Big Data & Aggregated Risk

The answer to the global aggregated risk data problem?

New technologies seem to be pre-occupying the banking industry today in a big way.  Blockchain, the Cloud and Big Data are three technical ideas that seem to be top of mind in today’s banking engine room. These solutions confront, as banks have to, a set of legal/regulatory challenges and management and reporting requirements that are imposed upon them.

A recent survey (‘Big Data as the Key to better Risk Management’) published by the Intelligence Unit of the Economist (interviewing 208 bankers of all types and in all geographies) queries the extent to which Big Data technology is used or likely to be used and the uses to which it would be put.  The report is authoritative and worth reading, but summary relevant conclusions would be:

  • The bankers surveyed believe that liquidity and credit risk will pose the greatest challenges for their institutions over the next three years
  • These two areas of risk reflect the greatest potential for big data and its associated tools to make an impact on improving risk management
  • If the occurrence of these risks is rare, then their effect is huge, given the fact that finance firms operate on relatively small amounts of capital and large amounts of more volatile financing
  • The most successful approaches require a centralised bank-wide reporting structure

What is the problem?

Banks, their customers, their transactions and their trading activities generate huge amounts of data, much of which is either discarded or not used.  This is particularly relevant to the area of risk management, in which the large and internationally active banks are being challenged by regulators to report intraday – on an aggregated basis, for all currencies, trading activities, legal entities, consolidating customer positions.  (BCBS 239)

To report an intraday credit position with a bank’s largest counterparty (including all of its legal entities and associate companies across all currencies and instruments) requires a bank to model the position from transactions, balances, FX rates and other sources from numerous systems and data sources.  Key intraday data for the bank as a whole – notably intraday liquidity and credit positions – is simply not available in most banks today. Most institutions have been built to operate with batch overnight processes, procedures and systems!

The bank has to manage intraday and report at a 90? angle different to the way in which all systems are working today – and across all of them.  Is it unreasonable that large institutions should be held to account for this kind of data?  Of course it isn’t.  And the fact that it is hard to do will not stop regulators from requiring it to happen, and, ultimately, from attempting to break up the “Too Big To Fail” institutions if it can’t be done.

What does a Solution look like?

Let’s forget for a moment the means of building the solution, and rather concentrate on what it must look like to the risk manager users.

  • Firstly, and most importantly, this feels like an air traffic control operation in which risk managers have access to a variety of dashboards that monitor those risk positions which matter – either to the regulator or to the bank itself, which will have its own views on its appetite for different types of risk. These dashboards are as real-time as you can make them – timely is the word from Basel.
  • Secondly, it is a data source, which is designed to take current data to model against specific risks and potential changes which might affect values. This needs tools to visualise the results of specific changes in market circumstance – effectively stress tests.  The SEPA Consultancy’s Intraday Liquidity Simulator would be one example of a tool to model the historic and future use of liquidity against changed circumstances.  What happens if the oil price drops by 50%?  What happens if a major and systemically significant corporate gets into trouble, say AIG, BP, Glencore?  What happens if the largest counterparty has an operational problem in its USD high value payment system, or if, say, the CHAPS system in the UK went down for the day?  Most banks can answer some or all of these questions locally, but on a global and aggregated basis?
  • And it is a given that intraday monitoring is a centralised bank-wide activity, even if it is also needed within legal entities.

How do you get there?

Today, you don’t necessarily know which data is going to be important in the future.  Up to recently, for example, the idea that you would need the time of settlement for a correspondent payment, as opposed to the time at which the transaction was applied to the accounting system, would have been neglected.  For a full risk management capability, you need all balance and transaction data, and potentially a whole lot more besides to build the capability.  Indeed, the largest part of building the facility is to provide the data feeds from bank systems and other relevant feeds and sources, for which the data is in multiple formats if indeed it is properly structured at all.  This is a huge soup of data, which needs to be complete in all of its ingredients, before it can be stirred and readied for consumption by its risk management users.

This is where the Big Data or Business Intelligence technologies potentially show their value.

What is Big Data?

The Gartner Group definition is as follows.  “Big data is high volume, high velocity, and/or high variety information assets that require new forms of processing to enable enhanced decision making, insight discovery and process optimization.”

New techniques are required because of the speed, the volume, the structured and unstructured nature of data.  The tried and tested tool of IT – the relational database – is simply not up to the task of handling the volumes and structures of data.  There is a new set of vendors and a new IT lexicon which probably starts with Hadoop to be learned, but this note is an encouragement to consider the use of the technology and not a primer for it.  We will leave the last word to the Economist survey.

“The ability to harness larger and more diverse data pools in support of business decision-makers holds the promise of both reducing losses by managing risks and increasing revenue by highlighting business opportunities. Successfully managing risks today requires that bankers identify, access and analyse trusted data and share their results across the bank.”

Peter Miller, TSCL Associate – November 2015