In a recent speech APRA Chair John Laker set out the basis for APRA’s approach to supervision of credit standards in housing lending by giving details of its comprehensive review of the debt serviceability policies of a large group of ADIs, which it followed up with a survey of all housing loan approvals by that group in September 2006. This analysis has given APRA a detailed insight into the riskiness of ADI housing loan portfolios and a solid ‘peer group’ basis to pursue any prudential concerns with individual lenders.

APRA reviewed the credit policy manuals of 47 ADIs, which account for around 80 per cent of total ADI housing loans. For banks, APRA was able to compare these policies to those from a similar study conducted by the Reserve Bank of Australia in 1998.

Phase one of APRA’s debt serviceability review confirmed that the net income surplus model is now the most common debt serviceability test used by ADIs. Ninety per cent of the ADIs reviewed use this model (some in conjunction still with the previous common debt servicing ratio of 30 per cent of the share of gross income).

Under the net income surplus model, the maximum loan amount is calculated on the basis of eligible sources of income, an estimate of basic living expenses, the interest rate, and any discount or margins applied to these components. APRA found that the methodology and assumptions used, and therefore the maximum loan amounts, vary significantly across ADIs, and more so than it expected.

In phase two, APRA invited those ADIs that participated in phase one to provide data on every housing loan approved during the month of September 2006.

APRA’s general assessment is that housing lending remains a sound asset class for its regulated lenders. However, the review has confirmed that housing lending portfolios have become somewhat more risky, whether measured in terms of higher LVRs or debt servicing burdens on borrowers.