How do I access a valuation manifest?

The valuation process produces a manifest, a body of information that explains how the results of every call to the GetValuation API are generated.

The manifest is a key element in the audit process. It can answer questions like ‘how were my market rules resolved?’ and ‘how many times did I access pricing data from a particular vendor for a particular instrument?’.

At the time of writing, the only way to access the manifest is to use Luminesce, LUSID's data virtualisation and reporting tool. Note a separate license may be required; contact us for more information. Note we intend to make the manifest available via the REST API, SDKs and web app, and also enable administrators to query across multiple users, soon.

Using Luminesce 

1. Sign in to the LUSID web app using the credentials of a LUSID administrator.
2. From the left-hand menu, select Insights > Request Logs:
   
3. Enter GetValuation in the Operation column (highlighted in red below) and choose the API call that corresponds to the date and time at which you performed the operation:
   
4. Copy the Request ID for that API call (highlighted in green above), for example 0HMGJ1RDFFFHG:0000001B.
5. From the left-hand menu, select Data Virtualisation > Query Editor:
   
6. Paste the following Luminesce SQL query into the editor, substituting:
   • Your Request ID retrieved from Insights as the argument to the UserRequestId filter, encapsulated in single quote marks.
   • Your valuation schedule datetime as the argument to the EffectiveAt filter, encapsulated in hash marks. Note this is not the Date and Time of the GetValuation API call as recorded in Insights above, but rather the exact datetime supplied to the valuationSchedule.effectiveAt field in the body of the original request to GetValuation:

     select *
     from Lusid.Logs.Valuations.Manifest
     where UserRequestId = '0HMGJ1RDFFFHG:0000001B'
     and EffectiveAt = #2022-03-07T16:30:00Z#

7. Press the Run button. Luminesce should return the manifest as a table of data like this: