Blog article
See all stories »

Dirty Data

No not the title of a chapter in a John Le Carre book, nor an invitation to spice up my pc. But in a recent survey looking at impediments to accurate counterparty risk measurement (and management) the top answer was, dirty data. If you extend counterparty risk measurement to include cash and liquidity exposure I'd say you're beginning to see the tip of an iceberg. My company receives account information in SWIFT or proprietary format; we transform, cleanse, aggregate and pass on to the receiving party. We see and resolve many problems for both parties to the transaction. For the provider bank, often the data comes from legacy systems and is not real time, and if it is, it’s not necessarily sourced from a truly real time system. The bank can tell you your account balance every 5 minutes but if the core system is not updating the account ledgers until every hour.

If the two system times are in sync, the next issue is the quality of the data itself. That’s no problem, I hear you say, we use SWIFT. Marvelous, but there's many a slip between mandatory fields and interpretation; blank spaces and non ref fields etc. If the receiving bank wants a full intraday service it will require real time MT900 & MT910 backed up by MT942's throughout a day with an end of day statement for ultimate control purposes. Does the MT900 & MT910's reconcile to the intraday MT942's, is there a floor limit, are the reference fields consistent, are fwds advised in different ways, does the MT942 replicate items on a preceding MT942?

All being well, the receiving bank has received the MT900 and MT910 in real time, and all references fields match, there’s no duplication, and the 942's accurately reflect all the preceding MT900 and MT910's, I should now be in a position to know my exact real time cash balances and can therefore manage my liquidity and counterparty exposure in real time? That should satisfy my regulators.

Not so fast, it seems the Provider bank neglected to tell me that some large bulk transfers e.g. securities cash are not credited until the last hour of the day. Added to the problem is that some large account items, normally book transfers, are not covered by intra day SWIFT advices, in fact they are not passed to the accounts until long after the currency clearing day is closed and I’ve already covered my supposed position.

If you’re experiencing some or all of the above and dirty data really is a problem, drop me a line. 

4478

Comments: (1)

Ketharaman Swaminathan
Ketharaman Swaminathan - GTM360 Marketing Solutions - Pune 06 May, 2011, 11:10Be the first to give this comment the thumbs up 0 likes

@Roy M: 

Props for pointing out the role of reference data in liquidity risk management.

To stretch the Le Carre reference, hope the survey you quoted makes data stewards "A Most Wanted Man" in payments shops. In a recent blog post, I'd narrated my experience of the acute risk caused to a bank's liquidity positon when a new processing software insisted upon 11-character BIC codes whereas the legacy feeder software could only supply 8-character ones.  

The above experience also showed that, apart from dirty data, technical stability of the payments landscape should become "The Mission Song" for payments shops in order to manage liquidity risk appropriately.

Now hiring