What is bitemporal data?

This is a way of storing information where each piece of data is captured with two datetime values:

• effectiveAt: the datetime the piece of data is considered ‘valid’, in a business context.
• asAt: the datetime the data was physically recorded in LUSID.

The general principle is explained in this Wikipedia article, this FINBOURNE blog post, and also by this third-party blogger.

Consider a live price quote for a stock, for example:

Vodafone Group PLC (VOD) @ 20 Aug 2018 10:02 UTC = 176.14 GBX

The effectiveAt for this quote is 2018-08-20 10:02:00 UTC. UTC is Universal Time Coordinated, and represents time at longitude 0°.

The asAt would be the datetime the quote was actually saved into the system. This datetime may be slightly later than the effectiveAt, as there may have been a delay in the quote being published or uploaded into LUSID.

Whenever any data is retrieved from LUSID, the user must supply datetimes for both effectiveAt and asAt (if omitted, the default is 'now', or latest). These datetimes are then used to filter the universe of data, to locate the most appropriate value.

The primary benefit of storing data bitemporally is that it allows all data changes to be recorded and audited. Consider if this quote had been saved with the wrong value. Using the bitemporal model, we are able to correct the value retrospectively. The corrected piece of data would have the same effectiveAt as before, but the asAt datetime on the updated record would correspond to the datetime the correction was made in LUSID. Since LUSID stores both versions of the piece of data, you are able to retrieve the latest (i.e. corrected) value using an asAt of 'now', or retrieve the original value using the asAt from the original record.

Understanding perpetual and time-variant data items

In LUSID:

• A perpetual data item has a value that is considered permanent, for example a person's place of birth. It's still possible to correct wrongly-entered data and roll back the asAt timeline to see a history of values, but it's not meaningful to roll back the effectiveAt timeline; for a given asAt datetime, the value is the same from the beginning to the end of time. Perpetual data items are considered monotemporal in LUSID.
• A time-variant data item has a value that can vary over time, for example a person's address. When you enter such a value, you specify an effectiveFrom datetime; you can either specify an explicit effectiveUntil datetime, or LUSID can apply one for you if/when you supersede the value (that is, enter one with a later datetime). For a time-variant data item, it's possible to both:
  • Correct wrongly-entered data and roll back the asAt timeline to see a history of values.
  • Roll back the effectiveAt timeline and see a history of a time-series of values.

  Time-variant data items are considered fully bitemporal in LUSID. For an examination of perpetual vs time-variant properties, see this article.