.
• Each token is represented once in the universe counts; tokens are identified by the horakl market-data identifiers used across the platform.
• The reference period is the filter selected above the module: 24H, 7D, 30D or 90D, defaulting to 7D.
• Missing observations are not interpolated. A token with insufficient data is excluded from the percentages rather than estimated.
• Results are cached in the browser for five minutes. Market data is refreshed by the horakl data pipeline, not in real time.

How to interpret Market Intelligence

The module answers two separate questions about participation, and the difference between them is often the most informative part of the reading.

• Positive describes how broad the movement is: the share of measurable tokens that ended the period at or above zero.
• Beat Market describes how much of the universe is ahead of the average: the share at or above the benchmark.
• Because the benchmark is a mean of the universe, a wide participation reading and a narrow Beat Market reading can coexist: many tokens can rise while fewer than half of them are above the average, and the reverse can also happen.
• The Market indicator sets the scale of the period. A −18% benchmark changes the meaning of a token with a −5% Change, which is a fall but a relative gain.

Market Intelligence describes the broad universe. Narrowing the filters turns it into a category or chain-level reading, because both the universe and the benchmark are recalculated for the filtered set. A single asset is never described by this module: an asset with a strong return does not define the market, and one exceptional mover does not explain a participation figure. Use Crypto Compare for entity-level analysis and the token table for the individual record.

Limitations and considerations

• Coverage is limited to the tokens horakl tracks and to the tokens that have usable data for the selected period. It is not a census of the crypto market.
• Stablecoins are excluded, so the universe describes volatile assets.
• Eligibility and filter effects matter: a category, chain or market-cap filter changes both the universe and the benchmark, so readings from different filters are not directly comparable.
• Market data can be delayed, corrected or revised, and horakl does not claim real-time pricing.
• Data may differ from other providers because of different universes, timestamps and revision policies.
• Historical aggregates describe the period they cover and are not a forecast.
• An aggregate indicator cannot describe the quality, liquidity or prospects of the underlying assets, and it is not investment advice.