Insights into ASIC’s Responsible Lending Hearings – Part 2

Insights from ASIC's Responsible Lending Hearings
Part 2

In Part 1 of my insights into ASIC's hearings into responsible lending, I noted that uncertainty remains over responsible lending expectations with regards to a customer's pre-loan and post-loan financial situation.

I continue this blog series here, publishing my key insights on the LIXI Blog.


There are vastly different opinions as to how effectively expense verification can be automated.

There was significant time dedicated by the commissioners to understanding the extent to which each participant had automated the use of transaction data from bank statements to verify living expenses.

The wide-ranging responses highlight a key issue faced by the industry. Some respondents indicated that they were able to perform expense verification in a purely automated manner from transaction data. Others indicated that verification remains mostly manual, and they are sceptical that transactions can be used to automatically verify living expenses.

The fundamental disconnect between the HEM analysis that is derived from the 'what and why' of a household's spending, and efforts to use transactions that provide the 'where' is challenging.

Many lenders use the Household Expenditure Measure (HEM) to evaluate the plausibility of a customer's declared expenses. The HEM was built upon a statistical analysis of the ABS Household Expenditure Survey (HES) which asked a sample of households to track how much they spent on a range of goods and services. The HES then overlays what the household uses those goods and services for. For example, spending at a hardware store for repairs to a primary residence is treated differently by HEM to the identical spend used to repair an investment property. In this situation, identical products being used to repair different properties has an impact on how that purchase is treated in determining the HEM.

Note that bank statement data can be extracted from bank statements directly, automated via a data aggregator, or obtained from a future Open Banking framework under the Consumer Data Right. In each case, however, the difference of opinion is more to do with how to use the data, not so much in how to get the data.

Transaction data from bank statements only provide short descriptions that provide very limited information. Merchant codes are provided for purchases, indicating where (i.e. which merchant) an amount was spent with, but not what was purchased.

In the case of the repairs, there is no available information that can be derived from the bank statement data to definitively know which property the spending was used for. There are many transactions that fall into this category where the transaction data cannot uniquely identify the HEM category into which that transaction maps. Cash withdrawals are even more challenging, providing no data at all with respect to how the cash was spent, or even if it was spent at all.

To map a transaction back to a HEM category with 100% certainty is not possible 100% of the time. A credit provider has two choices: either make assumptions about how to allocate spending, or engage further with the consumer to clarify the 'what and why' of a transaction.

Key Questions & Challenges

  • If assumptions are made in order to enable automatic mapping to HEM categories, do these assumptions undermine the verification?
  • Alternatively, after further engagement with the consumer (which is clearly a departure from complete automation), do the customer's responses need to be verified, and if so how?

Possible Outcome

The Commissioners' focussed their questioning in this area, and I am interested to see if ASIC provide some additional guidance around their expectations as to what constitutes an appropriate expense enquiry and verification. Another watch this space ...

Shane Rigby
CEO - LIXI Limited
22nd August 2019

Last Updated: 26th August 2019

See also: Part 1 & Part 3.