Coin Metrics is pleased to announce the version 2.11 release of our CM Market Data Feed. This release contains support for new exchanges, expanded coverage for many of our data types and metrics, significant quality-of-life improvements to our API, and many more improvements and bug fixes.

What’s New

  • New Exchanges: Coin Metrics continually expands our exchange coverage and adds additional data types for exchanges we already cover. This release includes the following new exchanges and their supported data types:
    • Coinbase Derivatives futures metadata, trades, order book snapshots, real-time order book, candles, and volume metrics.
    • Coinbase International futures metadata, trades, order book snapshots, real-time order book, candles, and volume metrics.
    • Bullish futures metadata, trades, order book snapshots, real-time order book, candles, and volume metrics.
    • ErisX futures metadata, trades, order book snapshots, real-time order book, candles, and volume metrics.
    • Deribit options order book snapshots, and real-time order book.
    • Deribit spot trades history was backfilled to the inception of their spot markets on 2023-04-24.
    • Deribit futures predicted funding rate and contract prices.
    • Deribit trades for spot, futures, and options markets now include several Deribit-specific fields: mark price, index price, implied volatility, and a liquidation indicator.
    • Bybit futures predicted funding rate and contract prices.
    • Binance options metadata, trades, order book snapshots, real-time order book, quotes, implied volatility, greeks, and contract prices.
    • OKEx options real-time book.
    • LMAX spot data was expanded to include sol-usd, matic-usd, link-usd, usdc-usd, and pyth-usd markets.
    • Kraken futures funding rates was expanded to include Kraken’s perpetual linear multi-collateral futures (PF) markets. We previously were only collecting data for Kraken’s perpetual inverse futures (PI) markets.
    • Binance spot and Coinbase spot additional metadata for markets is now being collected.
  • Expansion of our liquidity metrics. Coin Metrics publishes several liquidity metrics representing the depth of the order book at pre-defined percent away from mid-price and slippage at pre-defined U.S. dollar order sizes. Exchanges limit the amount of book depth that is provided, and Coin Metrics will assign a null value if the order book provided by the exchange is not sufficient to calculate the depth or liquidity for a pre-defined level. To deal with this issue, we have added additional granularity for our depth and slippage metrics so that users can more easily utilize more of the order book in determining liquidity. We added the following metrics:
    • liquidity_depth_0_2_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_3_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_4_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_5_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_6_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_7_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_8_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_0_9_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_1_5_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_3_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_4_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_6_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_7_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_8_percent_{ask|bid}_volume_{units|usd}
    • liquidity_depth_9_percent_{ask|bid}_volume_{units|usd}
    • liquidity_slippage_20K_{ask|bid}_percent
    • liquidity_slippage_30K_{ask|bid}_percent
    • liquidity_slippage_40K_{ask|bid}_percent
    • liquidity_slippage_50K_{ask|bid}_percent
    • liquidity_slippage_60K_{ask|bid}_percent
    • liquidity_slippage_70K_{ask|bid}_percent
    • liquidity_slippage_80K_{ask|bid}_percent
    • liquidity_slippage_90K_{ask|bid}_percent
    • liquidity_slippage_200K_{ask|bid}_percent
    • liquidity_slippage_300K_{ask|bid}_percent
    • liquidity_slippage_400K_{ask|bid}_percent
    • liquidity_slippage_500K_{ask|bid}_percent
    • liquidity_slippage_600K_{ask|bid}_percent
    • liquidity_slippage_700K_{ask|bid}_percent
    • liquidity_slippage_800K_{ask|bid}_percent
    • liquidity_slippage_900K_{ask|bid}_percent
  • Added support for our liquidations metrics for 5m periods. Our liquidation metrics aggregate buy and sell liquidations across various periods, and we previously only supported 1h and 1d frequencies. We added the following metrics served through our /timeseries/exchange-metrics, /timeseries-exchange-asset-metrics, and /timeseries-pair-metrics endpoints: liquidations_reported_future_{buy|sell}_{units|usd}_5m
  • Backfilled the history of our funding rate metrics. Our funding rate metrics futures_aggregate_funding_rate_{all|usd|coin}_margin_{8h|1d|30d|1y}_period represent the average funding rate weighted by open-interest. Our Bitcoin funding rate metrics now start on 2020-07-27 and other assets are similarly backfilled to the start time of our funding rate data.
  • Enhanced our market candles to ensure complete historical data and improved completeness.
    • When new markets are listed by exchanges, we will automatically backfill our candles data to the market’s inception, ensuring complete historical coverage.
    • We fixed a bug that could, in rare circumstances, cause gaps in our candles data during periods of no trades. All missing gaps were filled so that our candles are a continuous timeseries.
    • We fixed a bug that could, in rare circumstances, cause the volume in U.S. dollars to be incorrectly calculated as zero.
    • We fixed a bug for defi market candles where candles were being inproperly filled forward even if there was no activity and were not removed after determining the pool is no longer active.
    • Additionally, we now recalculate the previous 3 hours for each market, incorporating any delayed trades to ensure our candles are as accurate and complete as possible.
  • Expansion of our universe of assets and pairs for aggregated quotes. Our aggregated quotes are analogous to national best bid-offer for equities and are served through our /timeseries-stream/asset-quotes and /timeseries-stream/pair-quotes endpoints. We increased our aggregated quotes coverage from 45 to 139 assets.
  • Our API expanded support for wildcards to all of our /timeseries/* endpoints. Now users can specify patterns like assets=*, markets=exchange-*, exchange_assets=*-asset, pairs=btc-*, institutions=*, indexes=CMBI* to easily get data on a range of entities instead of enumerating each entity in a comma-seperated list.
  • Improved resampling logic when using granularity parameter. Our API supports a granularity paramter which allows users to resample high-frequency timeseries to a more user-friendly frequency like 1h or 1d. Previously, the API only returned observations exactly at the rounded timestamp defined by the granularity, which caused issues when no observation was available at that exact time. Now our API returns the first observation that is equal to or after the resampled rounded timestamp, ensuring a more accurate and complete timeseries, with no gaps, when using the granularity parameter.
  • Support for case-insensitive assets and pairs parameters in our /timeseries-stream/asset-metrics, /timeseries-stream/asset-quotes, and /timeseries-stream/pair-quotes endpoints. Previously, these endpoints only accepted lowercase assets and pairs.

Other Announcements

  • Expansion of our asset coverage for CM Prices. We released a new version of our CM Prices which expands our coverage universe for the CM Reference Rates and CM Principal Market Prices from 887 assets to 1,040 assets. Our latest release also improved our ability to select the highest-quality defi markets, regardless if the asset is the base or quote asset for a particular defi pool.
  • Trusted Exchange Framework v2.2. We published a new version of our Trusted Exchange Framework which quantitatively assesses exchanges along several dimensions.

Improvements

  • Significant upgrades in our API to reduce the latency of our real-time order book data served through /timeseries-stream/market-orderbook.
  • Decreased the likelihood of discontinuous jumps for metrics ReferenceRate and principal_market_price_usd served through /timeseries-stream/asset-metrics.
  • Migrated our feed handlers to use Bullish’s latest spot API version.
  • Reduced the delay in supporting new assets, pairs, and markets for our candles and metrics. There is a short delay in supporting new entities because new assets undergo human review at Coin Metrics to ensure that ticker conflicts are resolved and metadata is populated accurately, and this delay is now shorter.

Bug and Data Quality Fixes

  • Collected missing Binance futures trades data for a small number of futures markets.
  • Backfilled small gaps in the data for our volatility_realized_usd_rolling{24h|7d|30d} asset-metrics.
  • Removed a small period of duplicate trades from Bithumb. The affected candles and volume metrics were also recalculated.
  • Fixed a bug in which btc-quoted assets would receive improper weighting in our aggregated quotes served through /timeseries-stream/asset-quotes.
  • Fixed a bug which caused liquidity metrics for BitMEX futures to be calculated incorrectly.
  • Fixed a bug which could cause our liquidity metrics to be null in cases where order book data is delayed.
  • Fixed a bug where our API would return an empty response instead of status 400 in our /reference-data/*-metrics endpoints when a metric was not found.
  • Fixed a bug that would occasionally result in incomplete order book snapshots for bitFlyer.
  • Fixed a bug that would occasionally result in incomplete Gemini quotes data.
  • Fixed a bug that where our API would sometimes incorrectly return status 400 when a user is requesting asset-metric ReferenceRate with frequency 1d-ny-close.
  • Fixed a bug that prevented /timeseries-stream/market-orderbooks from sending update messages in rare instances.
  • Fixed a bug that caused empty order book snapshots served from /timeseries/market-orderbooks.