Data Exploration Workflow

Everysk’s data repository is equipped with powerful tools to help investors query, filter, run analytics, and explore data over time. The data exploration automation workflow is a user-friendly tool to do all this and more with no coding or SQL required.

Portfolio managers have a wealth of data that often goes underutilized. This can be for a host of reasons whether it’s because data is scattered across hundreds of different spreadsheets, only available as point-in-time snapshots, too cumbersome to extract out of pivot tables, or gated behind complex systems that require technical knowledge to query and access.

Everysk’s data exploration robot makes it straightforward to set up any sort of filter, aggregation, derived analytic, or time series that a portfolio or risk manager requires. The tool supports custom numerical calculations through a flexible expression engine and has a menu of time series-related functions for calculating properties such as cumulative sums and trailing z-scores.

Datastore Retriever Robot

Retrieves entire histories of daily portfolio data and analytics. Everysk helps clients consolidate all their holdings data, risk calculations and metadata in organized datastores tagged with dates and labels that make it easy to pull years’ worth of data for exploration.

Data Explorer Robot 1

Data explorer robots can do many tasks from big explorations to pulling out simple text values. For example, retrieve useful metadata for report titles like fund name, the most recent portfolio date, and Net Asset Value (NAV).

Data Explorer Robot 2

Create tables with current risk analytics aggregated by different dimensions. Use the data explorer’s configurable fields and aggregation menu to set up a table summing exposure, VaR and stress results grouped by your own custom strategy tag. Use the expression engine to scale the VaR values to a 20 day horizon. Use the filtering menu to exclude FX securities and cash positions.

Data Explorer Robot 3

Leverage the data explorer’s ability to explore data through time. By specifying a list of datastores rather than the latest, users can use built-in time series functions to quickly extract whole time series of any analytic in the data set. In this example, we’ll create a time series of cumulative YTD P&L in dollars.

Data Explorer Robot 4

The data explorer can offer insights by running analytics on time series, as well. For example, clients can pull their current tracking error, calculate their average tracking error and standard deviation over the last twelve months and display their current z-score to gain a sense of how their current performance relates to historical levels.

Report Generator Robot

Generate a daily report to summarize all the insights that were extracted above. This robot passes the data explorations into a report template that visualizes fund NAV, performance, exposures, risk, and tracking error in a format ready to be emailed out and used by stakeholders.