Release 3: Feeds and Integrations

Feed Services

Databricks Integration

Databricks is an ecosystem where users can analyse data, and build data products in an agnostic cloud-hosted environment. Databricks is also renowned for the analytics and AI tools they provide.

The specific area we are focusing on here is the Databricks Marketplace.

Macrobond customers can now ask to receive their bespoke data feed directly into Databricks.

The main advantage of doing so is to offer an additional way of consuming the data via the Macrobond Data web API or FTP service for instance, and reduce the ETL ('Extract, Transform, and Load'). Macrobond data would be readily available for consumption directly into the customers' Databricks instance.

The publicly available Macrobond Pay-As-You-Go data feed on Databricks is available here.

Key highlights:

- Reduced ETL

- Bespoke data feed

- Minimum latency

‍

Content

Macrobond ONE: Now-Casting Economics

The term now-casting is a contraction for now and forecasting and it is defined as the prediction of the present, the very near future and the very recent past. Now-casting is relevant in economics because key statistics on the present state of the economy are available with a significant delay. This is particularly true for those collected on a quarterly basis, with Gross Domestic Product (GDP) being a prominent example.

Now-Casting Economics delivers frequently updated near term predictions of key macroeconomic variables.  These predictions are generated by a set of statistical models, which run continuously on the Now-casting’s proprietary platform.  As soon as a new value for any of the input data series is published, the relevant Now-Casting models process the new information and update their predictions.

The dataset currently covers 33 countries, additional  key macroeconomic statistics which include proprietary indices such as the Now-Casting Surprise Index.

Data navigation improvements

1. From International to National Source

Some sources are both international and national sources. When we treat these sources as international only, users can only find the data within Source & Release.

We will now treat the below sources as national sources, so that we can expose structure within Country & Region, allowing users to navigate thematically within the tree to find the concepts they are after. First release includes ECB and Eurostat.

2. US Labor Market

Some of the steps involve:

• Removing discontinued nodes

• Moving wages & earnings from Labor Market to Income & Earnings

• Taxonomy harmonization e.g. ‘underemployment’

• Deeper tree-nodes to denote better specific concepts

3. Customer-driven fixes, involving custom tree creation to ensure thematic navigation (after) vs. flat or no hierarchy (before)

Including, but not limited to:

• US Dallas Fed Texas Manufacturing Outlook Survey

• US Retail Trade

‍

Revision History expansion

Users will now be able to benefit from Revision History (access to vintages and associated time stamps) on International sources.

We are turning the feature on for Eurostat first; ECB will be next.

As a reminder, in the last delivery cycle, we also turned on Revision History for daily time series to capture the vintageTimeStamps, to be used as publication dates.

Facilitating cross-regional comparison - International & Subnational RegionKeys

1. International sources’ RegionKeys

RegionKeys allow users to change the region of analysis in their documents, or to link together similar time series programmatically, around a unique concept.

We are starting with Eurostat and will prioritize expansion based on customer requirements.

2. US subnational RegionKeys

Macrobond will now allow users to flip the Region being analysed at the sub-regional level e.g. state, MSA, city etc.

This enhancement covers 132 releases in the US

3. Chinese subnational RegionKeys

Similar to above, this enhancement covers 56 releases in China.

‍