Home » Capital » Increasing the threshold for applying internal rating to boost SME lending?

Increasing the threshold for applying internal rating to boost SME lending?

The Australian government has kicked off an inquiry into the banking system.   It will be chaired by David Murray the former CEO of Commonwealth Bank of Australia.  A draft terms of reference has been released and a number of interested parties have commented on it.

The Commercial Asset Finance Brokers Association of Australia Limited (CAFBA) submission caught my eye as it had a specific suggestion to change the regulatory capital requirements for banks to make it easier to lend to SMEs.  The CAFBA submission outlines a number of issues that potentially makes it harder for SMEs to borrow money.

APS 113 details the prudential requirements for banks who are accredited to use their internal models to calculate regulatory capital requirements.  According to CAFBA, banks have more onerous requirements for loans in excess of $1m, and hence are reluctant to lend over this amount.  CAFBA recommends lifting this threshold.

What is APS 113?

Australian Prudential Standard 113 details the requirements a bank has to meet to use an internal ratings based (IRB) approach.  It is APRA’s implementation of Basel’s IRB approach for Australia.

Under Basel II rules, there are essentially 3 approaches that banks can use to calculate capital requirements for credit risk: Standardised, Foundation IRB and Advanced IRB.  While standardised is easiest to apply, it is also more conservative.  On the other hand the Advanced IRB approach can result in lower levels of capital, however banks have to meet some requirements in order to have approval to apply it.

Under the advanced IRB approach, the bank determines key parameters of credit risk: Probability of Default (PD), Loss Given Default and Exposure at Default using its own models.  APS 113 details the processes, features, and governance requirements for the models.

How is retail treated differently to non-retail?

One key difference is that non retail exposures must be individually risk rated.  Risk ratings are similar to say credit ratings such as from credit rating agency like S&P.  Each bank will have their own rating scale with specific criteria and guidelines for assigning ratings.  When a new loan is made the bank has a process to assign a risk grade based the clients risk.  This assessment must be done independently to the division making the loan.  The rating must be reviewed at least annually or as soon as new information is received.  For a  business loan, this will mean reviewing the client and collecting updated financials annually.

Retail exposures on the other hand are statistically managed.  This means loans are pooled based on characteristics of the loan.  While the PD attached to the pool must be reviewed annually and the bank must ensure that the loan is assigned to the correct pool, this is less onerous than annual reviews.

Are the requirements for non-retail really that much more onerous?

Individual risk rating is time consuming.  I would guess that given the option, banks may prefer to statistically manage the client.  The margin on a loan around the $1m to $3m mark is not large, so I can see the attraction of statistically managing the client and not carrying out individual rating.

On the other hand, having a detailed review of the client will help to manage risk and price the loan better.

Will lifting the threshold have a material impact?

This measure alone will not change the situation.  However this should help when applied with other measures.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: