18 December 2015
To: All authorised deposit-taking institutions
The Basel Committee introduced a globally harmonised liquidity framework by developing two minimum standards with the objective of promoting short-term and long-term resilience. The Liquidity Coverage Ratio (LCR) and the Net Stable Funding Ratio (NSFR) were developed to fulfil these objectives and to also enable regulators and investors to make meaningful comparisons between banks. APRA’s expectation is that ADIs with similar business models, balance sheets and customer groups would generate similar cash outflows under the LCR.
In early 2015 APRA conducted a consistency review1 across 14 large ADIs to determine whether they were taking a consistent approach to the interpretation and application of key terms in APS 210 Liquidity (APS 210) relating to the LCR. The area that demonstrated the greatest level of inconsistency was the assumptions relating to retail and qualifying SME deposits2.
This letter provides APRA’s observations of better practice in the approaches taken to determine whether retail deposits are considered stable or less stable. As part of ongoing supervision and the Committed Liquidity Facility ‘all reasonable steps’ assessment process in 2016, APRA will consider the extent to which ADI’s meet the expectations in this letter.
To qualify as ‘stable’, a deposit needs to be fully insured by the Financial Claims Scheme (FCS) and meet either the ‘established relationship’ or ‘transactional account’ criteria.
APRA views the ‘established relationship’ criteria as particularly important because once a customer has been deemed to pass the test, all retail deposits that are fully covered by the FCS may achieve a five per cent outflow assumption. The test is in principle a simple one for the relevant deposits, though APRA expects that ADIs will take care that the criteria applied are robust. For example, APRA expects that criteria applied would generally not allow new customers to be quickly and easily classified as having an ‘established relationship’ with the ADI.
Responses indicated that ADIs had adopted a wide variety of interpretations of the ‘established relationship’ test. Some ADIs focussed solely on the number of products that a customer had with the ADI while others took a multi-faceted approach.
Overall, the results of the review indicated that many ADIs had a low threshold for a new customer to meet the definition of ‘established relationship’. While the number and type of products are relevant considerations, APRA considers better practice is to adopt a multi-faceted approach, which includes a minimum relationship length (for example 12 months) and consideration of the relationship strength for determining the existence of an ‘established relationship’. In addition, where an ADI has an internal definition of a customer for whom the ADI is its single ‘prime’ or ‘main bank’, it may be appropriate to consider this in its definition of ‘established relationship’.
‘Transactional accounts’ are those commonly used for customers’ day-to-day banking requirements. These accounts are used to facilitate basic banking activity, such as receiving salaries, paying bills and other expenses, and thus would not be considered overly sensitive to liquidity stress events. As such, the criteria for ‘transactional accounts’ should be detailed enough to exclude deposits, such as savings accounts, which attract a higher run-off rate (assuming the other criteria for ‘stable’ deposits have not been met).
The results of the review indicate that tests applied by ADIs for ‘transactional account’ generally fell into two categories; a ‘product feature’ test (e.g. no specified term, multiple means of access to the account) and an ‘active use’ test (e.g. one or more transactions in the last month). Similar to the test for ‘established relationship’, many ADIs did not implement tests for ‘transactional accounts’ that were sufficiently stringent to appropriately segregate deposits.
APRA accepts that where a deposit is in a product designed to be a transactional account, and is non-interest bearing or pays a rate of interest significantly below similar products with the same ADI, then it may be appropriate to classify such accounts as ‘transactional accounts’. However, APRA considers that better practice is to incorporate both product features and an ‘active use’ test that considers not just the number but also the nature of transactions.
Less stable deposits
The LCR recognises that there are certain types of deposit accounts that demonstrate higher levels of liquidity risk than other deposit accounts. Deposits that fail the ‘stable’ deposit definition need to be considered against the criteria in the scorecard in paragraph 39 of Attachment A of APS 210.
The use of the internet and other mobile access channels for retail deposit accounts continues to rapidly evolve with most deposits accounts now having some form of on-line or mobile access. However, in looking at the 30-day LCR period, a customer seeking to withdraw funds from an at-call deposit account will most likely be able to do so by any means available for that account.
APRA expects ADIs to continue to assess whether the internet is integral to the design, marketing and usage of the product. However, unless the on-line features of the account are such that the propensity for withdrawal materially increases, such as where the primary means for withdrawal is via transfer to an external account at another ADI, APRA accepts that a small proportion of accounts may be defined as an ‘on-line account’.
APRA will continue to review developments and the associated depositor behaviour for deposit accounts with internet or other mobile access.
Heavily rate driven deposit
Depositors that have been attracted to an ADI primarily due to the interest rate on offer are more likely to be active in moving their funds among ADIs and are therefore considered to be more likely to withdraw their deposit in a liquidity stress event.
APRA noted a wide variety of interpretations of ‘heavily rate driven’, but in practice very few accounts were determined to meet these criteria.
In APRA’s view better practice is that any deposit product subject to the scorecard criteria that is currently in an ‘introductory’ or ‘promotional’ rate period - and where that rate will decrease after a set period - would be classified as ‘heavily rate driven’ for the duration of the promotional period.
APRA also noted that few ADIs identified any of their term deposits as heavily rate driven. However, some term deposit customers are very interest-rate sensitive and show little ‘stickiness’ based on factors other than the rate. ADIs should consider whether some of their term deposits are at times heavily rate driven, particularly if term deposit rates are increased for the primary purpose of attracting new customers or increased balances from existing customers. If an ADI determines that a term deposit is heavily rate driven, it should be classified as such for whatever period the term deposit impacts the LCR.
Finally, APRA noted that few ADIs benchmarked their offered rates against rates offered by peer competitors for the purposes of this classification and suggests that such benchmarking would constitute good practice.
APRA will continue to engage with individual ADIs on these matters as part of its ongoing supervision. In addition, APRA is considering whether any amendments to APS 210 and APG 210 are necessary.
Executive General Manager
Supervisory Support Division
1 Comprised of a set-format questionnaire and additional meetings where required.
2 For the purposes of the remainder of this letter, the term retail should be read to include qualifying SME customers/deposits.